The Pan-Pacific Entomologist

EDITORIAL BOARD

J. T. Sorensen, Editor R. M. Bohart

R. V. Dowell, Associate Editor J. T. Doyen

R. E. Somerby, Book Review Editor J. E. Hafernik, Jr.

Julieta F. Parinas, Treasurer J. A. Powell

Published quarterly in January, April, July, and October with Society Proceed-

ings usually appearing in the October issue. All communications regarding non-

receipt of numbers, requests for sample copies, and financial communications

should be addressed to: Julieta F. Parinas, Treasurer, Pacific Coast Entomological

Society, Dept. of Entomology, California Academy of Sciences, Golden Gate Park,

San Francisco, CA 94118-4599.

Application for membership in the Society and changes of address should be

addressed to: Stanley E. Vaughn, Membership Committee chair, Pacific Coast

Entomological Society, Dept. of Entomology, California Academy of Sciences,

Golden Gate Park, San Francisco, CA 94118-4599.

Manuscripts, proofs, and all correspondence concerning editorial matters (but

not aspects of publication charges or costs) should be sent to: Dr. John T. Sorensen,

Editor, Pan-Pacific Entomologist, Plant Pest Diagnostics Center, California Dept.

of Food & Agriculture, 3294 Meadowview Rd., Sacramento, CA 95832-1448.

See the back cover for Information-to-Contributors, and volume 66(1): 1-8, Jan-

uary 1990, for more detailed information. Information on format for taxonomic

manuscripts can be found in volume 69(2): 194-198. Refer inquiries for publi-

cation charges and costs to the Treasurer.

The annual dues, paid in advance, are $25.00 for regular members of the Society,

$26.00 for family memberships, $12.50 for student members, or $40.00 for in-

stitutional subscriptions or sponsoring members. Members of the Society receive

The Pan-Pacific Entomologist. Single copies of recent numbers or entire volumes

are available; see 67(1): 80 for current prices. Make checks payable to the Pacific

Coast Entomological Society.

Pacific Coast Entomological Society

OFFICERS FOR 1995

Curtis Y. Takahashi, President Vincent F. Lee, Managing Secretary

Julieta F. Parinas, Treasurer Keve Ribardo, Recording Secretary

THE PAN-PACIFIC ENTOMOLOGIST (ISSN 0031-0603) is published quarterly by the Pacific

Coast Entomological Society, % California Academy of Sciences, Golden Gate Park, San Francisco,

CA 94118-4599. Second-class postage is paid at San Francisco, CA and additional mailing offices.

Postmaster: Send address changes to the Pacific Coast Entomological Society, c/o California Academy

of Sciences, Golden Gate Park, San Francisco, CA 94118-4599.

This issue mailed 31 March 1995

The Pan-Pacific Entomologist (ISSN 0031-0603)

PRINTED BY THE ALLEN PRESS, INC., LAWRENCE, KANSAS 66044, U.S.A.

This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

PAN-PACIFIC ENTOMOLOGIST

71(1): 1-2, (1995)

OBITUARY:

CELESTE GREEN, SCIENTIFIC ILLUSTRATOR,

1913-1994

E. GORTON LINSLEY AND JOHN CHEMSAK

Department of Environmental Science Policy and Management

University of California, Berkeley 94720

The literature of biology, especially entomology, lost one of its finest artists on

10 Jan 1994. Celeste Green provided accurate and beautiful pictorial support for

hundreds of new species of insects, primarily Coleoptera (Cerambycidae), but also

Diptera, Hemiptera and a few other orders. Her finest illustration, of a bee (Pro-

toxaea), a magnificent framed 8 x 10” painting, presented to EGL upon his re-

tirement, was unfortunately lost in the Oakland-Berkeley Hills Fire of 1992.

Celeste was born on 16 Nov 1913. Her early childhood was spent in Carmel,

California, where her father managed the famous Highland Inn Hotel. Her artistic

ability became evident early, and she won several local talent contests. Later, in

Oakland, she attended high school and upon graduation pursued her art training

at the California College of Arts and Crafts. Her first professional work was in

the field of fashion illustrating in Oakland, and later she taught this subject at the

prestigious Jean Turner Art School in San Francisco.

Since the mid-1950s, when she was employed by the Department of Ento-

mology, as Senior Scientific Illustrator, her many contributions have enriched

publications in The Pan-Pacific Entomologist, the University of California Pub-

lications in Entomology, the Bulletin of the California Insect Survey, the Pro-

ceedings of the California Academy of Sciences, and other journals, domestic and

foreign.

She also contributed examples of her work to a general text book on scientific

illustration and as chapter headings to an autobiography of R. L. Usinger. Un-

fortunately, the high cost of color illustration limited the use of her talent in this

field. However, two particularly fine examples may be seen: one of Plinthocoelium

suaveolens plicatum (LeConte), as the Frontispiece of Linsley (1964), and another

of Crossidius spp., as plates 1 and 2 of Linsley & Chemsak (1961).

Retirement, in 1980, took Celeste to Indiana to be closer to her children, and

in 1986 she moved to the northwest (Gig Harbor, Washington), where she re-

mained until her death. Her daughter, Kathleen Petrilli, gives this account of these

chapters in her life:

She first moved to Indiana, where she became fascinated in researching and

making corn husk dolls, i.e., doll-size figures made from dried corn husks.

Many of these dolls depicted pioneer life in mid-America. During this time,

Celeste was living in Carmel, Indiana and she created a replica of the historic

train station complete with corn husk figures. This train station was on

display in the city of Carmel for several months.

During this time in the mid-West, both Celeste and her sister Pat developed

a small antique/pottery mending business. They did repair work for local

antique dealers. Although they kept this business small, their reputation

quickly grew and there was continual demand for their skills and services.

2 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Figure 1. Celeste Green. 1967.

Her final move was to the Pacific Northwest and her life in her beloved

“brown house” on the waterfront of Gig Harbor, Washington.

This was the period when Celeste’s artistic interests turned to undersea life

and Native American Indians of the N.W. She began making fish and sea

creatures out of a clay-like substance called sculpey which could be baked

and hardened in the oven. In 1992, she created.'a complete coral reef depicting

undersea life in the Puget Sound region.

During her sojourn in the N.W., Celeste became interested in. Indian history

and folk lore particularly the history of Indian Kachina dolls. Both she and

her sister Pat created wonderful Indian masks and dolls that now adorn our

home. (For me, they truly embody the creative spirit that was my mother’s

special gift.)

But these tributes to her talent and artistic ability do not record Celeste’s

personality and humanity. She was a lovely individual, loved by those who knew

her well, and respected by those with less frequent contacts. Within the University, .

these included faculty, staff, and students—graduate and undergraduate. We had

the unique privilege of sharing a museum laboratory with her and we are honored

to have an opportunity to publish this tribute to her memory.

LITERATURE CITED

Linsley, E.G. 1964. The Cerambycidae of North America. Part V. Taxonomy and classification of

the subfamily cerambycinae, tribes Callichromini through Ancylocerini. Univ. Calif. Publ.

Entomol., 22.

Linsley, E.G. & J. A. Chemsak. 1961. A distributional and taxonomic study of the genus Crossidius

(Coleoptera, Cerambycidae). Misc. Publ. Entomol. Soc. Am., 3: 26-64.

PAN-PACIFIC ENTOMOLOGIST

71(1): 3-12, (1995)

CANOPY ARTHROPOD DIVERSITY IN A

NEW CALEDONIAN PRIMARY FOREST

SAMPLED BY FOGGING!

Eric GUILBERT, MICHEL BAYLAC, AND JUDITH NAJT?

Laboratoire d’Entomologie, Muséum National d’Histoire Naturelle,

Paris, France

Abstract.—We collected 9608 specimens during three fogging samples in a New Caledonian

rainforest. The most abundant arthropod taxa in number of specimen were Collembola, Diptera,

Coleoptera, Hymenoptera and Psocoptera. The fogging in New Caledonia indicated a high

percentage of Psocoptera and Diptera; however the effects of sample biases should be assessed

before any comparison can be drawn.

Key Words. —Arthropoda, biodiversity, canopy, fogging, New Caledonia

Fogging as a collecting technique was first employed in 1966 by Martin (Erwin

1983). Since then, numerous studies of canopy arthropods have used this tech-

nique: in Ontario (Martin 1966), Costa Rica (Roberts 1973), Hawaii (Gagné 1979),

Panama (Erwin & Scott 1980), South Africa and Great Britain (Southwood et al.

1982), Amazonia (Erwin 1983, 1989; Adis et al. 1984), Japan (Hijii 1983, 1986),

Borneo (Stork 1987, 1988, 1991), Sulawesi (Noyes 1989), Thailand (Watanabe

& Ruaysoongnern 1989) and Australia (Basset 1988, 1990, 1991). However, few

studies dealt with the arthropod fauna of West Pacific islands and no comparison

among arthropod faunas of the locations listed above has been attempted.

New Caledonia includes 3.5% of the area covered by extant tropical rainforests,

and is one of the 10 “‘hot spots” identified by Myers (1988) on the basis of its

species diversity and fragility. Studies of the fauna and flora from the Riviére

Bleue reserve by the Muséum National d’Histoire Naturelle and ORSTOM show

high endemicity (80% of animal species, 73% of plant species).

The present study determined the composition of the arthropod fauna from

the tree crowns in the rainforest of the Riviére Bleue Reserve sampled by insec-

ticidal fogging.

METHODS AND MATERIALS

Study Area.—Samples were gathered in the Riviére Bleue reserve (20° S, 166°

E), which has been a territorial park since May 1980. Geological and floristic

characteristics of the location were described by Bonnet de Larbogne et al. (1991).

The reserve is located in the great Southern massif of New Caledonia in the region

of the Haute Yaté basin. The study was carried out in evergreen forest on alluvium

in the flat bottom of the valley, 50 m from the river at a mean altitude of 160

m; this parcel is occasionally flooded.

Samples. —Fogging was carried out from the ground with a Dyna-fog Golden

eagle backpack 2980, using a mixture of 200 cc cyfluthrin in water and polyhydric

1 Authors’ page charges partially offset by a grant from the C. P. Alexander Fund, PCES.

2 CNRS et UC 1338 MNHN, 45, rue Buffon, F-75005 Paris.

4 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

alcohols (Solfac EW 050) in 2 liters Maxifog solvent. Each fogging lasted 15 min.

The collecting surface was arranged in six rows of five white plastic sheets of 1

m7, separated by 0.5 m, giving a total surface area of 60 m?. One half was placed

on the soil and the other half on stands, 50 cm above the soil. The specimens

that fell on the sheets were collected either by brushing (first fogging) or by washing

with water (second and third fogging).

Three fogging samples were collected: the first (Fog1) and second (Fog2) in the

same place, and the third (Fog3) 100 meters away. The first fogging (18 Jan 1991,

20° C, hygrometry: 90%) took place during the rainy season, the second (23 Jul

1991, 22° C, hygrometry: 85%) and third ones (1 Aug 1991, 20° C, hygrometry:

90%) during the dry season. Fog3 was carried out one week after Fog2. The aim

was to sample the biotope, not particular tree species.

All arthropods were sorted at least to order level, and to family level for the

most abundant orders: Hymenoptera, Diptera, Coleoptera in Fog] and Fog3 and

Orthoptera, Hemiptera in Fog!. When taxonomic expertise was available and in

order to allow ecological and trophic precision, some families from Fogl were

sorted into subfamilies and tribes (Staphylinidae), species (Formicidae) and tro-

phic guilds (Cecidomyiidae). The Collembola were sorted to species in the three

foggings.

RESULTS

Faunistic Composition. —Over the three foggings, 9608 specimens were col-

lected (106.76 specimens/m7) belonging to 107 families within 20 orders: 2227

in Fogl (74.23/m7), 3191 in Fog2 (106.37/m?) and 4190 in Fog3 (139.67/m7).

These seem low when compared with other fogging samples: 35 to 161/m? in

Amazonian samples (Adis et al. 1984), and 51 to 218/m? in Bornean samples

(Stork 1991). Hii (1983) found 200 to 3500/m? by the smoking method. The

lower abundance in Fog] within our samples is due to the collection method.

When the three samples are considered together, the dominant orders are Col-

lembola (18.6% of the specimens), Diptera (18.4%), Coleoptera (13.6%), Hyme-

noptera (12.7%) and Psocoptera (11.4%). The frequencies of the other orders do

not exceed 7%. Nematocera are more abundant than Brachycera and the majority

of Hymenoptera were Formicidae (Table 1). Diptera and Coleoptera are repre-

sented respectively by 27 and 31 families; Hymenoptera are represented by at

least 14 families (Chalcidoidea were considered a single taxon).

As reflected by the standard deviation of their abundance in subsamples within

each fogging (Table 2), some abundant groups were not distributed evenly across

the sample surface: Coleoptera, Nematocera, Collembola, and Psocoptera. For-

micidae are patchy in Fog! and Fog2, Nematocera in Fog2 and in Fog3, Acarina

in Fogl and Fog3, Thysanoptera in Fog2 and Homoptera in Fog3. Other groups,

which are less abundant, are well spread throughout the collecting surface: Lep-

idoptera in the three samples, Acarina, Amphipoda, Dermaptera in Fogl and

Fog2, Orthoptera in Fog2 and Fog3. Among samples, some groups vary greatly

in abundance: Collembola, Acarina and Nematocera; while others are quite con-

stant: Hymenoptera, Brachycera, Lepidoptera and Dermaptera.

The abundance of Collembola varies greatly among foggings, from 6.3% to 24%

to 21% of the sampled insects. A total of 29 species of Collembola were collected.

The number of species among samples is nearly the same; but only nine species

1995 GUILBERT ET AL.: CANOPY ARTHROPODS 5

Table 1. Absolute and relative frequencies of individuals belonging to major taxa collected by

fogging.

Fogging 1 Fogging 2 Fogging 3 Total fogging

Taxa “Nber%~—S—™*é<“<‘«<é“Ner~~™~*=<“St*é<i=i‘<‘<«~NC:C(“‘é‘“SS””CN CU”

1. Other Hymenoptera 144 6.5 188 5.9 197 4.2 529 5.5

2. Formicidae 126 se 471 14.8 94 24) 691 Te

3. Nematocera 612 2a 323 10.1 299 6.4 1234 12.8

4. Brachycera oe. 3.4 72 5.4 253 5.4 500 <a

5. Heteroptera 63 2.8 61 1.9 49 Ted 173 1.8

6. Homoptera 40 1.8 64 2.0 233 5.0 337 3.5

7. Orthoptera 31 1.4 17 0.5 26 0.6 74 0.8

8. Coleoptera 269 12.1 301 9.4 736 15.8 1306 13.6

9. Araneae 197 8.8 174 5.5 Zit 6.0 648 6.7

10. Collembola 140 6.3 766 24.0 885 19.0 1791 18.6

11. Lepidoptera 10 0.4 17 0.5 31 0.7 58 0.6

12. Psocoptera 392 17.6 428 13.4 275 5.9 1095 11.4

13. Thysanoptera 28 L2 65 2.0 145 3.1 238 25

14. Dictyoptera 42 1.9 9 0.3 81 L,7 132 1.4

15. Acarina 15 0.7 102 3.2 413 8.9 530 5.5

16. Amphipoda 2 0.1 5 0.2 30 0.6 37 0.4

17. Others 7 0.3 6 0.2 18 0.4 31 0.3

18. Pseudoscorpionida 1 0.04 0 0.0 44 0.9 45 0.5

19. Dermaptera 1 0.04 5 0.2 17 0.4 23 0.2

20. Larvae undetermined 32 1.4 17 0.5 87 1.9 136 1.4

21. Total _ 2227 = 100 3191 100 4190 100 9608 100

are common to the three samples, and three of them are abundant (Table 3). Four

species are abundant in Fog2 and Fog3, but not in Fog1. The scarcity and absence

of some species in the Fog! may be due either to seasonal variation or to collecting

technique. So the occurrence and the abundance of Rastriopes fuscus Yosii and

R. sp. could characterize respectively Fog2 and Fog3 or could be a consequence

of the sampling method or seasonal variation.

The low percentage of Collembola from Fogl may be an artifact related to the

collecting technique: according to Hijii (1983), one should wash the sheets with

appropriate liquids after fogging in order to avoid the loss of microarthropods

such as Collembola and Acarina. However, the percentage in Fogl is high in

comparison with the results of previous studies, and could be due to contamination

of the sample by ground-dwelling Collembola jumping onto the sheets during

fumigation and collection. However, Hijii (1983) and Watanabe & Ruaysoongnern

(1989) found a great abundance of Collembola in Japan and Thailand respectively.

Watanabe & Ruaysoongnern (1989) suggested that the great number of Collembola

is related to seasonal flooding of the forest floor. Most Collembola found in fogging

samples may also come from tree trunks, and/or from organic matter in the canopy

(Stork 1988, Nadkarni & Longino 1990).

When comparing Collembola fauna from canopy and soil, their average richness

is similar in the fogging samples (18 species) and in malaise trap sample (19

species) (Guilbert, unpublished data). Similarity in the ground and canopy fauna

has also been observed in Costa Rica (Nadkarni & Longino 1990). However,

Martin (1966) found 22 soil species compared to 10 species from crown stratum

6 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Table 2. Simple statistics within samples of the major taxa sorted by sheets.

Taxa Range Sum Mean SD

Fogl

Hymenoptera 0-9 144 4.80 2.63

Formicidae 1-17 126 4.20 3.45

Nematocera 7-40 612 20.40 7.29

Brachycera 0-5 gi 2.50 1.55

Heteroptera 0-7 63 2.10 [AA

Homoptera Q-7 40 1.33 1.52

Orthoptera 0-5 31 1.03 1.30

Coleoptera 0-68 269 8.97 12.42

Araneae 1-21 197 6.57 4.04

Collembola 0-22 140 4.67 5.02

Lepidoptera 0-2 10 0.33 0.61

Psocoptera 1-54 . 392 13.07 12.74

Thysanoptera 0-4 28 0.93 1.01

Dictyoptera 0-6 42 1.40 1.61

Acarina 0-3 15 0.50 0.82

Amphipoda 0-1 2 0.07 0.25

Pseudoscorpionida 0-1 l 0.03 0.18

Dermaptoptera 0-1 l 0.03 0.18

Larvae 0-4 32 1.07 Lig

Fog2

Hymenoptera 1-14 188 6.27 3.05

Formicidae 0-345 471 15.70 62.33

Nematocera 4-18 323 10.77 3.70

Brachycera 0-49 172 5.73 8.62

Heteroptera 0-10 61 2.03 2.11

Homoptera 0-14 64 2.13 3.03

Orthoptera 0-3 17 0.57 0.86

Coleoptera 3-24 301 10.03 5.05

Araneae 0-10 174 5.80 2.82

Collembola 0-73 766 25.53 19.30

Lepidoptera 0-2 17 0.57 O77

Psocoptera 0-30 428 14.27 7.58

Thysanoptera 0-20 65 217 3.60

Dictyoptera 0-3 9 0.30 0.65

Acarina 0-9 102 3.40 2.79

Amphipoda 0-2 7 0.17 0.46

Pseudoscorpionida 0-0 0 0 0

Dermaptoptera 0-3 5 0.17 0.59

Larvae 0-3 wi 0:57 0.77

Fog3

Hymenoptera 2-14 197 6.57 2.74

Formicidae 0-11 94 3.13 2.67

Nematocera 4-24 299 9.97 4.54

Brachycera 2-23 253 8.43 5.24

Heteroptera 0-6 49 1.63 1.67

Homoptera 3-16 233 7.77 3.62

Orthoptera 0-3 26 0.87 0.97

Coleoptera 8-46 736 24.53 6.87

Araneae 2-26 277 9.23 5.26

Collembola 4-55 885 29.50 11.61

Lepidoptera 0-4 31 1.03 0.93

1995 GUILBERT ET AL.: CANOPY ARTHROPODS i

Table 2. Continued.

Taxa Range Sum Mean SD

Psocoptera 1-22 275 9.17 4.56

Thysanoptera 0-11 145 4.83 3.01

Dictyoptera 0-11 81 2.70 255

Acarina 2-46 413 13.77 9.28

Amphipoda 0-7 30 1.00 172

Pseudoscorpionida 0-5 44 1.47 1.48

Dermaptoptera 0-6 17 0.57 1.45

Larvae 0-9 87 2.90 a7

in Ontario, and Hijii (1989) found a greater abundance of specimens of Collembola

in the soil than in the canopy in Japan.

Two epiedaphic groups of Collembola, the Entomobryomorpha and Symphy-

pleona, are very abundant in both New Caledonian canopy fogging and malaise

trap samples. Most species belonging to these taxa found in our samples exhibit

long claws, which probably constitute an adaptive character and when associated

with their large ecological valency, allows invasion of all above-ground habitats.

Nevertheless, some species probably live in epiphytic plants or suspended soil

and may be restricted to the canopy (e.g., Pseudoparonella (Pseudoparonella) sp.,

Lepidosira (Nusasira) sp. and Dicyrtomina sp. 1 that occur in the fogging but not

in the malaise trap samples).

Diptera represent 14.9% of specimens of the entire study. Nematocera are more

abundant and diverse than the Brachycera. Ceratopogonidae, Chironomidae, Cec-

idomyliidae and Sciaridae are dominant and constitute 91.2% of Nematocera.

Sixteen families of Brachycera were found, of which three accounted for 80.2%

of the Brachycera individuals: the Chloropidae, Dolichopodidae and Drosophil-

idae.

With 27 families of Diptera, the canopy of the Riviére Bleue forest is poorer,

particularly in Brachycera than forest in Ontario (44 families: Martin 1966); but

it is richer than Queensland (12 families, mostly Nematocera: Basset 1991). The

abundance of Diptera is greater in our samples (13.2 to 30.8%) than anywhere

else (around 20%) except for Borneo (48 to 83%; Stork 1991). Family frequency

depends on their activity around the tree: some flies pass through the sample area

and are caught by the insecticidal cloud. Other families such as Phoridae, Chlo-

ropidae and Dolichopodidae are typically arboreal (Basset 1991); the occurrence

of some families such as the Phoridae in only one of the two sites sampled may

be related to differences in floristic composition. Greater abundance of Cerato-

pogonidae, Chironomidae, Cecidomyiidae and Sciaridae in Queensland, New-

Caledonian and Bornean samples may due to the proximity of rivers or water

spots (Stork 1991) as Riviére Bleue samples were collected 50 meters from a river.

The Cecidomyliidae belong to three different trophic guilds: predators, phyto-

phages and mycetophages (Table 4). Mycetophage species (that belong mainly to

the two primitive subfamilies Lestremiinae and Porricondylinae) are absent from

the canopy, whereas they constitute 4.49% of malaise trap sample (EG, unpub-

lished data). Larvae from the latter subfamilies are known to inhabit soil litter,

mosses and dead wood. Their absence in the fogging samples is striking and should

8 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Table 3. Number of specimens of Collembola per species and per collection obtained by fogging

(Fogl, Fog2, Fog3).

Species Fog] Fog2 Fog3

Neanuridae

Pseudachorutes tillieri Najt & Weiner 0 0 0

Hypogasturidae

Xenylla sp. 0 5 0

Isotomidae

Proisotoma sp. 0 1 0

Cryptopygus sp. 0 3 10

n.g. 1 0 0 Z

Entomobryidae s.1.

Pseudoparonella (Plumachaeta) oceanica Yoshii 4 0 18

Epimetrura sp. 29 127 124

Willowsia sp. 2 11 13

Salina (Salina) oceanica Yoshii 63 86 295

Pseudoparonella (Oceaniella) shibatai Y osii 16 13 96

Pseudoparonella (Pseudoparonella) sp. 1 1 6 8

Pseudoparonella (Pseudoparonella) novae-caledoniae Yosii 0 5 34

Lepidocirtoydes novae-caledoniae Y osii 0 0 5

Pseudoparonella (Oceaniella) griseocoerulea Yoshii 8 0 4

Pseudosinella (Austrocyrtus) speciosa Y oshii 0 0 0

Lepidosira (Nusasira) vicina Yoshii 2 11 15

Seira sp. 2 0 0

Pseudoparonella (Pseudoparonella) sp. 2 0 0 4

Lepidosira (Nusasira) sp. 0 6 4

Lepidosira sp. 1 0 0 3

Lepidosira sp. 2 0 1 0

Pseudoparonella (Oceaniella) bicincta Y oshii 1 0 0

Entomobrya (Entomobrya) sp. 1 0 0

Sminthuridae

Parasphyrotheca sp. 1 15 15

Sphaeridia sp. 0 0 0

Sphyrotheca sp. 1 1 27 0

Sphyrotheca sp. 2 2 0 0

Bourletiellidae

Bourletiellidae n.g. 0 l 0

Rastriopes fuscus Yosii 4 128 0

Rastriopes sp. 0 0 119

Dicyrtomidae

Dicyrtomina sp. 1 1 37 10

Dicyrtomina sp. 2 2. 283 106

Total 140 766 885

be further investigated. It could be related either to seasonal variations or to the

canopy microclimate. The samples are also characterized by the reverse propor-

tions in the two major predator genera (Table 4). This result is similar to obser-

vations in Europe where Jrisopsis is mainly confined to the forest litter while

Lestodiplosis has a broader ecological valence (MB, unpublished data). The phy-

1995 GUILBERT ET AL.: CANOPY ARTHROPODS 9

Table 4. Trophic guilds of Cecidomyiidae from canopy (Fog1) and ground layers (malaise trap).

Trophic guild Taxa Fogging percentage Malaise percentage

Predators Cecidomyiinae

Trisopsis 9.24% 28.41%

Lestodiplosis 14.55% 5.98%

Mycophages Porricondylinae 0.% 1.09%

Lestremiinae

Micromylini spp. 0.% 3.08%

Lestremiini spp. 0.% 0.32%

Phytophages Cecidomyiinae 76.21% 61.12%

Total specimens 149. 1556.

tophagous Cecidomyiidae represent 76.5% in the canopy, vs. 61.1% at the soil

level.

The Coleoptera are the most diversified order in terms of number of families

(31). Only 16 families are present in both Fog! and Fog3; most of the other

families are represented by fewer than five specimens each in a single sample.

There were fewer families than in Queensland (54 families: Basset 1991) or in

Amazonian samples (57 families: Erwin 1983). As in those regions and in Panama,

the Curculionidae, Corylophidae, Chrysomelidae and Staphylinidae are the most

abundant (Basset 1991, Erwin 1983, Erwin & Scott 1980). Pselaphidae are more

abundant in our samples than in any other published results. Coccinellidae are

more abundant than in other regions (Basset 1991, Erwin & Scott 1980). In

contrast to Queensland and Zaire, no Scolytidae and few Cerambycidae were

found (Basset 1991, Sutton & Hudson 1980).

Most Hymenoptera collected belong to Parasitica and Formicidae. Formicidae

are dominant with 7.2%, whereas other Hymenoptera represent 5.5% of the three

fogging samples. Very few Symphyta are present, all of them belonging to the

Sphecidae. Formicidae constitute 5.7, 14.8 and 2.0% of the specimens in each

fogging sample. They are represented by 14 species and 7 genera, 11% of the

genera cited by Taylor (1987) for the whole territory. Paratrechina, with 42.2%

and Camponotus, with 16.4% of the total Formicidae in the Fog1, are the most

abundant genera. The presence of Pheidole, which is terricolous, reflects the oc-

currence of epiphytes and suspended soil in the trees. Although not cited by Taylor

(1987) and up to now unknown in New Caledonia, Crematogaster has been found

in the fogging samples and, therefore, occurs in the canopy. Representatives of

Iridomyrmex, Monomorium and Adelomyrmex are scarce.

Formicidae are less abundant here than in Amazonian or Bornean samples

(Erwin 1983, Adis et al. 1984, Stork 1988). As in Queensland, the paucity of ants

is more typical of temperate forests than of tropical ones (Basset 1991). Parasitica

like Scelionidae are usually abundant in tropical samples (Basset 1991, Noyes

1989), but are poorly represented here; because they reflect the abundance of their

host taxa in the canopy, the latter are probably scarce in New Caledonia.

The Psocoptera, although usually scarce in fogging samples (1.9 to 12.1% in

previous studies), represent 11.4% of the three samples. Psocoptera abundance is

higher in Japan and Ontario (Hijii 1986, Martin 1966). They are more frequently

found on tree trunks than in the canopy. Stork (1987) suggested that the distance

10 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

of the sheets from the tree trunks may considerably influence Psocoptera collection

efficiency, and their abundance in the Riviére Bleue rainforest may indicate a high

density of tree trunks when compared with other forests.

Eleven families of Araneae were found. The Clubionidae, Lyniphiidae, Saltic-

idae and Pholcidae are the most abundant with respectively 29.4%, 19.8%, 12.7%

and 12% of the specimens. No other family exceeds 3.5%. The greater abundance

of Araneae here than in other regions of the world may be related to the reduced

abundance and richness of Formicidae, which may supplant the spiders in the

canopy (Majer 1990). However, Stork found 21 families in Bornean samples (Stork

1991). Fewer Salticidae (6.6%) and more Thomisidae (38%) were recorded in

Ontario (Martin 1966). Theridiidae (1%) are less abundant than in Queensland

(31.2%: Basset 1991) and Borneo (22.9%: Stork 1991).

The Acarina are more abundant in Fog2 and Fog3, possibly because the sheets

from these foggings were washed. With 5.5% of all the specimens from fogging

samples, they are more abundant than in most other regions. However, they were

recorded as a dominant group in Japan (Hijii 1983, 1986), in Ontario (Martin

1966) and in canopy organic matter in Costa Rica (Nadkarni & Longino 1990).

Other arthropod orders, Lepidoptera, Dermaptera and Neuroptera, are scarce

in the three fogging samples. Some groups, such as Thysanoptera, Pseudoscor-

pionida, Amphipoda, are scarce in Fog! and Fog2 and abundant in Fog3. This

variation may be related to seasonal flooding, as observed by Watanabe & Ruay-

soongnern (1989) for Thysanoptera in Thailand, or to the floristic composition

of the canopy.

DISCUSSION

New Caledonian Fauna Composition. —The abundance of Collembola and Ac-

arina varies strongly among samples from other locations (see literature cited

above). This variation may be related to canopy cover, presence of suspended

soil or to the method used by the authors. These taxa are usually not abundant

when collected by fogging. Acarina were the most abundant group collected by

hand by Nadkarni & Longino (1990). According to some authors, variations of

abundance and composition of soil faunal taxa such as Collembola are due to

transfer between canopy and soil layers (Adis et al. 1984).

The abundance of Hymenoptera, Heteroptera, Orthoptera, Araneae, Lepidop-

tera and Psocoptera exhibits little variation in the three Caledonian samples. They

seem to be unaffected by changes in the composition of the flora between the two

sites, or by seasonal variations. These taxa are more abundant in the dry season

in Thailand (Watanabe & Ruaysoongnern 1989). According to Basset (1991),

some arthropod taxa are more abundant during flowering and leaf flush in Queens-

land.

Our observations are based upon three samples only and major taxa and sam-

pling biases may be expected. For example, the abundance of Formicidae in the

Fog? is due to an aggregation of one species which represents 73% of all Formicidae

in this fogging. Stability in abundance of major taxa could hide variations of

composition inside these taxa. Further sampling is needed to test these obser-

vations.

1995 GUILBERT ET AL.: CANOPY ARTHROPODS 11

CONCLUSIONS

Riviére Bleue’s fauna shows differences with other faunas in the world at family

levels. In addition, it differs from faunas of other forests in New Caledonia (EG,

unpublished data) and at the species level (Collembola) between two locations in

the same forest. The arthropod fauna of New Caledonia is characterized by the

importance of the Psocoptera and of the Diptera. It is similar to the fauna of

Queensland because of the importance of the Psocoptera, and of the Bornean

fauna, which is characterized by the importance of the Diptera.

Composition of the sample is influenced by multiple variables of the sampling

method: escape of strong fliers from the insecticidal cloud (Noyes 1989), death

of insects under bark and inside epiphytes without falling down, position of the

sheets laying on the ground or suspended, collecting technique, position of the

fogger and collecting time. All these factors possibly contribute to obfuscate general

patterns that may have historical (i.e., biogeographical), as well as functional (i.e.,

ecological), explanations. Fogging and sorting methods should be standardized in

order to eliminate these sources of noise and to increase significance of observed

patterns.

ACKNOWLEDGMENT

Jean Chazeau and Lydia Bonnet de Larbogne (ORSTOM Nouméa) have sam-

pled the canopy fauna and assisted in every aspect of field work. Nicole Berti,

Thierry Bourgoin, Loic Matile, Christine Rollard, Claire Villemant, and Janine

Weulersse (Muséum national d’Histoire naturelle) kindly provided their taxo-

nomical assistance. Simon Tillier (Muséum national d’Histoire naturelle) man-

aged the program: “Evolution et vicariance en Nouvelle-Calédonie”’ which made

this study possible and supported the author while preparing this manuscript.

LITERATURE CITED

Adis, J., Y. D. Lubin & G. Montgomery. 1984. Arthropods from the canopy of inundated and Terra

firme forest near Manaus, Brazil, with critical considerations on the pyrethrum-fogging tech-

nique. Studies Neotr. Fauna Environ., 19: 223-236.

Basset, Y. 1988. A composite interception trap for sampling arthropods in tree canopies. J. Aust.

Entomol. Soc., 27: 213-219.

Basset, Y. 1990. The arboreal fauna of the rainforest tree Argyrodendron actinophyllum as sampled

with restricted canopy fogging: composition of the fauna. The Entomologist, 109: 173-183.

Basset, Y. 1991. The taxonomic composition of the arthropod fauna associated with an Australian

rainforest tree. Aust. J. Zool., 39: 171-190.

Bonnet de Larbogne, L., J. Chazeau, A. Tillier & S. Tillier. 1991. Milieux naturels néo-calédoniens:

la Réserve de la Riviére Bleue. pp. 9-17. Jn Chazeau, J. & S. Tillier (eds.). Zoologia Neoca-

ledonica, 2. Mém. Mus. natn. Hist. nat., Paris.

Erwin, T. L. 1983. Beetles and other insects of tropical forest canopies at Manaus, Brazil, sampled

by insecticidal fogging. pp. 59-79. In Sutton, S. L., et al. (eds.). Tropical rainforest ecology and

management. Blackwell Press, Oxford.

Erwin, T. L. 1989. Canopy arthropod biodiversity: a chronology of sampling techniques and results.

Rev. Entomol. Peruana, 32: 71-77.

Erwin, T. L. & J. C. Scott. 1980. Seasonal and size patterns, trophic structure and richness of

Coleoptera in the tropical arboreal ecosystem: the fauna of the tree Luehea seemannii Triana

and Planch in the canal zone of Panama. Coleopt. Bull., 34: 305-321.

Gagné, W.C. 1979. Canopy-associated arthropods in Acacia koa and Metrosideros tree communities

along an altitudinal transect on Hawaii island. Pac. Ins., 21: 56-82.

12 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Hijii, N. 1983. Arboreal arthropod fauna in a forest: seasonal fluctuations in density, biomass, and

faunal composition in a Chamaecyparis obtusa plantation. Jap. J. Ecol., 33: 435-444.

Hijii, N. 1986. Density, biomass, and guild structure of arboreal Arthropod as related to their

inhabited tree size in a Cryptomeria japonica plantation. Ecol. Res., 1: 97-118.

Hijii, N. 1989. Arthropod communities in japanese cedar (Cryptomeria japonica D. Don) plantation:

abundance, biomass and some properties. Ecol. Res., 4: 243-260.

Majer, J. D. 1990. The abundance and diversity of arboreal ants in northern Australia. Biotropica,

22: 191-199.

Martin, J. L. 1966. The insect ecology of red pine plantations in central Ontario. IV. The crown

fauna. Can. Entomol., 98: 10-27.

Myers, N. 1988. Threatened biotas: “hot spots” in tropical forests. The Environmentalist, 8: 187-

208.

Nadkarni, N. M. & J. T. Longino. 1990. Invertebrates in canopy and ground organic matter in a

neotropical montane forest, Costa Rica. Biotropica, 22: 286-289.

Noyes, J.S. 1989. A study of five methods of sampling Hymenoptera (Insecta) in a tropical rainforest,

with special reference to the Parasitica. J. Nat. Hist., 23: 285-298.

Roberts, H. F. 1973. Arboreal Orthoptera in the rain forest of Costa Rica collected with insecticide:

a report of the grasshoppers (Acrididae) including new species. Proc. Acad. Nat. Sc., Philadel-

phia, 125: 46-66.

Southwood, T. R. E., V. C. Moran & C. E. J. Kennedy. 1982. The richness, abundance and biomass

of the arthropod communities on trees. J. Animal. Ecol., 51: 635-649.

Stork, N. E. 1987. Arthropod faunal similarity of Bornean rain forest trees. Ecol. Entomol., 12: 219-

226.

Stork, N. E. 1988. Insect diversity: facts, fiction and speculation. Biol. J. Linn. Soc., 35: 321-337.

Stork, N. E. 1991. The composition of the arthropod fauna of Bornean lowland rain forest trees. J.

Tropical Ecol., 7: 161-180.

Sutton, S. L. & P. J. Hudson. 1980. The vertical distribution of small flying insects in the lowland

rain forest of Zaire. Zool. J. Linn. Soc., 68: 111-123.

Taylor, R. W. 1987. A checklist of the ants of Australia, New Caledonia and New Zealand (Hy-

menoptera: Formicidae). CSIRO Aust. Div. Entomol. Rep., 41: 1-92.

Watanabe, H. & S. Ruaysoongnern. 1989. Estimation of arboreal arthropod density in a dry evergreen

forest in northeastern Thailand. J. Trop. Ecol., 5: 151-158.

PAN-PACIFIC ENTOMOLOGIST

71(1): 13-17, (1995)

FEEDING AND PREY PREPARATION IN THE SOLPUGID,

EREMORHAX MAGNUS HANCOCK

(SOLPUGIDA: EREMOBATIDAE)

FRED PUNZO

Department of Biology, University of Tampa,

Tampa, Florida 33606

Abstract.—Prey preparation as an important component of handling time is demonstrated for

the first time in a solpugid (Eremorhax magnus Hancock). Prey body parts (from the grasshopper,

Trimerotropis pallidipennis Walker) characterized by high chitin content (head, antennae, wings,

legs) are selectively removed prior to ingestion. Head capsules were removed in 77-84% of the

feeding trials, depending on the size of the prey, followed by forewings (54%) and hindwings

(37%). Body parts possessing lower amounts of chitin (abdomen, thorax, hind femur) are pro-

cessed and ingested thereby supporting the nutrient concentration hypothesis. Prey is initially

detected via the palpi which are then used to pull the prey toward the chelicerae. The prey is

then grasped by the chelicerae which are then used to fragment and grind the prey for ingestion.

Ingestion time ranged from 6.2—17.4 min for small hoppers, and 11.6—28.3 min for larger prey.

Key Words.—Arachnida, Solpugida, Eremorhax, prey preparation, feeding

Previous studies have shown that predators frequently consume only certain

parts of their prey (Haynes & Sisojevic 1966, Sih 1980) and often show strong

preferences for specific tissues and body regions (Curio 1976; Punzo 1989, 1992).

For example, insectivorous birds frequently remove the wings, legs and head

capsule, and swallow the thorax and abdomen (Sherry & McDade 1982). Some

lycosid and thomisid spiders preferentially ingest the softer tissues of an insect’s

abdomen while rejecting other body regions depending on the degree of hunger

(Haynes & Sisojevic 1966, Nentwig 1987, Punzo 1991). In many cases insectiv-

orous birds and mammals will modify or remove specific prey parts before in-

gestion is initiated (Curio 1976). This has also been reported for a few arthropod

predators such as mantids and some decapod crustaceans (Krebs & McCleery

1984). Although this type of behavior, known as prey preparation, increases the

overall handling time, it can help to optimize energy budgets by targeting the

ingestion of those body parts possessing a higher concentration of essential nu-

trients (Hespeheide 1973, Kaspari 1990). One way for insectivores to maximize

nutrient intake rate would be to reject those prey parts having a high chitin content.

Chitin is either indigestible or poorly digested by insectivores in general (Punzo

1989, Scott et al. 1976).

Research on optimal foraging has focused on energy expenditure associated

with search, pursuit, capture, ingestion and resource depression (Charnov 1976,

Lucas 1983, Punzo 1989, Punzo & Garman 1989) whereas prey preparation has

received little attention (Kaspari 1990). The few available studies focus on ver-

tebrate predators (see reviews by Curio 1976, Krebs & McCleery 1984, O’Brien

et al. 1990). In this paper, I explore the relationship between chitin content and

prey preparation in the solpugid, Eremorhax magnus (Hancock). This is the first

demonstration that solpugids make decisions concerning which prey parts should

be selectively consumed.

14 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

MATERIALS AND METHODS

Eremorhax magnus is a common inhabitant of the desert regions of southern

California (Muma 1951). Adult females (29-35 mm, total body length). were

collected as they wandered the surface at night during June-August, 1992. Sol-

pugids were collected within a 10-km radius of Victorville (San Bernardino Coun-

ty), CA. A helmet-mounted light with a red filter was used to locate and observe

solpugids as described by Punzo (in press). A total of 56 females were collected

and transported back to the laboratory. Solpugids were housed individually in

plastic cages (30 x 14 x 8 cm), provided with water, .and fed once. per week on

a diet of mealworm larvae (Tenebrio molitor L.).

Twenty solpugids were randomly assigned to. one of two experimental groups.

Each experimental group was allowed to feed on one of two prey size classes: (1)

small (juveniles): total body length (TBL): 12-16 mm; body weight (BW): 0.31

+ 0.02 g; (2) large: TBL: 17-22 mm; BW: 0.84 + 0.03 g. I chose pallid-winged

grasshopper females (Trimerotropis pallidipennis Walker) as the prey species for

all feeding experiments. This grasshopper is common in this area (personal ob-

servation) and three of the solpugids had a pallid-winged grasshopper in their

chelicerae when they were collected. Specimens of 7. pallidipennis were collected

with a sweep net and also brought back to the laboratory for subsequent use in

feeding trials. All solpugids were deprived of food for 72 h prior to testing.

Grasshoppers from each prey size class were used to assess the chitin content

(mean weight and percent chitin) of various body parts: head, antennae, abdomen,

hind femur, foreleg, midleg, thorax, forewing and hindwing. Chitin weight was

determined according to the method described by Zach. & Falls (1978). Body parts

were freeze-dried, weighed on a Metler electronic analytical balance, immersed

in 2.0 M KOH for 72 h, rinsed, dried again and reweighed. KOH dissolves all

tissues except chitin.

For feeding trials, each solpugid was presented with a grasshopper from one of

the designated size classes. Feeding trials were recorded with a Cine-8 High Speed

Camera (Visual Instrumentation Corp.) at 100 frames/sec. A Lafayette Super 8

Analyzer (Model 1026) was used for frame-by-frame analysis as described by

Punzo (1989). I recorded the removal time (sec), defined as the amount of time

that elapsed from the moment the prey was grasped until a particular body part

was detached.

I used the data recorded for chitin content to determine whether or not there

was any evidence of nutrient concentration. According to the nutrient concentra-

tion hypothesis (Foster 1987, Kaspari 1990), the removal of prey parts possessing

high amounts of indigestible chitin (prey preparation) should result in the con-

centration of utilizable nutrients while maximizing the amount of space in the

gut available for additional food items. I calculated the difference in nutrient

concentration when a particular body part was removed from the grasshopper

using the data collected on chitin content. The various body parts were subse-

quently ranked by dividing the mean removal time for each body part by its chitin

content as described by Kaspari (1991).

Statistical analyses followed procedures described by Sokal & Rohlf (1981).

Prey-part rankings were obtained by Tukey’s multiple comparison test; this yield-

ed statistical clusters of body parts. These clusters related to predicted perfor-

1995 PUNZO: EREMORHAX MAGNUS PREY PREPARATION 15

Table 1. Mean chitin weight (mg) and percent chitin (%) of several body parts for two size classes

of the grasshopper, Trimerotropis pallidipennis.

Grasshopper size class

Small Large

Body part n Mean weight SD % n Mean weight SD %

Head 10 2.34 0.32 37.7 9 5.85 0.84 40.4

Antennae 10 0.13 0.02 41.2 10 0.32 0.03 43.6

Abdomen 9 ry 3 0.41 0.7 8 6.47 0.72 12.3

Thorax 10 0.51 0.14 21.1 10 1.24 0.31 25.2

Hind femur 10 1.91 0.17 17.4 9 2.87 0.26 18.3

Foreleg 10 0.31 0.04 43.4 10 0.54 0.07 41.2

Midleg 8 0.20 0.02 35.4 10° 0.42 0.03 39.7

Front wing 8 1.10 0.16 48.7

Hindwing 10 1.57 0.38 61.4

mances of the solpugids at each combination of predator and prey size. Kendall’s

measure of concordance was used to assess between-predator and between-prey

size. similarity in consumption frequencies. For all solpugids, I determined the

mean consumption frequency for each prey body part in order to estimate any

possible preferences as described by Lucas (1983) and Kaspari (1990). Tukey’s

multiple comparison test clustered prey parts according to similar consumption

frequencies. All tests were two-tailed with significance levels set at P = 0.05.

RESULTS AND DISCUSSION

Values for mean chitin weights and percentages for various body parts of T.

pallidipennis are listed in Table 1. Head capsules, antennae, forelegs, midlegs and

both pairs of wings are all characterized by relatively high chitin content (35.4—

61.4%) as compared to the abdomen (9.7—12.3%), thorax (21.1-25.2%) and hind

femur (17.4—18.3%). Analyses of feeding trials indicate that E. magnus selectively

removes the head capsule and wings (Table 2) and focuses its feeding on those

body parts containing the least amount of indigestible chitin such as the abdomen,

thorax and hind femur.

Table 2. Removal time (sec) of Eremorhax magnus for grasshopper body parts from two different

size classes.

Grasshopper size class

Small Large

Body part ~~ na Mean(SD) n Mean (SD)

Head and antennae? 20 37.3 (7.4) 23 72.4 (9.1)

Abdomen 17 NR? 15 NR

Thorax 19 NR 20 NR

Hind femur 18 NR 14 NR

Forewing 12 14.8 (3.6)

Hindwing 14 17.1 (6.1)

a Significant between-grasshopper size differences (P < 0.01).

> NR = body part not removed (grinded vigorously between chelicerae).

16 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Solpugids consumed prey parts from each prey size class in similar frequencies

(Kendall’s W = 0.57, P < 0.05 for small prey; W = 0.84, P < 0.01 for large prey)

except for fore- and hindwings which were very small in the smaller hoppers and

usually ingested with the rest of the thorax. The Tukey tests indicated the following

clusters of consumption frequencies: for the larger prey size category, the head

capsules were removed in 84% of the feeding trials whereas forewings and hind-

wings were removed in lower frequencies (54 and 37%, respectively). For the small

grasshoppers, head capsules were removed in 77% of the feeding trials. Kendall’s

concordance was significant (W = 0.852, P < 0.01) for the mean consumption

frequencies of body parts for each prey size class indicating that the same criteria

were involved in decisions to remove prey parts from both small and large prey.

Video recordings also showed a rather stereotyped feeding behavior pattern for

these solpugids feeding on grasshoppers. In all cases, E. magnus females responded

quickly to tactile stimuli upon contact of prey with their palpi or legs. Following

initial contact, the grasshopper is pulled toward the chelicerae by the palpi. The

prey is then grasped firmly with the chelicerae. This is followed by a vertical

motion of the movable cheliceral finger against the upper fondal teeth resulting

in the fragmentation and grinding of prey tissues. During the movement of the

prey through the cheliceral mill, certain body parts are severed and removed, and

others are retained for further processing and subsequent ingestion (Table 2).

Although the forelegs, midlegs and hind tibiae were discarded, the hind femur

was processed through the chelicerae allowing these solpugids to ingest the mass

of muscle tissue associated with these saltatorial legs. This was not observed when

E.. magnus fed on other types of arthropods such as beetles and spiders. Previous

observations on feeding behavior in other species have indicated that in some

cases the prey is actually stabbed with the chelicerae upon initial contact (Bolwig

1952, Cloudsley-Thompson 1977, Turner 1916). This was not observed in E.

magnus for any feeding trial. Some investigators have reported stalking of prey

by some solpugids such as Hemerotrecha californica Chamberlin and Galeodes

sp. (Muma 1966) but this behavior was not observed for FE. magnus. The amount

of time required by E. magnus to ingest small grasshoppers ranged from 6.2—17.4

min; for larger grasshoppers ingestion time ranged from 11.6—28.3 min.

The results from this study are the first demonstration that prey preparation is

an important component of handling time for a solpugid. By removing body parts

difficult to digest, E. magnus is maximizing the concentration of nutrients that

can be digested and absorbed as well as the amount of space available in the gut

to receive additional food. These benefits may outweigh the cost associated with

an increase in the overall handling time that accompanies prey preparation.

ACKNOWLEDGMENT

I thank T. Punzo and J. Bottrell for assistance in the collection of specimens

in the field, B. Garman for consultation on statistical procedures, D. Donohue

for permission to collect on private property, and W. Price and T. Snell for critical

input and constructive criticism provided on earlier drafts of the manuscript. A

Faculty Development Grant from the University of Tampa made much of this

work possible.

1995 PUNZO: EREMORHAX MAGNUS PREY PREPARATION 17

LITERATURE CITED

Bolwig, N. 1952. Observations on the behavior and mode of orientation of hunting Solifugae. J.

Ent. Soc. S. Aff., 15: 239-240.

Charnov, E. L. 1976. Optimal foraging: the marginal value theorum. Theor. Pop. Biol., 9: 129-136.

Cloudsley-Thompson, J. L. 1977. Adaptational biology of Solifugae (Solpugida). Bull. Br. Arachnol.

Soc., 4: 61-71.

Curio, E. 1976. The ethology of predation. Springer Verlag, New York.

Foster, M.S. 1987. Feeding methods and efficiencies of selected frugivorous birds. Condor, 89: 566-

580.

Haynes, D. L. & P. Sisojevic. 1966. Predator behavior of Philodromus rufus Walckenaer (Araneae:

Thomisidae). Can. Entomol., 98: 113-136.

Hespenheide, H. A. 1973. Ecological inferences from morphological data. Annu. Rev. Ecol. System.,

4: 213-229.

Kaspari, M. 1990. Prey preparation and the determinants of handling time. Anim. Behav., 40: 118-

126.

Kaspari, M. 1991. Preparation as a strategy to maximize nutrient concentration in prey. Behav.

Ecol., 2: 234-241.

Krebs, J. R. & R. H. McCleery. 1984. Optimization in behavioral ecology. pp. 91-121. In Krebs,

J. R. & N. B. Davies (eds.). Behavioral ecology: an evolutionary approach. Sinauer & Assoc.,

Sunderland, Massachusetts.

Lucas, J. R. 1983. The role of foraging time constraints and variable prey encounter in optimal diet

choice. Am. Nat., 122: 191-210.

Muma, M. H. 1951. The arachnid order Solpugia in the United States. Bull. Am. Mus. Hat. Hist.,

97: 34-141.

Muma, M. H. 1966. Feeding behavior of North American Solpugida (Arachnida). Fla. Ent., 49:

199-216.

Nentwig, W. 1987. The prey of spiders. pp. 249-263. In Nentwig, W. (ed.). Ecophysiology of spiders.

Springer Verlag, New York.

O’Brien, W. J., H. Browman & B. I. Evans. 1990. Search strategies and foraging animals. Am. Sci.,

78: 152-160.

Punzo, F. 1989. Effects of hunger on prey capture and ingestion in Dugesiella echina Chamberlin

(Orthognatha, Theraphosidae). Bull. Br. Arachnol. Soc., 8: 72-79.

Punzo, F. 1991. Field and laboratory observations on prey items taken by the wolf spider, Lycosa

lenta Hentz (Araneae, Lycosidae). Bull. Br. Arachnol. Soc., 8: 261-264.

Punzo, F. 1992. Dietary overlap and activity patterns of sympatric populations of Scaphiopus

holbrooki (Pelobatidae) and Bufo terrestris (Bufonidae). Fla. Scientist, 55: 38-49.

Punzo, F. (in press). Diet and feeding behavior of the solpugid, Erembates palpisetulosus Fichter

(Solpugida: Eremobatidae). Psyche.

Punzo, F. & B. Garman. 1989. Effects of encounter experience on the hunting behavior of the spider

wasp, Pepsis formosa (Say) (Hymenoptera: Pompilidae). Southwest. Nat., 34: 573-578.

Scott, M. L., M. C. Nesheim & R. J. Young. 1976. Nutrition of the chicken. M. L. Scott Assoc.,

New York.

Sherry, T. W. & L. W. McDade. 1982. Prey selection and handling in two neotropical hover-gleaning

birds. Ecology, 63: 1016-1028.

Sih, A. 1980. Optimal foraging: partial consumption of prey. Am. Nat., 116: 281-290.

Sokal, R. R. & F. J. Rohlf. 1981. Biometry (2nd ed.). W. H. Freeman and Co., New York.

Turner, C. H. 1916. Notes on the feeding behavior and oviposition of a captive American false-

spider (Eremobates fornicaria Koch). J. Anim. Behav., 6: 160-168.

Zach, R. & J. B. Falls. 1978. Prey selection by captive ovenbirds (Aves: Parulidae). J. Anim. Ecol.,

47: 929-943.

PAN-PACIFIC ENTOMOLOGIST

71(1): 18-23, (1995)

NEST AND COLONY STRUCTURE IN THE PRIMITIVE ANT,

HARPEGNATHOS VENATOR (SMITH)

(HYMENOPTERA: FORMICIDAE)

MICHAEL W. J. CROSLAND

Department of Biology, Chinese University of Hong Kong,

Shatin, N.T., Hong Kong

Abstract.—Colonies of the primitive ant Harpegnathos venator were found to make a simple

nest of uniform design with only two nest chambers. Nests were found in clay embankments.

Such simple nests are unusual in mature colonies of ants. Interesting characteristic funnels were

found constructed both between the chambers and at the nest entrance. Colonies comprised a

mean of only 35 workers and a maximum of 72 workers (from 26 colonies excavated). Dealate

queens were present in nests. Twenty-one nests contained one queen and 5 nests contained 2

queens. This contrasts with some other primitive ant species where the queen caste has been

secondarily lost. Queens, though morphologically distinct, were little larger than workers. Queen

weight averaged only 1.4 times the mean worker weight.

Key Words.—Insecta, Hymenoptera, Formicidae, ant, primitive ant, Harpegnathos, nest

Harpegnathos is amongst the most bizarre and little-known genera of ants.

Recent investigations of the southern Indian species, Harpegnathos saltator Jer-

don, were prompted by the 1990 International Social Insect conference (IUSSI)

held in India, where this ant was made the symbol of the conference because of

its extraordinary appearance. Harpegnathos is frequently compared (Shivashankar

et al. 1989, Nascimento et al. 1993) with the primitive Myrmecia ants from

Australia which also have large size, elongated mandibles, large eyes and primitive

traits, such as workers foraging individually. Harpegnathos, however, belongs to

a different subfamily (subfamily Ponerinae, tribe Ponerini). Its very elongated

mandibles are upturned at the end in a most bizarre fashion and it readily makes

jumps of more than 10 cm to catch insect prey (Musthak Ali et al. 1992).

In this paper, I investigate a second species of Harpegnathos. Nothing is pub-

lished about H. venator, except its taxonomy (Wu 1941). In Hong Kong, where

it has been previously recorded (Wu 1941), it is known as “The Jumping Ant”

and described as “uncommon” (Hill et al. 1982).

Ponerine ants are particularly interesting because of the secondary loss of the

queen caste in a number of species. In this situation, reproductive workers take

over the function of the queen caste (Peeters 1991). Nothing has been published

about colony size, nest structure, or presence or absence of a queen caste in any

species of Harpegnathos. I provide this information in this paper for H. venator

and describe the remarkably simple nest structure of this species.

MATERIALS AND METHODS

Twenty-six colonies of Harpegnathos venator were excavated in 1992 and 1993

from Tai Po Kau Nature Reserve and Sha Lo Tung in the New Territories in

Hong Kong. Detailed notes and measurements of nest structure were made during

nest excavation. All ants present in colonies were collected and brought to the

laboratory where queens and workers were counted and measured.

1995 CROSLAND: HARPEGNATHOS VENATOR NEST STRUCTURE 19

RESULTS

Nest Locations.—All nests were found on steep slopes, typically cut into a

hillside beside a constructed footpath. Slopes with Harpegnathos nests were all

in partially shaded locations in mixed deciduous wooded areas (largely replanted

since 1945). All slopes were of exposed clay and sparsely vegetated with occasional

moss and small ferns.

Workers. —Nests contained from 8 to 72 workers with a mean of 35.0 workers

(interquartile range 25-49). Most workers were found in the upper nest chamber.

Workers were remarkably timid. Many remained stationary at (or retreated to)

the back wall of their nest chamber, with few workers moving forward out of their

nest chamber, even when the nest chamber was broken open. Workers were

monomorphic and had a slow “tempo” (i.e., slow, deliberate movements, HG6ll-

dobler & Wilson 1990).

Queens.— A queen caste was found in H. venator. Of the 26 colonies excavated,

21 colonies contained a single queen and 5 colonies contained 2 queens. Queens

were usually found in the upper nest chamber. Queens were morphologically very

similar to workers though they could be clearly distinguished by their maximum

thorax width (i.e., thorax width at the widest point of the fused meso- and meta-

thorax). (Mean maximum thorax width of queens = 2.25 mm, range 2.20—2.35

mm, ” = 10. For workers, mean = 1.67 mm, range 1.45-1.90 mm, n = 100.)

Queens were found to possess the full complement of flight-associated sclerites,

typical of a generalized hymenopterous thorax (e.g., Wheeler 1910, Bolton 1986,

Hdlldobler & Wilson 1990). Virgin queens, found newly-eclosed in summer, all

possessed full-length wings with hind-wings extending past the abdominal tip

(none were brachypterous). Overall, however, queens were little larger than work-

ers. Queens were an average of only 1.4 times average worker weight. (Mean dry

weight of queen = 16.5 mg, range 15.5-17.0 mg, m = 10. For workers, mean =

11.5 mg, range 8.0-14.5 mg, n = 100.)

Nest Entrance.—Nest entrances extend 1-4 cm out from the slope to form a

characteristic funnel (Fig. 1). The funnel is oval, typically about 3 cm in diameter

(though occasionally larger) but often narrower where it joins the slope. A narrow

nest entrance, open throughout the year in most nests, is present in the middle

of the funnel.

Nest Chambers. — All 26 nests excavated had two nest chambers, although col-

onies collected varied considerably in size (8 to 72 workers). Larger colonies also

had only two nest chambers but enlarged. (The nests were mature since they

seasonally contained multiple alates of both sexes). However, one large colony

(59 workers) had a third rather irregular nest chamber.

The two nest chambers are typically disk-shaped and slightly domed in the

center (Fig. 1). The wall to the nest chamber is very smooth and evenly sculptured

in the clay. Workers often congregate in occasional small extensions to the cham-

bers (“‘alcoves,” Fig. 1b). The upper and lower chambers are separated by a thin

partition and connected by a single funnel.

Funnel Connection.—A remarkable funnel provides the only connection be-

tween the upper and lower nest chambers (Fig. 2). The funnel is typically about

2 cm in diameter with a narrow 5-8 mm hole for the ants to pass through. The

funnel is above the highest part of the domed lower chamber and connects to the

20 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Figure 1. Typical nest structure of Harpegnathos venator. (a) Vertical lateral cross-section of nest.

(b) View from above. Typical measurements are: A = 1.5—2 cm; B = 5-12 cm and; C = 6-16 cm. D

= Nest entrance, E = Lower chamber, F = Funnel connection between lower and upper chamber, G

= Upper chamber, H = “Alcove.”

1995 CROSLAND: HARPEGNATHOS VENATOR NEST STRUCTURE 21

a] B

Figure 2. Funnel between upper and lower nest chambers. (a) Vertical lateral cross-section. (b)

View of circular funnel from below. A = Lower nest chamber; B = Upturned lip of funnel; C = Upper

chamber; D = Diameter of funnel = about 2 cm.

front of the upper chamber. The circular funnel is sculptured from clay and has

an upturned lip (Fig. 2).

DISCUSSION

The present paper describes the two-chambered nest of H. venator which is

amongst the simplest recorded in mature colonies of ants, though the nest also

has a remarkable internal funnel connection. Primitive characteristics of H. vena-

tor included small colony size and little size differentiation between queens and

workers.

Nests of mature ant colonies typically contain a complex of multiple nest cham-

bers, (often dozens or even hundreds, Sudd 1967, Dumpert 1981, Hélldobler &

Wilson 1990). Even primitive Myrmecia ants in Australia have large complex

nests with multiple nest chambers (Gray 1971). The uniformity in the design of

the simple two-chambered nest among H. venator colonies was most striking.

Only small variations occurred among nests, for example, as a result of large rocks

in the clay interfering with the typical disk-shaped nest chambers.

The funnel connection between the two nest chambers is interesting, with no

Le THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

similar structure being recorded in the major reviews on ant nest structure (Whee-

ler 1910, Sudd 1967, Dumpert 1981, HGlldobler & Wilson 1990). The funnel is

typically about 2 cm above the floor of the lower chamber. R. Winney (personal

communication) speculates that H. venator may need their jumping ability to get

from the lower chamber into the funnel. Most ants were found in the upper

chamber and clearly evacuation to the upper chamber would facilitate escape

against arthropod predators entering the nest.

Primitive characteristics found in H. venator included nesting in soil, small

colony size, slow tempo of worker movement and little size differentiation between

queens and workers. Nests of ants are thought to have first evolved in soil and

later developed a diversity of locations, e.g., nesting inside plants, tree-trunks or

using leaves of trees to build nests (Dumpert 1981). The small colony size found

in H. venator (a mean of only 35 workers) further points to their primitiveness

(no large complex society having evolved). This colony size is amongst the lowest

known in ants (H6lldobler & Wilson 1990), though at least three ant species are

known with even smaller colonies (Peeters et al. 1991). However, most ant species

including the primitive Myrmecia ants have evolved the ability to develop much

larger colonies (typically several hundred workers in Myrmecia, H6lldobler &

Wilson 1990).

True morphologically-distinct queens were found in nests of H. venator. This

contrasts with some other primitive ponerine ants where the queen caste has been

replaced by reproductive workers (Peeters 1991). Interestingly in the other major

group of primitive ants, the subfamily Myrmeciinae, the queen caste is always

found in all species so far examined (Haskins 1970; Crosland, unpublished data).

Queens of H. venator are little larger than workers. Similar queens, little larger

than workers, are also found in some primitive Myrmecia species (reviewed in

Wilson 1971). Worker-like queens are sometimes referred to as “‘ergatoid.”” How-

ever, the recent suggestion is to reserve the term “‘ergatoid’”’ for intermediates

between normal queens and workers that can replace the ordinary queen caste

(Hélldobler & Wilson 1990). Hence H. venator queens are not ergatoid. Queens

of Myrmecia and H. venator contrast with those of higher ants such as Camponotus

where queens are approximately 20 times the weight of minor workers (Crosland

1990).

Ant queens typically found colonies claustrally, in a closed underground cham-

ber where the first larvae are fed on food reserves and digested alary muscles from

the body of the lone queen. Worker-like Myrmecia queens are primitive amongst

ants in that the single colony-founding queen periodically leaves her underground

chamber to forage, presumably because her worker-like body has insufficient food

reserves to rear her first brood (Wilson 1971). Experiments are planned to test

whether similar queen foraging also occurs in H. venator.

ACKNOWLEDGMENT

Thanks to Phil Ward for pointing out this interesting species and to Dick

Winney, Christian Peeters and an anonymous referee for help with this work.

This work was supported by a Direct Grant from the University and Polytechnic

Grants Council to the Chinese University of Hong Kong.

1995 CROSLAND: HARPEGNATHOS VENATOR NEST STRUCTURE 23

Note added in proofs: The nest architecture of Harpegnathos saltator will soon be published in: Peeters,

C., B. Hélldobler, M. Moffett & T. M. Mustak Ali. 1994. “‘Wall-papering”’ and elaborate nest

architecture in the ponerine ant Harpegnathos saltator. Insectes Sociaux (in press).

LITERATURE CITED

Bolton, B. 1986. Apterous females and shift of dispersal strategy in the Monomorium salomonis-

group (Hymenoptera: Formicidae). J. Natur. Hist., 20: 267-272.

Crosland, M. W. J. 1990. The influence of the queen, colony size and worker ovarian development

on nestmate recognition in the ant Rhytidoponera confusa. Anim. Behav., 39: 413-425.

Dumpert, K. 1981. The social biology of ants. Pitman, London.

Gray, B. 1971. Notes on the biology of the ant species Myrmecia dispar (Clark). (Hymenoptera:

Formicidae). Insectes Sociaux, 18: 71-80.

Haskins, C. P. 1970. Researches in the biology and social behavior of primitive ants. pp. 355-388.

In Aronson, L. R., E. Tobach, D. S. Lehrman & J. S. Rosenblatt (eds.). Development and

evolution of behavior. W. H. Freeman, San Francisco.

Hill, D. S., P. Hore & I. W. B. Thornton. 1982. Insects of Hong Kong. Hong Kong University Press,

Hong Kong.

Hdlldobler, B. & E. O. Wilson. 1990. The ants. Belknap Press, Cambridge, Massachusetts.

Musthak Ali, T. M., C. Baroni Urbani & J. Billen. 1992. Multiple jumping behaviors in the ant

Harpegnathos saltator. Naturwissen., 79: 374-376.

Nascimento, R. R., J. Billen & E. D. Morgan. 1993. The exocrine secretions of the jumping ant

Harpegnathos saltator. Comp. Biochem. Physiol., 104B: 505-508.

Peeters, C. 1991. The occurrence of sexual reproduction among ant workers. Biol. J. Linn. Soc., 44:

141-152.

Peeters, C., S. Higashi & F. Ito. 1991. Reproduction in ponerine ants without queens: monogyny

and exceptionally small colonies in the Australian Pachycondyla sublaevis. Ethology Ecology

& Evolution, 3: 145-152.

Shivashankar, T., H. C. Sharathchandra & G. K. Veeresh. 1989. Foraging activity and temperature

relations in the ponerine ant Harpegnathos saltator Jerdon (Formicidae). Proc. Indian Acad.

Sci. (Anim. Sci.), 98: 367-372.

Sudd, J. H. 1967. An introduction to the behaviour of ants. E. Arnold Publishers, London.

Wheeler, W.M. 1910. Ants, their structure, development and behavior. Columbia University Press,

New York.

Wilson, E. O. 1971. The insect societies. Belknap Press, Cambridge, Massachusetts.

Wu, C. F. 1941. Catalogus insectorum sinensium. Volume 6. Published by Yenching University,

Peiping, China.

PAN-PACIFIC ENTOMOLOGIST

71(1): 24-30, (1995)

TETHERED FLIGHT CHARACTERISTICS OF MALE AND

FEMALE PEAR PSYLLA (HOMOPTERA: PSYLLIDAE):

COMPARISON OF PRE-REPRODUCTIVE AND

REPRODUCTIVE INSECTS

Davip R. HORTON AND TAMERA M. LEwIs

USDA-ARS,! Yakima, Washington 98902

Abstract.—In tethered flight experiments, female winterform pear psylla (Cacopsylla pyricola

Foerster) exhibited longer mean flight durations and a higher frequency of long duration flights

(> 15 min) than did males. Pre-reproductive psylla (collected from the field in mid-winter)

exhibited longer duration flights than did psylla collected from the same orchard after egg-laying

had commenced. Flight durations decreased with consecutive flights. Although tibia length and

wing size were larger for females than males, there was no evidence within sexes or collection

dates that body size and flight duration were correlated.

Key Words. —Insecta, tethered flight, pear psylla, body size

Pear psylla, Cacopsylla pyricola Foerster, is a specialist pest of pears in many

pear growing regions of the world. The species is seasonally dimorphic (Oldfield

1970). The overwintering generation (winterform) undergoes a reproductive dia-

pause in the fall, characterized by a lack of mating and immature ovaries (Krysan

& Higbee 1990), and (for a large fraction of the population) dispersal from pear

orchards (Horton et al. 1992). Reentry into pear orchards, ovarian development,

and egg-laying occur in early spring as temperatures rise (Krysan & Higbee 1990,

Horton et al. 1992).

For many insect species, diapause, reproductive status, and dispersal are closely

linked life history components (Johnson 1969). Earlier work failed to demonstrate

large differences between tethered flight activity of diapausing and post-diapause

(but pre-reproductive) female winterforms (Horton, unpublished data). This study

compared flight characteristics of psylla collected from the field in mid-winter

(pre-reproductive) with those of psylla collected from the same orchard once egg-

laying had commenced. We also compared flight characteristics of males and

females for both pre-reproductive and reproductive psylla. In several other insect

species, females appear to be the more dispersive sex (Adesiyun & Southwood

1979, Reader & Southwood 1984, Davis 1986). Finally, we recorded several wing-

and body-size measurements to determine whether flight capacity was related to

body size.

MATERIALS AND METHODS

Flight Methods.—Winterform pear psylla were collected from a commercial

pear orchard in Yakima, Washington on two occasions: 19-22 Jan 1993 and 29-

31 Mar 1993. Egg-laying had commenced in the field by the March collection;

dissections indicated that psylla collected in mid- to late-January had immature

ovaries (first mature eggs in dissections were not noted until the first week of

March: DRH, unpublished data). Psylla were separated by sex and placed in 150

13706 W. Nob Hill Blvd.

1995 HORTON & LEWIS: PEAR PSYLLA FLIGHT 25

ml screened vials (5 per vial) on field-collected pear shoots. Vials and psylla were

placed in the flight room (25° C; 85% RH; 10:14 h [L:D]) for 24 h; psylla were

flown in the morning (0900 PST) and afternoon (1300 PST) of the day following

collection.

The tip of a nylon brush bristle (5-8 mm in length) was attached to the me-

sothorax of psylla using a small drop of a cyanoacrylate glue (INSTA-CURE®,

Bob Smith Industries, Atascadero, CA). The other end of the bristle was attached

to an insect pin, which in turn was stuck in a styrofoam block. Psylla were

immobilized for gluing by lightly etherizing them. After being glued, psylla were

placed in a cold room (3° C) for 30 min to recover from handling.

The flight chamber was a 0.5 m X 0.5 m X 0.5 m box with white sides, a

plexiglass top, and an open front. The chamber was illuminated by two 20 watt

cool white fluorescent bulbs that were suspended | m over the box. Flight durations

were timed with the aid of a personal computer and BASIC program. Upon

removal from the cold room, psylla were allowed a 15 min “warm-up” period

before initiation of the experiment. If flight was not initiated voluntarily, we lightly

blew a puff of air at the posterior end of the psylla. Each psylla was flown 5 times,

with a 5 min rest period between consecutive flights. Exceptions were psylla that

flew continuously for at least an hour; for these psylla, the experiment was ended

at 1 h. Sample sizes were 32 to 39 psylla for each sex x collection date combination.

Body Size Measurements. — Following the flight experiments, body size mea-

surements were made using a dissecting scope and ocular micrometer. Measure-

ments were made of tibia length and four wing characteristics. For tibia length,

a hind leg was cut off at mid-femur and laid flat on the platform, with the outward

portion of the leg facing up. The tibia was measured from the posterior distal

point to the knob at the anterior proximal end. For wing size, a forewing was cut

off near the base, laid convex side up, and then covered with a glass cover slip.

Wing measurements included the distances between the following: (1) fork of the

median (M) and cubitus (Cu) veins to the fork of Cu,, and Cu,, veins (measure

LRC in Nguyen 1985); (2) fork of the Cu,, and Cu,, veins to the fork of the M,.,

and M;,, veins (measure LCM in Nguyen 1985); (3) fork of the M and Cu veins

to the apex of the wing; (4) termination of the Cu,, vein to the costal vein break.

The latter two estimates were included to provide some indication of overall wing

length and width, respectively.

Data Analysis. —For each insect, three measures of flight duration were mon-

itored: duration of the longest flight, mean flight duration (of consecutive flights

[up to 5 flights per insect]), and duration of the first flight. Flights of 60+ min

were scored as 60 min. These data were analyzed with two-way (sex x reproductive

status) analysis of variance (ANOVA). We caution here that the data departed

significantly from the assumption of normality (even if transformed), results not

uncommon in studies of tethered flight (Davis 1980). Because ANOVA assump-

tions were not met, marginally significant results should be accepted only cau-

tiously. For insects that completed 5 flights, we included flight number (first

through fifth) as a factor in a repeated measures ANOVA; this analysis allowed

us to determine whether flight duration changed between one flight and a sub-

sequent flight. Body size measurements were compared between sexes and col-

lection dates using multivariate analysis of variance (MANOVA). All analyses

were done in the PROC GLM package of SAS (SAS Institute 1987).

6 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

0.754 4 PRE-REPR. MALES

A REPROD. MALES

0.30

0.25

@ PRE—REPR. FEMALES

O REPROD. FEMALES

RELATIVE FREQUENCY

0.00

SHORT INTERMED. LONG

FLIGHT DURATION

Figure 1. Observed (symbols) and fitted (lines) proportion of pear psylla for which the longest

flight was short-duration (< 5 min), intermediate-duration (5—15 min), or long-duration (> 15 min).

Lines show predictions from multiway categorical analysis: sex—x? = 6.81, P = 0.033; reproductive

status— x? = 5.46, P = 0.065; likelihood ratio (= interaction): P = 0.996. For first flight (frequencies

very similar to those shown for longest flight): sex— x? = 7.1, P = 0.029; reproductive status— x? =

6.4, P = 0.04; likelihood ratio (= interaction): P = 0.994. For average flight (data not shown), P-values

both > 0.25.

We also classified flight durations (longest, first, mean) into one of three cate-

gories: short flight (< 5 min); intermediate flight (5-15 min); long flight (> 15

min). These data were analyzed as a 2 X 2 X 3 (sex X reproductive status x

flight category) multiway categorical table using PROC CATMOD (SAS Institute

1987).

Product-moment correlation was used to determine whether there was any

relationship between tibia length or the four wing measurements and the three

estimates of flight duration (longest, mean, first). Analyses were done separately

for each sex X collection date combination using the PROC CORR package of

SAS (SAS Institute 1987).

RESULTS

For longest flight and first flight, proportion of psylla falling in the short-,

intermediate-, or long-duration flight categories varied with both sex and repro-

ductive status (Fig. 1; results shown only for longest flight, as results for first flight

were virtually identical). For both sexes, probability of intermediate- or long-

duration flight was larger for pre-reproductive psylla than for reproductive psylla,

whereas the converse was true for short-duration flights; i1.e., for the latter, about

40% of reproductive psylla exhibited short-duration flight, whereas only about

25% of pre-reproductive psylla exhibited flights of this duration (Fig. 1). The

1995 HORTON & LEWIS: PEAR PSYLLA FLIGHT zt

MINUTES OF FLIGHT

PRoek.= GREP IPRE=h. KE PRE-K REP

LONGEST FIRST AVERAGE

Figure 2. Mean (+ standard error) flight durations (longest, first, average) for pre-reproductive

(PRE-R) and reproductive (REP) pear psylla. Results of ANOVA (all effects df = 1,142); longest: sex—

F = 4.8, P = 0.029; reproductive status— F = 8.1, P = 0.005; interaction—P = 0.27; first flight: sex—

F = 4.6, P = 0.034; reproductive status—F = 8.7, P = 0.004; interaction—P = 0.23; average: sex—

F = 5.8, P = 0.017; reproductive status—F = 4.0, P = 0.049; interaction—P = 0.31.

greatest difference between the sexes was in frequency of intermediate- and long-

duration flights. Females exhibited a higher frequency of long-duration flights

than did males, whereas males flew primarily intermediate-duration flights. Re-

sults are not shown for average flight duration, as reproductive status and sex

effects were not significant (see Fig. 1 caption).

Mean flight durations varied with sex and reproductive status for all three

estimates of flight duration (Fig. 2). For all measures, females exhibited larger

mean durations than did males, and pre-reproductive psylla flew longer than did

reproductive psylla (although P = 0.049 for reproductive status [average flight

duration]; see Fig. 2 caption). There is a suggestion of a sex X reproductive status

interaction for all three response variables (Fig. 2); however, in no case was the

interaction significant (Fig. 2 caption).

For psylla that completed 5 flights, there was a significant decay in mean flight

duration between the first flight and subsequent flights (Fig. 3). There was also a

significant flight number x reproductive status interaction (Fig. 3 caption); the

interaction was apparently due to large differences in flight durations between

reproductive and pre-reproductive psylla for the first flight, but not for subsequent

flights (univariate ANOVA for the first flight: reproductive status—F = 7.8; df=

1,133; P = 0.006; similar analyses for flights 2-5 were non-significant).

Males were significantly smaller than females (Table 1). However, within each

sex X reproductive status category, correlations between the body size measure-

ments and the three measures of flight duration (longest, mean, first) were non-

significant.

28 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

PRE—REPR. FEMALES

REPROD. FEMALES

PRE—REPR. MALES

3 REPROD. MALES

MINUTES OF FLIGHT

FLIGHT NUMBER

Figure 3. Mean (+ standard error) flight durations of consecutive flights for pear psylla that

completed 5 flights. Results of repeated measures ANOVA (within-subjects statistics from Wilks’

lambda): flight number—F = 28.3, df = 4,130, P < 0.0001; flight number x reproductive status—F

= 2.5, df = 4,130, P = 0.04; flight number x sex—F = 1.5, df = 4,130, P = 0.21; flight number =x

reproductive status x sex—F = 2.0, df = 4,130, P = 0.10; reproductive status—F = 5.7, df = 1,133,

P= 0.018; sex—F < 1.0; reproductive status x sex—F' < 1.0.

DISCUSSION

Studies have suggested or shown for other insect species that females are more

dispersive than males (Adesiyun & Southwood 1979, Reader & Southwood 1984,

Stewart & Gaylor 1991) or that females exhibit longer flight durations than males

in laboratory flight experiments (Dingle 1966, Davis 1986). In this study, female

winterform pear psylla exhibited larger mean flight durations (Fig. 2) and a higher

frequency of long (> 15 min) flights than did males (Fig. 1). Also, of the 146

insects tested, three flew continuously for 1+ h before being interrupted; all three

were females (longest male flight was 46 min). As females are larger than males

(Table 1), it is tempting to hypothesize that body size affected flight durations, as

shown for other species (Dingle et al. 1980, Davis 1986, Saks et al. 1988, Roff

1991); however, we were unable to demonstrate a relationship between body size

and flight duration within any of the 4 sample groups.

Fall-collected and spring-collected pear psylla did not differ substantially in

flight duration, despite differences between the collections in ovarian development

of psylla (DRH, unpublished data). Long-duration flights in many insects occur

before significant ovarian development (Johnson 1969), so the results of the earlier

study were unanticipated; but, see the discussion in Sappington & Showers (1992).

In this study, there were significant differences in flight activities between the two

collection dates, as psylla collected in late-March had reduced flight durations

relative to psylla collected in January. One difference between this study and a

previous study (DRH, unpublished data) is that the spring sample taken here

occurred well after egg-laying had commenced in the field, whereas the spring

sample in the other study occurred before egg-laying (March 1). This result suggests

1995 HORTON & LEWIS: PEAR PSYLLA FLIGHT 29

Table 1. Mean (+ standard error) body size measures (mm) of male and female winterform pear

psylla collected in January (pre-reproductive) and late March (reproductive), 1993.

Pre-reproductive Reproductive

Measure*>

Females Males Females Males

Tibia 0.56 (0.004) 0.54 (0.004) 0.55 (0.004) 0.54 (0.004)

LRC 0.65 (0.006) 0.56 (0.006) 0.63 (0.006) 0.57 (0.006)

LCM 0.55 (0.006) 0.51 (0.006) 0.55 (0.006) 0.50 (0.006)

Length 1.87 (0.014) 1.66 (0.012) 1.86 (0.012) 1.66 (0.012)

Width 1.07 (0.008) 0.97 (0.008) 1.06 (0.008) 0.97 (0.008)

4 Final four measures are wing measures (see Materials and Methods).

+’ MANOVA (F-statistics from Wilks’ lambda): sex— F = 60.1, df = 5,130, P < 0.0001; reproductive

status—F = 0.5, df = 5,130, P = 0.79; sex x reproductive status—F = 2.3, df = 5,130, P = 0.048.

Univariate tests: all five measurements differed between males and females (P < 0.0001, df = 1,134);

for LRC, the sex x reproductive status interaction is significant (P = 0.012, df = 1,134); reproductive

status main effects were non-significant for all five measures (P > 0.25).

that long duration flights decreased in frequency as psylla became fully engaged

in egg-laying.

ACKNOWLEDGMENT

The comments of Rick Redak (University of California, Department of En-

tomology, Riverside), Dave Thompson (New Mexico State University, Las Cru-

ces), Tom Unruh (USDA-ARS, Yakima, Washington), and Tom Weissling (USDA-

ARS, Yakima, Washington) on an earlier draft are appreciated. We thank Mark

Weiss (USDA-ARS, Yakima, Washington) for writing the BASIC program used

in recording flight data. The Washington Tree Fruit Research Commission (Yak-

ima), Winter Pear Bureau (Portland, Oregon), and the Western Regional IPM

Grants Program (WRCC 69) provided financial support.

LITERATURE CITED

Adesiyun, A. A. & T. R. E. Southwood. 1979. Differential migration of the sexes in Oscinella frit

(Diptera: Chloropidae). Entomol. Exp. Appl., 25: 59-63.

Davis, M. A. 1980. Why are most insects short fliers? Evol. Theory, 5: 103-111.

Davis, M. A. 1986. Geographic patterns in the flight ability of a monophagous beetle. Oecologia,

69: 407-412.

Dingle, H. 1966. Some factors affecting flight activity in individual milkweed bugs (Oncopeltus). J.

Exp. Biol., 44: 335-343.

Dingle, H., N. R. Blakely & E. R. Miller. 1980. Variation in body size and flight performance in

milkweed bugs (Oncopeltus). Evolution, 34: 371-385.

Horton, D. R., B. S. Higbee, T. R. Unruh & P. H. Westigard. 1992. Spatial characteristics and effects

of fall density and weather on overwintering loss of pear psylla (Homoptera: Psyllidae). Environ.

Entomol., 21: 1319-1332.

Johnson, C. G. 1969. Migration and dispersal of insects by flight. Methuen, London.

Krysan, J. L. & B.S. Higbee. 1990. Seasonality of mating and ovarian development in overwintering

Cacopsylla pyricola (Homoptera: Psyllidae). Environ. Entomol., 19: 544-550.

Nguyen, T. X. 1985. Establishment of a morphometric range of Psyllidae (Insecta-Homoptera);

seasonal polymorphism of Psylla pyri L. C.R. Acad. Sc. Paris, 301: 369-372.

Oldfield, G. N. 1970. Diapause and polymorphism in California populations of Psylla pyricola

(Homoptera: Psyllidae). Ann. Entomol. Soc. Amer., 63: 180-184.

. Reader, P. M. & T. R. E. Southwood. 1984. Studies on the flight activity of the Viburnum whitefly,

a reluctant flyer. Entomol. Exp. Appl., 36: 185-191.

30 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Roff, D. A. 1991. Life history consequences of bioenergetic and biomechanical constraints on mi-

gration. Amer. Zool., 31: 205-215.

Saks, M. E., M. A. Rankin & R. E. Stinner. 1988. Sexually differentiated flight responses of the

Mexican bean beetle to larval and adult nutrition. Oecologia, 75: 296-302.

Sappington, T. W. & W. B. Showers. 1992. Reproductive maturity, mating status, and long-duration

flight behavior of Agrotis ipsilon (Lepidoptera: Noctuidae) and the conceptual misuse of the

oogenesis-flight syndrome by entomologists. Environ. Entomol., 21: 677-688.

SAS Institute. 1987. SAS/STAT guide for personal computers. Version 6 Edition. Cary, North

Carolina.

Stewart, S. D. & M. J. Gaylor. 1991. Age, sex, and reproductive status of the tarnished plant bug

(Heteroptera: Miridae) colonizing mustard. Environ. Entomol., 20: 1387-1392.

PAN-PACIFIC ENTOMOLOGIST

71(1): 31-60, (1995)

NON-MONOPHYLY OF AUCHENORRHYNCHA

(“HOMOPTERA”), BASED UPON 18S rDNA PHYLOGENY:

ECO-EVOLUTIONARY AND CLADISTIC IMPLICATIONS

WITHIN PRE-HETEROPTERODEA HEMIPTERA (S.L.) AND

A PROPOSAL FOR NEW MONOPHYLETIC SUBORDERS

JOHN T. SORENSEN,! BRUCE C. CAMPBELL,? RAYMOND J. GILL,!

AND Jopy D. STEFFEN-CAMPBELL2

‘Insect Biosystematics, Plant Pest Diagnostics Center,?

California Dept. of Food & Agriculture, Sacramento, California 95832-1448.

“Western Regional Research Center, USDA-ARS,

Albany, California 94710-1100.

Abstract.—Parsimony-based phylogenetic-analyses of full 18S rRNA genes (18S rDNA) were

conducted to determine the basal clade topology of Hemiptera. The single most parsimonious

topology, which attenuated homoplasy and retained only the most conservative base sites,

showed: (a) Sternorrhyncha and Euhemiptera are sister-clades; (b) Cicadomorpha (composed of

Cicadidae and sister-clade Cercopidae+ Membracidae) is sister-clade to the remaining Euhem-

iptera; and (c) Fulgoromorpha is sister-clade to Heteropterodea (Coleorhyncha+ Heteroptera).

Supportive morphological synapomorphies for the 18S rDNA topology are listed. Less parsi-

monious, but competitive topologies indicate association of Heteroptera with extant Cicado-

morpha. Thus, Auchenorrhyncha is unlikely (< 10%) to be monophyletic, as previously assumed,

and its morphological synapomorphies (tymbal acoustic systems, aristoid antennae, ScP+R vein

fusion) are homoplasious; the misinterpretation, selection, and convergence of these traits is

discussed. Current paleontological assessments of the basal Hemiptera are reviewed and also

suggest non-monophyly for Auchenorrhyncha. A Lower Cretaceous fossil, Megaleurodes me-

gocellata Hamilton, previously assigned to Aleyrodoidea: Boreoscytidae, is tentatively reassigned

to fossil superfamily Fulgoridioidea of Fulgoromorpha. Use of paraphyletic Auchenorrhyncha

should be abandoned as a hemipteran suborder; instead recognition of the four monophyletic

basal clades of Hemiptera as its suborders is appropriate. Three new suborder names are proposed

because of potential confusions or varying definitions (discussed) involving existing names:

Clypeorrhyncha (= extant, monophyletic Cicadomorpha), Archaeorrhyncha (= Fulgoromorpha),

and Prosorrhyncha (= Heteropterodea, as clade Coleorhyncha+ Heteroptera); Sternorrhyncha

is retained. Clade name Neohemiptera is proposed for the clade Fulgoromor-

pha-+ Heteropterodea. An eco-evolutionary scenario for cladogenesis among the basal hemipteran

clades is presented. Evidence indicates a saltational, punctuated equilibrium mode of evolution

occurred among the clades during, or near, the Permian.

Key Words. —Insecta, Cicadomorpha, Fulgoromorpha, molecular phylogeny, cladistics

The hierarchical, ordinal relationship among the names “Hemiptera,” ‘“‘Het-

eroptera,” and “Homoptera” has been confused since Latreille (1810) first rec-

ognized the latter two names as sections of his ““Hemiptera’”’ (sensu lato). This

was done in response to Fabricius’ (1775) mouthpart-based modification of Lin-

naeus’ (1758) original wing-based classification of insect orders; see Henry &

Froeschner (1988: xii—xiii) for discussion. Current schemes for recognition of (an)

order(s) for all hemipterans differ confusingly among workers and regions. Some

prefer the separate orders Homoptera and Hemiptera (sensu strictu, sensu La-

3 3294 Meadowview Road.

32 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

treille’s Heteroptera), following the revisions of Borror & DeLong’s (e.g., 1971)

books. Others use order Hemiptera (s.1., sensu Latreille) with suborders Homop-

tera and Heteroptera. Still others use the separate orders Homoptera and Het-

eroptera.

Although sometimes presenting operational problems for a classification (Sor-

ensen 1990: 402), application of monophyly through cladistic philosophy is being

used to solve the dilemma of hierarchical grouping in Heteroptera (Schuh 1986),

and hopefully for all hemipterans here. On the basis of morphology, cladistic

workers (Kristensen 1975, 1991; Hennig 1981; Popov 1981; Schuh 1979; Wootton

& Betts 1986) now recognize that Homoptera is a paraphyletic grade at the base

of monophyletic Heteroptera, with the entire greater clade usually recognized as

Hemiptera. Accordingly, order Homoptera must be abandoned under a mono-

phyly criterion, despite the resistance of many homopterists to having their groups

incorporated into Hemiptera because they associate that name with a usage now

replaced by Heteroptera (e.g., Henry & Froeschner 1988).

Recent treatments (e.g., Carver et al. 1991: 443) of Hemiptera retain Sternor-

rhyncha and Auchenorrhyncha as hemipteran suborders, based on their respective

assumed monophyly*. Sternorrhyncha is now considered a sister-group (Schuh

1979, Carver et al. 1991, Wheeler et al. 1993) to Auchenorrhyncha+ Heteroptera?

(= Euhemiptera sensu Schuh 1979). Now, Campbell et al. (1994) show irrefutable

evidence of the monophyly of Sternorrhyncha, as a synapomorphy having a unique

nucleotide expansion area of 18S rDNA. Thus, Sternorrhyncha is a cladistically

valid hemipteran suborder. In Campbell et al.’s (1994) analysis, however, Au-

chenorrhyncha was paraphyletic, a result that is cladistically incompatible with

its use as a hemipteran suborder.

Wheeler et al. (1993) used discontinuous, short sections of 18S rDNA and

morphological data, alone and in combination, in a parsimony analysis to show

in their most resolute indications that their ““Auchenorrhyncha’”’ were a mono-

phyletic grouping. However, because that analysis was chiefly concerned with

relationships within Heteroptera, it included only minimal representatives of

Cicadomorpha‘* (sensu Carver et al. 1991, and here), and showed their monophyly

to be based upon two 18S rDNA sites that were homoplasious when considered

over their entire generated tree. Unfortunately, they excluded Fulgoromorpha,

the putative sister-group to Cicadomorpha (Carver et al. 1991: 445). As a con-

sequence, Wheeler et al.’s (1993) analysis established only: (a) tentative mono-

phyly, based upon nucleotide homoplasy under parasimony, for their treated

cicadomorphans rather than among all auchenorrhynchous groups; and (b) that

(in the absence of Fulgoromorpha) their cicadomorphan taxa (+ Cercopidae,

Membracidae) formed a sister-group to Heteropterodea, the latter as clade Co-

leorhyncha+ Heteroptera (sensu Schlee 1969, Schuh 1979).

4 Hamilton (1981) considered Auchenorrhyncha to be polyphyletic, based on head morphology,

with its groups surrounding his “Aphidomorpha” (= Sternorrhyncha); however, he considered the

Homoptera, itself, to be monophyletic and the sister-group of Heteropterodea (sensu Schuh 1979).

> Carver et al. (1991) use Heteroptera [sensu lato] to include Coleorhyncha+ Heteroptera [sensu

stricto: e.g., Henry & Froeschner (1988)], a clade considered Heteropterodea (Schuh 1979) [= Het-

eropteroidea (Schlee 1969)] here.

6 Wheeler et al. (1993) used a single Cicadidae [Tibicen sp.] and two Cicadellidae [Graphocephala

coccinea (Forster), Oncometopia orbona (Fabr.)] species.

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA 33

In contrast, Campbell et al.’s (1994) analysis included exemplar taxa from

Fulgoromorpha (Flatidae), Cicadomorpha (Cercopidae, Cicadidae, Membraci-

dae), Sternorrhyncha (Psyllidae, Aphididae, Diaspididae, Aleyrodidae) and Het-

eroptera (Miridae). They mention, but do not discuss, the paraphyly of Auchen-

orrhyncha because that study’s purpose was only to establish monophyly for

Sternorrhyncha and its internal phylogeny with reference to the derivation of

Aleyrodidae.

This article: (a) analyzes an expanded set of the 18S rDNA nucleotide sequences

used by Campbell et al. (1994) and derives the basal phylogenetic topology among

major clades of hemipterans; (b) discusses the morphological characters previously

assumed to be valid synapomorphies for Auchenorrhyncha, but that must rep-

resent homoplasies according to our 18S rDNA analyses; (c) lists supporting

morphological synapomorphies for the 18S rDNA-based tree; (d) discusses the

eco-evolutionary scenario involved with cladogenesis of the 18S rDNA-based

tree; and (e) proposes cladistically compatible category names to reflect the re-

alignment of the basal phylogeny of Hemiptera.

DISCUSSION OF METHODS

Chemically-Based Procedures. —Preparation followed Campbell et al. (1994).

Total genomic DNA was purified by homogenizing fresh insects, or parts thereof,

in micro-centrifuge tubes with a pestle in 200 ul of sterile buffer (10 mM Tris,

2.5 mM MgCl, 50 mM KCl), 200 ul phenol and 10 ul 20% SDS. The phases

were separated using centrifugation, and the DNA was precipitated using ethanol

and resuspended in 20 wl TE (10 mM Tris [pH 8.0], 1 mM EDTA).

PCR (Polymerase Chain Reaction) was performed using the GeneAmp® Kit

(Perkin Elmer Cetus, Norwalk, Connecticut) with 25-yl reactions: 1 ul DNA

template (~ 100 ng), 2.5 wl PCR buffer, 0.5 ul each dNTP, 2 yl (50 nM) each

respective forward and reverse primer, 0.125 wl Tag DNA polymerase and 15.25

ul water. The PCR cycling program was: 30 sec at 95°C, followed by 39 cycles of

1 min at 95°C, 2 min at 50°C and 4 min at 74°C, with 7 min at 74°C after the

last cycle.

Because the 18S rDNA used was difficult to PCR amplify as a single unit, it

was treated as two separate units (“front and “back’’). The front 18S rDNA

portion used (a) forward primer: 5'-CTG GTT GAT CCT GCC AGT AGT-3’;

and (b) reverse primer: 5'-GGT TAG AAC TAG GGC GGT ATC-3’. The back

18S rDNA portion used (c) forward primer: 5’-GAT ACC GCC CTA GTT CTA

ACC-3’; and (d) reverse primer: 5’-TCC TTC CGC AGG TTC ACC-3’. These

primers, a—d respectively, correspond to the base positions (a) 4-24, (b) 1385-

1404, (c) 1385-1404, and (d) 2446-2463, of 18S rDNA for Acyrthosiphon pisum

(Harris), as determined by Kwon et al. (1991).

All PCR products were cloned to contend with potentially contaminating DNA

(from associated fungi, parasitic arthropods, etc.) that might be present in the

hemipteran (“‘template’’) preparations. Cloning used the plasmid and competent

cells supplied in the 7A Cloning® System (Invitrogen, LaJolla, California), and

cloning procedures followed the protocols in the instruction manual. Plasmid

DNA preparations were digested with Eco RI and separated by electrophoresis.

Candidate clones for sequencing were selected based upon appropriate size of the

inserted PCR product. Confirmation of the correct 18S rDNA was determined

34 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Table 1. Base sites used in SET 4 analysis. These most conservative informative sites were retained,

as discussed in the text, after attenuation and polarization of alignment A. Sites numbers are our’s

for alignment A after the initial attenuation (SET 1 analysis). See text for site position and secondary

structure of synonymous rRNA (sensu Kwon et al. 1991). Taxa species are given under methods,

STERN represents the site expression across all treated sternorrhynchan taxa. The full sequences

(>2000 sites) for all taxa are deposited in GenBank, and are available there or from us.

Base site number

1 1 1 1 1 1

1 1 1 1 2 2 4 4 0 1 1 1 2 2

5 6 6 7 7 3 5 5 9 4 6 3 5 2 2 1 1 2 5 6

Taxon 3 2 6 0 9 2 3 9 8 1 3 4 7 1 5 0 7 3 1 9

COLEO TAGtTGAACACGGGGGGG6s CT

STERN T GGtT Gee caA ££ GaGaeoatT a@wGeee¢

MEMBR A A G C A A A T A A AG T T T GGGT éT

CERCO A A GTA A A T A A A GT T GGGGT T

CICAD A A A T A A A T GA GG TTGGGA T T

DELPH A A GCA AA TAC GA TGtTGA A CC T

MIRID A A A T A A AT GA A AGA T GA GC T

a Site 79 is homoplasious, as A, in the dipterans, Aedes and Drosophila (see Carmean et al. 1992).

> Site 454 is homoplasious within several heteropteren lineages in Wheeler et al’s. (1993) data

sequences.

by restriction endonuclease analysis and nucleotide sequencing. Stock cultures of

clones used here are available from BCC and JDS-C at USDA, Albany, California.

Both top and bottom strands of double-stranded DNA were completely se-

quenced using the materials and protocols supplied with the Sequenase® (version

2.0) Sequencing Kit (U.S. Biochemical, Cleveland, Ohio), and [a?°S]dATP (Am-

resham, Arlington Heights, Illinois).

Exemplar Taxa Employed.—Sequences from our material are deposited with

GenBank under acc. nos. U06474 to U06481, except for Prokelisia marginata

(Van Duzee) (acc. no U09207). Identifications were made by RJG; voucher spec-

imens of most of the taxa are maintained at CDFA, Sacramento, California.

Families analyzed and their exemplars are:

ALEYRODIDAE: Pealius kelloggii (Bemis) [CALIFORNIA. SACRAMENTO Co.: Sacramento,

Mar 1993, Prunus lyoni (Eastwood) C. S. Sargent]. APHIDIDAE: Acyrthosiphon pisum (Harris) [18S

rDNA sequence ex Kwon et al. (1991), deposited in GenBank, acc. number X62623]. CERCOPIDAE:

Philaenus spumarius L. [CALIFORNIA. CONTRA COSTA Co.: Pinole, 29 Jun 1993, geranium].

CICADIDAE: Okanagana utahensis Davis [CALIFORNIA. SHASTA Co.: Milford, Jul 1993, Arte-

misia tridentata Nuttall]. DELPHACIDAE: Prokelisia marginata [CALIFORNIA. CONTRA COSTA

Co.: Richmond, 28 Sep 1993, Spartina foliosa Trin.]. DIASPIDIDAE: Aonidiella aurantii (Maskell)

[CALIFORNIA. SACRAMENTO Co.: Sacramento, 14 Jul 1993, Laurus nobilis L.]. MEMBRACI-

DAE: Spissistilus festinus (Say) [CALIFORNIA. YOLO Co.: Davis, 20 Sep 1993, alfalfa]. MIRIDAE:

Lygus hesperus Knight [CALIFORNIA. YOLO Co.: Davis, 20 Sep 1993, alfalfa]. PSYLLIDAE: Trioza

eugeniae Froggatt [CALIFORNIA. ALAMEDA Co.: Albany, 7 Apr 1993, Eugenia sp.]. TENEBRI-

ONIDAE: Tenebrio molitor L. [18S rDNA sequence ex Hendriks et al. (1988), deposited in GenBank,

acc. number X07801].

Phylogenetic Analyses. — Initial alignments of nucleotide sequences were achieved

using Gene Works® (version 2.3.1, subprogram: “DNA Alignment’; Intelli-

Genetics, Mountain View, California); final optimal alignments were done by

hand. Because of the length of our nucleotide sequences, we only present those

most conservative sites for Euhemiptera in Table 1; full sequences are deposited

in GenBank and are available there, or from us, upon request. The 18S rDNA of

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA 35

many of the hemipteran taxa, especially Sternorrhyncha, contained highly variable

expansion regions. These regions were synonymous to helices 10, E21, 41 and 47

of the secondary structure of synonomous 18S rRNA of A. pisum (Kwon et al.

1991, Campbell et al. 1994), and were largely unalignable (< 70%) at the higher

taxonomic levels studied here. Further, the 18S rDNA of several groups suggests

they have a higher clock-speed base substitution rate, causing unacceptable DNA

homoplasy between major clades. Because the effect of these swamped the more

conservative 18S rDNA regions that were needed to decipher the ancient topology

among major clades, we eliminated their influences through a sequential series of

attenuations.

PAUP (Swofford 1993: vers. 3.1.1), in both “branch and bound” and “ex-

haustive” search modes, was used for phylogenetic analyses. Gaps and deletions

were scored as missing (in SET 1, see below). Weighting (1:10) of transitions to

transversion did not affect tree topologies in any analyses. The PAUP algorithm

was employed because parsimony, as an optimality criterion, has been demon-

strated to show the greatest accuracy in converging on a phylogenetic topology

with equal rates of evolution, across the range of numbers of available base sites

(especially the least), for Kimura model of evolution and a 10:1 transition : trans-

version ratio (Hillis et al. 1994); also see Steel et al. (1993) and Sidow/Stewart

(1993) for further discussion of the parsimony criterion in nucleotide analyses.

Although our taxa initially indicated differential rates of base pair substitutions

among differing lineages (Campbell et al. 1994), the problem was dealt with by

selective removal of ancillary groups during the analyses to eliminate these effects

and increase resolution among retained taxa. Similar analytical procedures were

functionally employed on problematic 18S and 28S rRNA data for metazoans

and increased the resolution of their ancient phylogenetic topology (Christen et

al. 1991, Lafay et al. 1992, Smothers et al. 1994), and also have been employed

in phylogenetic reconstruction using continuous morphometric data that has been

transformed using ordinations (Sorensen 1992).

Of many sets of PAUP analyses that were run, four sets are presented here to

illustrate the effect of alignments, and of sequentially attenuating the homoplasy

encountered in the 18S rRNA gene in order to eliminate all but its most conser-

vative regions. This homoplasy was usually judged by relatively poor (1) consis-

tency indexes and (2) bootstrap numbers (but see Hillis & Jull 1993, Felsenstein

& Kishino 1993), and by (3) convergent site expression among only the more

terminal taxa between established sister-clades Sternorrhyncha and Euhemiptera.

Initially, our 18S rDNA extraction of a thrips (Frankliniella sp.) was considered

for tree rooting; however, it possessed an inordinate number of autapomorphic

nucleotides, which rendered it unsuitable, given the level of 18S rDNA homoplasy

in Hemiptera. We also considered a psocopteran for rooting, but were unable to

amplify the full 18S rRNA gene. Ultimately, we chose an available coleopteran,

because of the temporal (Permian) divergence involved, and because Coleoptera

is a basal clade in the Endopterygota, the sister-group to the hemipteroid lineage

(Hennig 1981, Kukalova-Peck 1991, Carmean et al. 1992). The beetle 18S rDNA

was used in conjunction with that of a psyllid, because Campbell et al. (1994),

and our initial analyses, determined that psyllids were the most basal group in

clade Sternorrhyncha; thus, Psyllidae are the nearest monophyletic out-group for

analyses of Euhemiptera.

36 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

In the first set of analyses (SET 1), all the 18S rDNA sites and all treated taxa

were employed, and generated trees were anchored using the beetle. SET 1 was |

divided into two subsets (1A, 1B), allowing an estimation of the effect of two

differing alignment orders (A, B) among their taxa. Each subset began with a

different taxon as its initial nucleotide alignment, and aligned the sequentially

varying remaining taxa to the first; this served as a check for potential homology

among sites, as deletions were inferred during alignment. The order of alignment

A (SET 1A) was: delphacid, mirid, cicada, cercopid, membracid, beetle, psyllid,

diaspidid, aphid, aleyrodid; this yielded 2738 (alignment inferred) sites, of which

336 were informative. The order of alignment B (SET 1B) was: membracid,

cercopid, cicada, mirid, delphacid, beetle, psyllid, diaspidid, aphid, aleyrodid; this

yielded 2773 (alignment inferred) sites, of which 307 were informative. Differences

in the number of sites between these alignments resulted from ambiguities in

aligning sites within variable helices.

The second set of analyses (SET 2) were also conducted on all treated taxa using

subsets with alignments A and B (SET 2A, SET 2B, respectively). The data from

both these subsets were attenuated, however, so that all inferred site deletions

were removed from each, along with all adjacent sites on both sides, back to

agreement across all taxa. This provided an objective and significantly more

conservative estimate of site homology and essentially eliminated subjectivity in

the interpretation of ambiguously aligned sites. The SET 2A attenuation yielded

1513 sites, of which 110 were informative; that of SET 2B yielded 1494 sites, of

which 100 were informative.

The third analysis set (SET 3) was conducted to eliminate the effect of site

homoplasy induced by the presence of more derived taxa within clade Sternor-

rhyncha, some members of which have greatly accelerated base substitution rates

for the gene (Campbell et al. 1994). In SET 3: the diaspidid, aphid and aleyrodid

were eliminated; the nucleotides were realigned in their absence using the align-

ment A (most informative sites) taxon order; and the tree was anchored using the

beetle. This yielded 1647 sites, of which 64 were informative. The total SET 3

site number increased over that of either SET 2 subset because deletions present

in the omitted taxa, and their pruning effect, were eliminated. The SET 3 number

of informative sites dropped from either of the SET 2 subsets, however, because

Synapomorphies among the omitted sternorrhynchans were also eliminated.

The final analysis set (SET 4) was conducted on the SET 3 taxa, but used the

most severe estimate of conservative sites available within Euhemiptera. The SET

4 analysis was based upon alignment A (most informative sites), but used: (1)

only those sites that could be individually out-group polarized in a Hennigian

sense, and (2) of those, sites showing parallel homoplasy between Sternorrhyncha

and Euhemiptera were excluded. Therefore, only those alignment A sites were

used that were plesiomorphic in both the beetle and psyllid (the sternorrhynchan

basal clade), but which were also nonhomoplasiously apomorphic within Euhem-

iptera, with respect to their lack of co-occurrence in Aleyrodiformes (diaspidid,

aphid, aleyrodid, sensu Campbell et al. 1994). Sites synapomorphic throughout

Sternorrhyncha, but plesiomorphic throughout Euhemiptera used, were also in-

cluded to give a measure of the support for clade Sternorrhyncha. Thus, SET 4

employed the 20 most conservative informative sites. The SET 4 topology was

manipulated using MacClade (Maddison & Maddison 1992), to explore its less

optimal alternatives.

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA 37

Topology Descriptors Used. —Here, brevity of text description for various tree

topologies favors the use of a slightly modified ‘““Newicks’s 8:45” tree description

standard, which is commonly used in phylogenetics. This format lists network

terminals as relative nested subsets within parenthetical enclosures; for further

definition, see Swofford (1993). We use curved brackets, { }, and italics to offset

the descriptors from the text. For example, {{4, B},{C, {D, E}}} describes the

topology: clade A+B as sister-group to clade C+D+E, and within the latter,

sister-group C to clade D+E. We also may inject bootstrap support numbers (BSS)

in the descriptors, as for example, {{A, B} 85, C} 73, where clade A+B bootstraps

at 85% and clade A+B+C at 73%. We abbreviate the taxa, sometimes including

larger recognized clade names where their internal topology are unimportant in

the given frame of reference, by their capitalized first five letters (e.g., {{MEMBR,

CERCO}, STERN} for {{Membracidae, Cercopidae}, Sternorrhyncha}).

RESULTS: THE 18S rDNA TREES

SET 1.—SET 1A, based on 2738 total sites [TS] and 336 informative sites [IS],

yielded a minimum length tree [MLT] (not shown) with a tree length [TL] of 847

and a consistency index [CI] of 0.58. In that topology: {{STERN} 92, {EUHEM}

60}. Within Sternorrhyncha: {{{APHID, DIASP} 100, ALEYR} 99, PSYLL} 92;

which supports clade Aleyrodiformes (sensu Campbell et al. 1994). Within Eu-

hemiptera: {{AUCHE} 53, MIRID} 60. Within Auchenorrhyncha: {{CICAD,

CERCO, MEMBR} 96, DELPH} 53; asa trichotomy within Cicadomorpha, with

sister-clade Fulgoromorpha. Thus, SET 1A supports an Auchenorrhyncha clade

but only at BSS 53.

SET 1B (based on 2773 TS, 307 IS) yielded a MLT (not shown) with TL 773

with CI 0.57. The MLT topology for SET 1B is similar to that for SET 1A in

that: {{STERN} 68, {EUHEM} 75}; showing lower bootstraps for Sternorrhyn-

cha, but higher for Euhemiptera. Also, within Sternorrhyncha: {{{APHID, DIASP}

100, ALEYR} 79, PSYLL} 68; showing lower bootstraps for the entire clade and

clade Aleyrodiformes. However, SET 1B shows paraphyly for Auchenorrhyncha,

with topology: {{CICAD, CERCO, MEMBR, MIRID} 54, DELPH} 54; where

the heteropteran forms a quadrachotomy at low bootstrap with the cicadomor-

phans, and fulgoromorpha is sister-clade to that grouping.

SET 1 resolves Sternorrhyncha, its internal topology, and Euhemiptera, but

does not resolve the origin of Heteroptera or potential monophyly for Auchen-

orrhyncha. The low Cls (0.58, 0.57) for SETs 1 indicate high homoplasy levels

in the data.

SET 2.—The same two equally parsimonious MLT topologies, shown in Figs.

1A, 1B, were produced by both attenuated SET 2A (based on 1513 TS, 110 IS)

and SET 2B (based on 1494 TS, 100 IS). For SET 2A, these MLT topologies had

TL 232 with CI 0.59; for SET 2B, they had TL 208 with CI 0.58.

Both these SETs 2 MLTs show topology: {{STERN} 93, {EUHEM} 97; the

increased euhemipteran bootstrap indicates that the first attenuation of the data

was successful in removing some homoplasy between it and Sternorrhyncha, due,

most probably, to the unique expansion areas of the 18S rDNA in the latter (see

Campbell et al. 1994). The internal sternorrhynchan topology is also preserved

with reasonable bootstraps, as: {{{APHID, DIASP} 99, ALEYR} 84, PSYLL} 93.

The two competing SET 2 MLTs, however, again differ in the placement of

Heteroptera within Euhemiptera, but both indicate polyphyly for Auchenorrhyn-

38 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

MEMBR

CERCO

CICAD

DELPH

MIRID

PSYLL

1A MEMBR a=

MIRID

ALEYR

DIASP

APHID

ALEYR

DIASP

APHID

COLEO COLEO

Figure 1. Two tying minimum length trees (Figures 1A, 1B) from PAUP analyses of SETs 2A and

2B. Alignments A (1513 TS, 110 IS) and B (1494 TS, 100 IS) both produced these tying MLTs. For

SET 2A: TL 232, CI 0.59; for SET 2B: TL 208, CI 0.58. Branch & Bound bootstrapping for this data

indicates support for clade Euhemiptera (BSS 97) and clade Sternorrhyncha (BSS 93); within Euhem-

iptera: quadrachotomy {MIRID, CICAD, DELPH, {MEMBR, CERCO} 52}; within Sternorrhyncha:

{{{APHID, DIASP} 99, ALEYR} 84, PSYLL} 93. The Fig. 1A MLT indicates a potential origin of

Heteroptera may be associated with cicadomorphans. The Fig. 1B MLT indicates clade Fulgoromor-

pha+ Heteroptera with sister-clade Cicadomorpha. Both MLTs show internal topology for Sternor-

rhyncha as per Campbell et al. (1994). Topology internode lengths are proportionate to number of

anagenic base substitutions present in SET 2A data set (alignment A), which is a function of the

induced groups and their informative sites in the nucleotide matrix.

cha. The first topology indicates: {{{{MEMBR, MIRID}, CERCO}, CICAD},

DELPH}; with Heteroptera originating from the more terminal end of an oth-

erwise paraphyletic Auchenorrhyncha and Cicadomorpha. The second indicates:

{{{MEMBR, CERCO}, CICAD}, {DELPH, MIRID}}; with monophyly for Ci-

cadomorpha, polyphyly for Auchenorrhyncha, and clade Fulgoromor-

pha+ Heteroptera as sister-group to Cicadomorpha. SET 2 bootstraps for Euhem-

iptera show the quadrachotomy: {{MEMBR, CERCO} 52, CICAD, DELPH,

MIRID} 97.

SET 2 confirms the topology of Sternorrhyncha. Within Euhemiptera, it does

not resolve the origin of Heteroptera or monophyly of Cicadomorpha, Although

it indicates polyphyly for Auchenorrhyncha. The low SETs 2 Cls (0.58, 0.59)

continue to indicate the presence of high homoplasy levels.

SET 3.—SET 3 (based on 1647 TS, 64 IS), which eliminated all Sternorrhycha

except Psyllidae, produced 945 possible trees; its MLT, with TL 117, and the

next five shortest trees, with TLs 118-120, are shown in Figs. 2A—F. The SET 3

MLT topology (TL 117) for Euhemiptera shows: {{{CERCO, CICAD}, MEMBR},

{DELPH, MIRID}}; indicating monophyly for Cicadomorpha with clade Ful-

goromorpha-+ Heteroptera as its sister-group, and polyphyly for Auchenorrhyn-

cha. The second and third shortest SET 3 topologies confirm this, and differ only

in their internal topology within Cicadomorpha, as: {{MEMBR, CICAD}, CER-

CO} at TL 118, and {{MEMBR, CERCO}, CICAD} at TL 119. The three to-

pologies tying for fourth most parsimonious place, at TL 120, all place Heteroptera

at various origin points within (extant) Cicadomorpha (Figs. 2D-F). Thus, the

several most parsimonious SET 3 trees indicate polyphyly for Auchenorrhyncha.

SET 4.—The out-group polarized data of SET 4 (20 IS only), yielded the MLT

in Fig. 3, with TL 29 with CI 0.72. In this analysis, only the most conservative

informative sites available for inference of euhemipteran topology were used (see

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA 39

2A MEMBR 2B MEMBR 2C MEMBR

CERCO CICAD CERCO

CICAD CERCO CICAD

DELPH

MIRID

STERN

COLEO (TL117)

DELPH

MIRID

STERN

COLEO (TL 118)

DELPH

MIRID

STERN

COLEO (TL119)

2D MEMBR OE MEMBR OF MEMBR

CICAD MIRID CERCO

MIRID CERCO MIRID

CERCO CICAD CICAD

DELPH DELPH DELPH

STERN STERN STERN

COLEO (TL 120) COLEO (TL 120) COLEO = _ (TL 120)

Figure 2. MLT (Figure 2A) from SET 3 PAUP analysis, based on 647 TS and 64 IS, yielding TL

117. Figures 2B—F show topologies for second (Figure 2B: TL 118), third (Figure 2C: TL 119) and

fourth (Figures 2D-F: TLs 120) best levels of parsimony. The MLT plus the second and third most

parsimonious topologies indicate clade Fulgoromorpha-+ Heteroptera with sister-clade Cicadomorpha;

a heteropteran association with Cicadomorpha does not occur until the fourth best parsimony level.

methods discussion for SET 4). The MLT topology for Euhemiptera was

{{{MEMBR, CERCO}, CICAD}, {DELPH, MIRID}}; again this indicates mono-

phyly for Cicadomorpha, with sister-group clade Fulgoromorpha+ Heteroptera,

and polyphyly for Auchenorrhyncha. i

Support for Non-monophyly of Auchenorrhyncha.—Given a parsimony crite-

rion, none of our 18S rDNA analyses indicate monophyly for Auchenorrhyncha.

Instead, Auchenorrhyncha was always indicated to be para- or polyphyletic be-

cause usually either Heteroptera arises: (1) as a sister-group to Fulgoromorpha,

the two forming a clade that itself assumes a sister relationship to clade Cica-

domorpha; or (2) from within the (then nonmonophyletic) Cicadomorpha. In fact,

clade Cicadomorpha with sister-group Heteroptera is more parsimonious than

clade Auchenorrhyncha.

In our most conservative and preferred analysis, SET 4, the clades in the MLT

are supported by the following numbers of synapomorphies (our alignment A

numbers for SET 2 sequences, first attenuation), with transitions indicated by *

and transversions by ** (Table 1). Sternorrhyncha: 5 nonhomoplasious unam-

biguous synapomorphies (sites: 62 [A — G*], 152 [A — C**], 153 [A — G*],

1110 [G— T**], 1269 [T — C*]). Euhemiptera: 2 nonhomoplasious unambiguous

synapomorphies (sites: 53 [T — A**], 159 [C — T*]), plus potentially 2 homo-

plasious ambiguous synapomorphies (sites: 241 [C — A**] with reversal in Del-

phacidae, 457 [G — T**] with reversal in Miridae); the latter two ambiguities are

equivocal in support of either Euhemiptera or Cicadomorpha, however; in ad-

dition, site 79 [G — A*] is apomorphic for the Euhemiptera, but is homoplasious

with some Diptera (see Table 1, also see Carmean et al. 1992). Cicadomorpha: 2

nonhomoplasious unambiguous synapomorphies (sites: 721 [G — T**] with in-

dependent mutation in Miridae [G — A*], 1251 [C — T*}); also potentially plus

the two ambiguous sites stated to be equivocal for Euhemiptera. Clade Cercop-

idae+Membracidae: 1 homoplasious unambiguous synapomorphy (site: 263 [G

40 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Mm +r

STERN 2 : MEMBR

SHS :

~rrn CERCO

[| m, zm

ee gee

ie al CICAD

(9p)

BE: eas

| f a E DELPH

BERED

7 MIRID

457 [|_|

* 454

1025 BE

reise

198

263

721

Figure 3. MLT produced from the SET 4 data set (20 IS only), where sternorrhynchan homoplasy

was excluded using out-group polarization. Bars on internodes represent synapomorphies labelled

with their site numbers (Table 1); black = nonhomoplasious and unambiguous site change; gray =

homoplasious (within Euhemiptera) but unambiguous site change; white = homoplasious (within

Euhemiptera) and ambiguous site change. Sites 79 and 454 (white to black gradients with asterisk)

are homoplasious outside this analysis; 79 is homoplasious in dipterans and 454 in some heteropteran

lineages (see Table 1). Sites 241 and 457 are ambiguous site changes, marked by ?, that may occur

either along the euhemipteran ancestral internode, or alternatively along the cicadomorphan ancestral

internode. Site 721 is an independent transformation on the cicadomorphan and heteropteran ancestral

internodes. The MLT supports clade Fulgoromorpha-+ Heteroptera.

— A*] parallelism in Miridae). Clade Fulgoromorpha+ Heteroptera: 1 nonhom-

oplasious unambiguous synapomorphy (site: 1117 [G— A*]) and 2 homoplasious

unambiguous synapomorphies (sites: 454 [G — A*], homoplasious within het-

eropterans in Wheeler et al’s. (1993) sequences, and 1025 [G — T**], a parallelism

in Membracidae). Although the single representatives for Fulgoromorpha and

Heteroptera used were thought to preclude informative sites as synapomorphies

for them, Heteroptera showed the mentioned independent mutation of site 721

[G — A*]. (Synapomorphies for each of Fulgoromorpha and Heteroptera are

available in our subsequent analyses, see footnote 7).

In the SET 4 MLT, clade Fulgoromorpha+ Heteroptera precludes Auchenor-

rhyncha monophyly, yet it is based on 1 nonhomoplasious transition synapo-

morphy (site 1117) and (in “opposition’”’?) 2 homoplasious synapomorphies, a

transition (site 454) showing homoplasy in some heteropteran lineages (Wheeler

et al. 1993), and a transversion (site 1025). Some authors suggest transversion/

transition mutation biases are present in some nucleotide data (e.g., primate

mtDNA), and that a 10:1 weight should be imposed in favor of transversions for

phylogenetic inference (Mishler et al. 1988, Patterson 1989, Michevich & Weller

1990). If so, such weighting could affect MLT generation towards a topology

optimizing transversions over transitions. In fact, even a philosophical preference

towards a transversion bias should tend to negatively affect the relative acceptance

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA 41

of a topology where transitions appear to dominate over tranversions on given

cladogram internodes (e.g., clade Fulgoromorpha+ Heteroptera on the SET 4 MLT).

However, 18S rDNA does not appear to show such bias for those secondary

structural portions of the molecule termed “‘bulges”’ or “‘loops” (Vawter & Brown

1993). The transition synapomorphy of clade Fulgoromorpha-+ Heteroptera oc-

curs on such a secondary substructure. Our site 1117 (= site 1715 of Kwon et al.

1991) occurs on a “‘bulge’”’ (Kwon et al. 1991: fig. 3, bulge position of helix 40),

where a transversion bias was not present for 18S rDNA (Vawter & Brown 1993).

Thus, for our 18S rDNA sequences, equal weights for transitions and transversions

are appropriate, so that a prejudice against the SET 4 MLT is unreasonable.’

Nevertheless, as an alternative to the SET 4 MLT, we explored other, less

parsimonious SET 4 topologies that would permit monophyly for Auchenor-

rhyncha. Using PAUP and the SET 4 data, we tabulated the probability of mono-

phyly for Auchenorrhyncha and other groups sequentially across all possible TLs

(29-42), as decreasing levels of parsimony (Table 2). For each rising TL level, we

noted the accumulative numbers of trees containing each of 5 possible clades: (a)

Euhemiptera, (b) Cicadomorpha, (c) Cicadomorpha-+ Heteroptera, (d) Fulgoro-

morpha-+ Heteroptera, and (e) Auchenorrhyncha; any internal topology was per-

mitted for the member taxa of each “clade.” These accumulations were trans-

formed to probabilities (of existence) for the clades, as their frequency of occurrence

(i.e., the accumulated total number of trees containing a clade at each TL, divided

into the number of trees possible at that TL). The probabilities of clades Cica-

domorpha+ Heteroptera and Fulgoromorpha-+ Heteroptera, and their total, can

also be taken as a function of probability for non-monophyly for Auchenorrhyn-

cha, because of conflicting relative association of Heteroptera. In Table 2, clade

Auchenorrhyncha does not exist until the third best parsimony level (TL 31),

where it occurs on only 2 of 23 possible trees (0.09), and that by that level,

competing clades Fulgoromorpha+ Heteroptera (0.52) and Cicadomor-

pha+ Heteroptera (0.39) both occur at greater frequencies (2 0.91). Auchenor-

rhyncha rises to its greatest frequency (0.18) at TL 32, where it remains the least

probable clade; it rises to its greatest occurrence (30 trees of 822 retained and 945

possible) at TL 38, where it ties with competing clade Cicadomorpha+ Heteroptera

7 This synapomorphy is supported in additional analyses involving a more extensive sampling of

taxa (Campbell et al., unpublished data), to be published elsewhere: [GenBank accession numbers in

parentheses] Cercopidae-Tomaspinae (U16264), Cicadellidae-Cicadellinae (U15213), Cicadellidae-

Deltocephalinae (U15148), Cixidae (U15215), Dictyopharidae (U15216), Flatidae (U06476), Gerridae

(U15691), Issidae (U15214), Lygaeidae (U15188). Given the fact, in matrix generation of MLTs, that

holding character number constant, and either decreasing average state number or increasing terminal

taxa number, effectively increases the probability of homoplasy, we chose here to increase the total

number of 18S rDNA base pairs analyzed to maximize the discovered synapomorphies. Based on the

distribution of synapomorphies throughout differing regions of the 18S rDNA gene, it may not be

possible to infer accurate phylogenetic conclusions using short segments (i.e., 6-700 base pairs) of the

gene. We have compared our sequences with those of Wheeler at al. (1993) and Carmean et al. (1992)

for site homoplasy. Functionally, “throwing more taxa’’ at this problem will merely (a) validate, or

negate, the existing synapomorphies among the presented basal topology, (b) supply synapomorphies

for morphologically obvious clades (e.g., Fulgoromorpha), or (c) permit insertion of excluded taxa

(e.g. Coleorhyncha). The Campbell et al. (to be published) analyses, which increase taxa, will verify

non-monophylly for Auchenorrhyncha and discuss mutation rate differences for regions of the 18S

rDNA gene. See note added at end of Literature Cited.

Table 2. Accumulative frequency of selected ‘‘clades”’ across all tree lengths for 945 possible trees from SET 4 data.

Tree length

29 30 31 32 33 34 35 36 37 38 39 40 41 42

Trees generated? 1 9 23 51 74 109 127 169 244 383 582 700 791 822

Euhemiptera 1 9 23 50 72 99 103 105 105 105 105 105 105 105

(1.00) (1.00) (1.00) (0.98) (0.97) (0.91) (0.81) (0.62) (0.43) (0.27) (0.18) (0.15) (0.13) (0.13)

Cicadomorpha 1 4 8 10 11 15 23 35 35 35 35 35 35 35

(1.00) (0.44) (0.35) (0.20) (0.15) (0.14) (0.18) (0.21) (0.14) (0.09) (0.06) (0.05) (0.04) (0.04)

Cicadomorpha+ 0 5 12 15 15 22 26 30 30 30 30 30 30 30

‘Heteroptera? (0) (0.56) (0.52) (0.29) (0.20) (0.20) (0.20) (0.18) (0.12) (0.08) (0.05) (0.04) (0.04) (0.04)

Fulgoromorpha+ 1 4 9 16 17 17 17 23 46 54 78 78 78 78

Heteroptera (1.00) (0.44) (0.39) (0.31) (0.23) (0.16) (0.13) (0.14) (0.19) (0.14) (0.13) (0.11) 0.10) (0.09)

Auchenorrhyncha 0 ) Z 9 11 16 18 22 25 30 30 30 30 30

(0) (0) (0.09) (0.18) (0.15) (0.15) (0.14) (0.13) (0.10) (0.08) (0.05) (0.04) (0.04) (0.04)

@ Retained.

+ Any internal topology allowed among member taxa.

(44

LSIDOTOWOLNG OWMIDVd-NVd FHL

(DIL TOA

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA 43

for fewest tree numbers. Thus, given the conservative data of SET 4, which was

designed to optimally eliminate the more recently derived homoplasious sites

that obfuscate evolutionary topological relationships among the older clades, we

find scant evidence for the possibility of monophyly for Auchenorrhyncha.

CLADISTIC IMPLICATIONS

Until now, Auchenorrhyncha was generally considered to be monophyletic on

the basis of either molecular data from insufficient subgroups (Wheeler et al.

1993), or morphological traits previously considered to be valid synapomorphies

(but see Hamilton 1981). A recent example of the latter is Carver et al.’s (1991:

464) statement that “The monophyly of the Auchenorrhyncha ... is firmly es-

tablished by the complex tymbal acoustic system and the aristoid antennal fla-

gellum characteristic of the group.”’ Other traits across Auchenorrhyncha have

been considered symplesiomorphies, with the exception of a fused ScP+R vein

apomorphy (Kukalova-Peck 1991: 170).

However, phylogenetic reconstruction using nucleotide sequencing is thought

to be superior to, and definitely more objective than, that based upon morphology

(Felsenstein 1982, 1983, 1988; Crespi 1992; Sorensen 1992). This is because, in

general, nucleotide substitutions are random, non-selective events, as opposed to

trying to determine how to code and weight morphological characters, which are

defacto a result of selection. Morphologically-based phylogenetics is conceptually

plagued by the inherent effects of selection and character correlation; although

these are easily recognizable, they are nearly impossible to handle (see Sorensen

1990, 1992). Use of nucleotides not only renders a portrait that is essentially free

of these problems (Lewontin 1989), but permits character transformation overlays

that allow recognition of morphological homoplasy. If the 18S rDNA phylogeny

derived here is correct, it is evident that the morphological synapomorphies for

Auchenorrhyncha must be convergences that are most probably selection-induced.

Tymbal Systems as Homoplasy. — Although the development of a complex tym-

bal system for sound production may seem like a strong synapomorphy for Au-

chenorrhyncha, this mechanism is homoplasious in Hemiptera and clearly is under

strong sexual selection. Tymbal systems not only exist in Cicadomorpha and

Fulgoromorpha, but they also occur in Pentatomomorpha (e.g., Pentatomidae:

Carpocoris; Chapman 1971), a highly derived and phylogenetically distant clade

(Wheeler et al. 1993), where their position and function appears to be similar to

that within most Auchenorrhyncha. Furthermore, despite many investigations

into tymbal sound production in various Auchenorrhyncha (Ossiannilsson 1949,

Smith & Georghiou 1972, Shaw & Carlson 1979, Mitomi & Okamoto 1984,

Zhang & Chen 1987, Zhang et al. 1988), except for Cicadidae (Pringle 1954, 1957),

precise and convincing physiological mechanisms of their function in leafhoppers

or planthoppers have not yet been published (Claridge 1985, Claridge & de Vrijer

1994) and remain, at best, controversial.

In Ossiannilsson’s (1949: 103-106) discussion of morphology, there are many

significant differences between the fulgoromorphans (Delphacidae [as ‘“‘Areopi-

dae’’], Cixidae, Issidae) and cicadomorphans (Cercopidae, Cicadellidae, Mem-

bracidae) that he examined. Examples of these differences include Fulgoromor-

pha’s lack of (a) a “striated tymbal” (shared with some cicadellids) and (b) a

“pilose surface”; their (c) “enlarged metapostnotum,” (d) “‘less developed meta-

44 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

postphragma,”’ (e) “well developed lateral dorsal longitudinal muscles” (except

in brachypterous forms); their (f) ““second abdominal tergum”’ being “‘devoid of

phragmata in spite of the dorsal longitudinal muscles of the first abdominal seg-

ment being strongly developed”; and their (g) “second tergum being strongly

vaulted into a convex, shield-like surface with inner strengthening lists,’ which

serve as the posterior attachment of the longitudinal muscles from the meta-

postnotum.

Perhaps the best reference to Fulgoromorpha’s tymbal variance is summarized

by Ossiannilsson’s (1949: 104) statement that homology across the Auchenor-

rhyncha for muscle J a dvm, “‘. . . might be uncertain for only the Fulgoromorpha,

as the conditions of this group are so deviating... .” It should be evident that

this uncertain homology between fulgoromorphan and cicadomorphan tymbal

systems does not seem adequate to be regarded as a convincing synapomorphy

for these groups, especially in light of the occurrence of an (at least superficially)

similar tymbal mechanism in the Pentatomomorpha. Clearly more detailed tym-

bal comparisons are needed.

Aristoid Antennae as Homoplasy.—It is easier to accept the reduction to an

aristoid antennae among the auchenorrhynchan groups as homoplasy if one re-

members that all Pterygota and Thysanura have annulated (or flagellar) antennae

(sensu Schneider 1964: type B; Chapman 1971: type A), as opposed to true

segmented antennae (sensu Schneider 1964: type A; Chapman 1971: type B),

which occur in the apterogote subclasses Collembola and Diplura. In segmented

antennae, each true segment, including the scape, pedicel and each flagellar seg-

ment has up to five intrinsic muscles connecting its base to the base of the next

distal segment, and these permit intersegmental movement. In annulated anten-

nae, however, only the scape has such segmental musculature, whereas each fla-

gellar “‘segment,” all of which are actually mere annulations, is connected to the

next by membrane only; annulated antennae are moved only by levator/depressor

muscles connecting the anterior tentorial arms to the scape, and flexor/extensor

muscles connecting the scape to the pedicel (Imms 1940). Thus the flagellum of

the Pterogota is a single, functional unit that has already undergone reduction

from true segmentation to mere annulation, and it has undergone many homo-

plasious further reductions across diverse taxa (i.e., larval Holometabola, adult

Mallophaga/Anoplura, adult Brachycera/Cyclorrhapha Diptera, etc.).

Among hemipterans, only those that jump have evolved aristoid flagella. How-

ever, differing forms of flagellar reduction occur independently at least twice in

Hemiptera (i.e., Peloridiidae and Nepomorpha) besides that noted in Auchen-

orrhyncha. Cicadomorphan and fulgoromorphan convergence towards an aristoid

antenna results, we believe, from selection to: (a) minimize injury; (b) enhance -

aerodynamic streamlining; and/or (c) allow acoustic receptions via Johnston’s

organ. Because of its sensory function, selection to avoid or minimize antennal

damage should be an extremely strong force. Antennal injury should be lessened

during the less controlled, head-first landings encountered in jumping. Further,

jumping with large antennae would enhance aerodynamic instability in small,

bullet-like auchenorrhynchans that leap with almost explosive force (K. G. A.

Hamilton, personal communication). In contrast, the more massive bodies of

larger jumpers with long antennae (e.g., orthopteroids) probably minimize anten-

nal contributions toward instability.

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA 45

Aristoid flagella on hemipterans occur only in the presence of tymbal trans-

missions systems, but not the reverse (e.g., Pentatomomorpha). Thus, aristoid

antennae may also serve as an acoustic reception device, picking up air-transmitted

vibrations and transferring them to Johnston’s organ, a chordotonal organ in the

antennal pedicel. Although many Auchenorrhyncha apparently transmit sound

through the substrate, the acoustic receptors remain largely unknown (Claridge

1985), but Johnston’s organ has been suggested as such a receptor (Howse &

Claridge 1970), and clearly serves such a function in other insects (Chapman

1971).

If the development of the scape+pedicel versus the flagellum are considered,

the antennal systems of cicadomorphans and fulgoromorphans appear only su-

perficially similar in their respective ultimate development. The fulgoromorphan

pedicel is exceptionally developed (e.g., Hamilton 1981: figs. 18, 19), with nu-

merous autapomorphic plaque sensilla (e.g., Baker & Chandrapatya 1993: figs. 1,

2) that vary across the group (Marshall & Lewis 1971). Fulgoromorphan flagellar

annulation is generally extreme, appearing as mere thin rings, except for a rela-

tively enlarged, bulbous flagellomere | (= antennal 3) (e.g., Baker & Chandrapatya

1993: fig. 6), which has an autapomorphic sensory organ (Bourgoin 1985: fig. 2,

“OSBF’’) throughout the group.

Supportive evidence for the homoplasious evolution of an aristoid antenna in

fulgoromorphans is provided inadvertently by Megaleurodes megocellata Ham-

ilton (1990: fig. 34), a fossil from Brazilian Lower Cretaceous deposits (AMNH

type 43608). Hamilton (1990: 96) thought it was a primitive whitefly with ful-

goromorphan traits, and assigned it to ““Aleyrodoidea: Boreoscytidae?”’. However,

the traits with which M. megocellata was assigned to Aleyrodidae are either quite

-homoplasious in Hemiptera (e.g., divided eye, ocellar position) or symplesio-

morphies (K. G. A. Hamilton, personal communication). Because of its facial ca-

rinae, tegulae and three-segmented tarsi (the latter two symplesiomorphies) we

believe it is a primitive fulgoromorphan that shows non-aristoid antennae that

arise fairly high on the face. Therefore, we tentatively reassign Megaleurodes

megocellata to the fossil superfamily Fulgoridioidea, but with an uncertain family

assignment. It is similar to the Jurassic Fulgoridiidae (sensu Bode 1953) in that

its antennae are multiarticulate (non-aristoid), a diagnostic plesiomorphy for that

(gradistic ?) family (Bode 1953, Hamilton 1990); but the head of Megaleurodes .

differs from that of Fulgoridium (Bode 1953: fig. 143) with its ocelli below the

eyes (K. G. A. Hamilton, personal communication). We consider the fossil Ful-

goridioidea to be an extinct grade to the modern Fulgoroidea, within Fulgoro-

morpha, and to demonstrate the lineage initially had non-aristoid antennae.

In contrast, in many cicadomorphans, the antenna generally has a less developed

scape and pedicel and a less reduced flagellum, where flagellar annulation (“‘seg-

mentation’) is still usually quite evident (e.g., Cwikla & Freytag 1983: fig. 4). In

some cicadas, the flagellum is still reasonably developed (e.g, Hamilton 1981: fig.

14), especially in nymphs (e.g., Hamilton 1981: fig. 2), which retain a developed,

definitely ““segmented,”’ but short flagellum. Interestingly, Cicadas do not jump,

and their tympana also serve as acoustic receptors. Nevertheless, some cercopids

have an aristoid antennal flagellum that appears to approach that of Fulgoro-

morpha. These have flagellomeres 2-n quite annulated and an enlarged flagello-

mere 1, possibly with a sensory organ that externally appears somewhat similar

46 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

to that of fulgoromorphans. Shcherbakov (1988), however, states that fossil Pro-

cercopidae and Karajassidae, the initial cercopoids and cicadelloids, respectively,

retained a “segmented” antennal flagellum. This would necessarily indicate a more

recent, independent derivation of the cicadomorphan arista than would have to

occur if it was synapomorphic on a postulated monophyletic auchenorrhynchan

ancestral stem.

If our 18S rDNA inferred relationships between Fulgoromorpha and Cicado-

morpha are correct, homoplasy for aristoid flagellar development is required. The

fulgoromorphan-like antennae of .cercopids appears necessarily unparsimonious

on any cladogram containing extant taxa, with sister-group Fulgoromorpha and

any internal topology for Cicadomorpha (unpublished data); the early fulgoro-

morphans present a similar problem. If aristoid antennae were derived only once,

on the euhemipteran ancestral phylogenetic internode, the character requires at

least two-steps on the 18S rDNA-based topology, with a reversal on the heter-

opterodean ancestral internode; independent derivation on each of the ancestral

internodes for cicadomorphans and fulgoromorphans is equally parsimonious.

Fused ScP+R Vein. —Kukalova-Peck, following her own venation terminology

(Kukalova-Peck 1983), which is also used here, states that for Auchenorrhyncha,

ScP- supports R, as a fusion apomorphy (Kukalova-Peck 1991: 170). She notes

that in Heteropterodea, however, ScP- is independent of R, asa symplesiomorphy;

yet she also shows an apomorphic ScP+R fusion in the Coleoptera (Kukalova-

Peck 1991: fig. 6.28E) in the hindwing (the beetle flight wing). Dworakowska

(1988) reviews the venation of Auchenorrhyncha, following Kukalova-Peck’s ter-

minology, and details many auchenorrhynchan wings, but her excellent study

shows the limitations of homological interpretation of venation. Also see Wootton

(1979) for discussion of problems in determining vein homologies.

Although the auchenorrhynchan fusion of ScCP+R seems reasonable as a syn-

apomorphy, we believe that it is a homoplasy. Convergence in venation occurs

commonly in hemipterans (Wootton & Betts 1986), particularly among their early

fossils (Wootton 1981), and is probably related to selection for various flight

dynamics parameters (Betts 1986a, b, c). The auchenorrhynchan ScP+R fusion

probably results from selection for rigidity in the basal region of the wing, coupled

with the developing need for.a point or area of flexion, just beyond, near midpoint

of the forewing margin. These modifications of the primary flight wing are required

for camber control during flight in heteropterans (Wootton & Betts 1986; Betts

1986a, b, c). Function-based similarities in wing geometries also appear to have

been derived in more phylogenetically advanced orders, for example Hymenoptera

(see Whitfield & Mason 1994: figs. 3-8). Another function-based homoplasy is

the development of an expansion of the wing’s precostal strip, to form an epi-

pleuron in Auchenorrhyncha and ‘Coleoptera (Kukalova-Peck 1991: 167).

To promote greater flight efiiciency, we believe differing wing geometries were

evolved and tested among early hemipterans. This resulted in structural conver-

gence in response to the selective constraints imposed by physical factors. We

feel such homoplasy can often be recognized by subtle differences among clades,

however. For example, where ScP eventually reaches the forewing margin in

Cicadomorpha, a venation break occurs where component C should merge

smoothly with liberated component ScP, as occurs in Fulgoromorpha. To illustrate

this point, consider the modifications of the three veins of the costal (pronating)

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA 47

complex, while bearing in mind the auchenorrhynchan ScP+R fusion. As Pc, CA

(= C+) and CP (= C—), these veins usually form a relatively tight beam-like triad

along the leading edge of the fulgoromorphan forewing (Dworakowska 1988: figs.

3-5 cross-sections). This wing-leading “‘beam’’ is undoubtedly for structural re-

inforcement.

In some groups (e.g., Eurybrachyidae, Flatidae, Lophopidae, Nogodinidae, Ri-

caniidae, Tropiduchidae, some Fulgoridae; Dworakowska 1988), Pc rolls ventrally

to more closely associate with CP (as fused Pc+CP), which leaves CA alone to

form the forewing’s functional anterior margin. However, CP never exists sepa-

rately at the wing base. In Fulgoromorpha, when CP posteriorly separates from

CA more distally, CP gives rise to several serial branches along the wing’s anterior

margin. This is what Kukalova-Peck (1991: 170) refers to as a false ‘subcosta’.

In such instances among fulgoromorphans, where this posteriorly moved and

serially branched CP occurs, ScP+R splits distally, and shortly thereafter the

liberated ScP curves to the forewing’s anterior margin to fuse with CP, as ScP+CP

(Dworakowska 1988: figs. 29, 37c, 41, 67); meanwhile, the abandoned R com-

ponent continues distally to the wing margin, also to split as RA and RP (and

usually each again). In contrast, in Cicadomorpha, where CP remains nearer the

anterior margin of the forewing throughout its course, this C/ScP abutment occurs

as an unfused association (Dworakowska 1988: figs. 94, 97).

Clearly structural selection is involved in this difference because the cicado-

morphan situation promotes flexibility at that point along the wing margin. The

homoplasious coleopteran ScP+R fusion allows hind(flight)wing folding under

their elytra, with the appropriate articulation. It may also be possible that the

auchenorrhynchan ScP+R fusion merely reflects a strengthening of the front-

basal or proximal area of the wing, enabling CP to travel distally to its ultimate

fusion/abutment with the ultimately liberated ScP, allowing the nodal flexion

point. If so, it should not be surprising that in some auchenorrhynchans, proximal

fusions of ScP and R with M also occur, permitting even greater stiffening, as

either ScP+R+MA (Dworakowska 1988: fig. 42), or, particularly among fossils,

ScP+R-+M (?) (Hamilton 1990: figs. 6, 41, 42, 58, also apparently 31, 33, 65,

69, 74, 75).

In some Fulgoromorpha, the ScP+C fusion point marks the distad border of

tegminization (e.g., Hamilton 1990: fig. 52), or an apparent “‘pterostigma’”’ in some

fossils (e.g., Hamilton 1990: figs. 46A, 55, 56, 58, 75, 80). In Cicadomorpha, the

C/ScP abutment break marks a quite primitive line of flexion (Hamilton 1990:

fig. 31, ““Cicadoprosbolidae”; Dworakowska 1988: figs. 92, 93, 95). It is the an-

terior of the flexion line that permits camber change during the wing beat (Wootton

& Betts 1986), and as such is under strong selection pressures.

Within clades Fulgoromorpha and Cicadomorpha, many other venation as-

sociations or fusions change, at least in part, sometimes quite notably. For in-

stance, the free base of Sc in Cercopoidea and Cicadoidea (Shcherbakov 1981:

66), or Pc+C+Sc+R amalgamation in some cicadas (Carver et al. 1991: fig.

30.4f). We believe that these differences between cicadomorphans and fulgoro-

morphans serve to: (a) cloud the potential questions of homology; (b) demonstrate

the “‘dancing,”’ but functionally related, fusion of the axial unit-radial complex

veins (sensu Dworakowska 1988) among auchenorrhynchans; and (c) illustrate

the apparent need of fusion in that region of the membranous wing of these highly

48 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

active insects, to achieve rigidity among the more basal components of these

veins.

In the other hemipteran clades, selection for a convergent ScP+R fusion may

have been alleviated by non-active habits or differential wing evolution. Ster-

norrhynchans are smaller, often with “passive”’ flight. Coleorhynchan forewing

venation is often quite thickened and pronounced (Kukalova-Peck 1991: fig. 6.25J;

Popov & Shcherbakov 1991: figs. 9, 10, 12, 22, 23, 35). Among the more derived

Heteroptera, the wing may be quite sclerotized proximally, with a developed

cuneus and costal fracture, while among the basal heteropteran Enicocephalo-

morpha, which have relatively membranous wings, it is the forewing venation

that is thickened.

Morphological Synapomorphies Supporting the 18S rDNA Tree.—The under-

standing of “homopteran’”’ phylogenetic topology has always been plagued by an

abundance of character homoplasy within and among groups, and the dearth of

convincing morphological synapomorphies indicating the relationships among

the major clades; the latter appears to be an artifact of limited local perspective

(sensu Sorensen 1992). The: 18S rDNA topology here cannot “correct” these

problems; it can merely illustrate those few nonhomoplasious synapomorphies

that appropriately structure the topology of the corresponding morphological tree.

We disagree with methods employed elsewhere (e.g., Wheeler et al. 1993), wherein

combinations of molecular and morphological traits are used in the same analyses

to increase phylogenetic resolution. We find such character amalgamations to be

philosophically and pragmatically untenable. We consider a combinable-com-

ponent approach (i.e., Bremer 1990, Lanyon 1993) among various competing

topologies that are based on differing dataset types, as appropriate to define to-

pological reliability, if required. We have avoided a separate, comparative mor-

phological analysis here, however, because we concur with Felsenstein (1988), at

least in this case, that morphological characters for phylogenetic inference are

inherently problematic. For example, we believe that a meaningful, morpholog-

ically-based phylogenetic analysis of hemipterans must, at the very least, reflect

an apriori understanding of their historical homoplasies, as well as their coding/

scoring consequences, and must appropriately include all fossil taxa.

The morphological synapomorphies that support the 18S rDNA-based topology

follow, as developments (gains), unless otherwise noted. Clade Sternorrhyncha—

(a) a sternorrhynchan-type filter chamber (Evans 1963: type A); see Fig. 3 and

Campbell et al. (1994) for 18S rDNA synapomorphies. Clade Euhemiptera—(a)

a vannus (Wootton & Betts 1986); (b) pronounced separation of costal and sub-

costal basivenale (Kukalova-Peck & Brauckmann 1992); (c) loss of ScA+ vein

(Kukalova-Peck & Brauckmann 1992). Clade Cicadomorpha—(a) a cicadomor-

phan-type filter chamber (Evans 1963: type B); (b) a ledge overhanging the antennal

insertion (Hamilton 1981; K. G. A. Hamilton, personal communication); (c) the

lorum with a wide connection to hypopharynx and a very narrow connection to

the gena (Hamilton 1981; K. G. A. Hamilton, personal communication); and (d)

a spiral-fold or -lobed wing-coupling system (D’Urso & Ippolito 1994: type A),

modified from the sternorrhynchan system wherein one or more spiral hooks

occur instead (D’Urso & Ippolito 1994: 223). Clade Fulgoromor-

pha-+ Heteropterodea—(a) slight reduction of the lorum (Hamilton 1981), as in-

termediate step to Heteropterodea (K. G. A. Hamilton, personal communication);

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA 49

(b) apical fusion of forewing veins 1A and 2A (Wootton & Betts 1986); (c) a long

and longitudinally-directed forewing vein CuA (Wootton & Betts 1986); ques-

tionably (e) often strong microspines, as accessory microsculpture, on the vein

opposite the fold of the wing-coupling apparatus (D’Urso & Ippolito 1994: 223);

and (d), of uncertain polarity, lack of a hindwing ambient vein; also, although

polarities are uncertain, we suspect that several alimentary canal modifications

shared by Fulgoromorpha, Coleorhyncha and Heteroptera (Goodchild 1966) rep-

resent plesiomorphies for a gut that lacks any filter chamber type—these include

a ‘pylorus’, a sac-like rectum, reduced rectal glands, and a midgut-hindgut junction

situated posteriorly in the body cavity. Clade Fulgoromorpha—(a) specialized

facial carina (Hamilton 1981); (b) a collar-like pronotum (Hamilton 1990); (c)

placate sensilla on the pedicel (Baker & Chandrapatya 1993); (d) a specialized

sensory organ on the base of flagellomere 1 (Bourgoin 1985); (e) a rolled, but

never spiral, folded wing-coupling system (D’Urso & Ippolito 1994: type B), with

(f) strong accessory microsculpture. Clade Heteropterodea (= Coleorhyn-

cha+ Heteroptera)—(a) a gula, or the beginning of its ventral fusion in Coleo-

rhyncha (Hamilton 1981: fig. 23); (b) a distinctive triangular mandibular lever

(Hamilton 1981; K. G. A. Hamilton, personal communication); (c) a non-aristoid

reduction of antennae to 3 or 4 (secondarily 5) segments (Schlee 1969, Emel’yanov

1987, Wheeler et al. 1993); (d) capture of the trachea of forewing vein 1A by

CuA2 and its invasion of the remigium (Wootton 1965, 1986), despite Wootton’s

(1979) summary of unreliability of tracheal capture for vein homology; (e) wings

capable of being folded flat (overlapping) over the body (Wheeler et al. 1993); (f)

ground plan for the abdominal segments (Schlee 1969); (g) structure of the anal

cone (Schlee 1969); and (h) development of the sclerites at the base of the aedeagus

(Schlee 1969); see Wheeler et al. (1993) for 18S rDNA synapomorphies. Clade

Coleorhyncha—see Popov & Shcherbakov (1991: 233) for synapomorphies. Clade

Heteroptera—see Wheeler et al. (1993) or Hennig (1981) for numerous synapo-

morphies.

PALEONTOLOGICAL EVIDENCE

Under the section on cladistic implications, we considered some fossil evidence

for character homoplasy. Here, we estimate the concordance of fossil lineages

with the 18S rDNA topology.

Prior to 1980, the relationships among hemipteran fossils were confused by

differing philosophical camps that often made interpretations despite a lack of

important character information. In a review article, Wootton (1981: 331-332)

states: ““Within the Permian, Hemiptera radiated spectacularly, leaving behind a

bewildering array of fossils, many of them just wings. Convergence is widespread,

and interpretation difficult and conflicting. ... Auchenorrhyncha occur in pro-

fusion and confusion in the L. and U. Permian...”; he cites as an example:

*““Prosbolidae may be primitive Cicadoidea, and Scytinopteridae may be Cerco-

poidea, but both these families have been conflictingly defined”’ (e.g., Evans 1956,

1964, vs Rohdendorf et al. 1961, Bekker-Migdisova in Rohdendorf 1962). Hennig

(1981: 273) aptly summarizes: “The differences of opinion do not inspire me with

much confidence in the decisions of specialists who have assigned the Permian

and other fossils to various subgroups of *Auchenorrhyncha.” There was no

apparent rigorous cladistic methodology for assignments and homology. Usually,

50 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

P, =P, T J K Cz R

Dysmorphoptili

Prosbolopseidae

he)

- Palaeontinoidea

. = Pereborioidea

=———= Ignotalidae |

presumed archescytinoidean ancestor

Pe)

Figure 4. Paleontological synopsis of (selected) taxa from euhemipteran lineages, largely following

Shcherbakov (1984, 1988) and Popov & Shcherbakov (1988, 1991) with insertion of Fulgoridiidae

(Bode 1953, Hamilton 1990). Abbreviations for geological times are standard; gray boxes demarcate

clade lineages; question marks (small and large) signify derivations implied as tenative in the literature.

Shcherbakov (1984) places Prosbolopseidae and Ingruidae in superfamily Prosboloidea.

early hemipterans were known only from forewing tegmina; body and head im-

pressions, sometimes distorted, usually are unknown until the Jurassic (e.g., Bode

1953, Hamilton 1990). However, despite this reliance on tegmina, only Evans

(1964) attempted to define early fossil auchenorrhynchan superfamilies by wing

venation.

More recently, Shcherbakov (1981, 1982), following Emel’yanov (1977), and

Dworakowska (1988), following Kukalova-Peck (1983), have treated the diag-

nostic venation of extant auchenorrhynchan families. Since then, reassessments

of older phylogenetic relationships, based on group diagnostics but not necessarily

apomorphies, have been made for both the ancestral hemipteroid lineage (e.g.,

Kukalova-Peck & Brauckmann 1992), and for earlier hemipterans themselves

(e.g., Shcherbakov 1984, 1988; Popov & Shcherbakov 1988, 1991). Figure 4 shows

a current paleontological synopsis of euhemipteran lines.

These assessments suggest that the extant (monophyletic) Cicadomorpha (Cly-

peata sensu Shcherbakov) and Heteropterodea were derived from the Permian

Prosboloidea [P,-J,] of the polyphyletic Cicadomorpha. Reputedly, the modern

Cicadomorpha lineage evolved from an ancestor in the prosboloidean Prosbolidae

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA 51

[P,], along a lineage that involved Hylicellidae [T,-K2] giving rise to: (a) Cica-

doidea [T;-R]; (b) Procercopidae [J,-K,], that begot Cercopoidea [J,-R]; and (c)

Karajassidae [J,_,], that begot Cicadelloidea [J,-R] (Shcherbakov 1988). A second

prosbolid lineage begot Dysmorphoptilidae [P,] (Shcherbakov 1984). A prosbol-

opseid Prosbolopseinae lineage gave rise to: (a) Palaeontinoidea [P,-K,]; and (b)

Pereborioidea [P,-T;], the latter probably deriving Ignotalidae [P,] (Shcherbakov

1984, 1988).

The Heteropterodea reputedly arose from a prosbolopseid Ingruidae [P,] an-

cestor. The Coleorhyncha lineage began when ingruids begot Progonocimicinae

[P,-K,], that begot Cicadocorinae [J,-K,] and Karabasiinae [J,-K,], the latter of

which begot Hoploridiinae [K,] and Peloridiidae [K?-R] (Popov & Shcherbakov

1991). The Heteroptera lineage reputedly arose when ingruids begot the Scytin-

opteroidae, the most primitive of which, Scytinopteridae [P,-T,], begot: (a) the

Serpenivenidae [P,-T,] and their probable descendents, Stenoviciidae [P,-T,] and

Paraknightiidae [T,]; and later, (b) Ipsviciidae [J,_,] (Shcherbakov 1984).

The origin of the Fulgoromorpha lineage is yet unclear. However, Shcherbakov

(1984) suggests it arose from Archescytinoidea, independently of Cicadomorpha,

towards the end of the early Permian. Assessment of Fulgoromorpha’s Permian

ancestors, reputedly Surijokocixidae [P,-J,] and Coleoscytoidea [P,_,], is more

tentative, and modern fulgoromorphan groups, such as Cixidae [K,-R] and Achil-

idae [K,-R], do not appear until the Cretaceous (Shcherbakov 1988), after the

intervening presence of Fulgoridiidae [J] in the Jurassic (Bode 1953, Hamilton

1990, but see Wilson et al. 1994).

Thus, the paleontologically supported origin of Fulgoromorpha remains the

most unsettling of the three major euhemipteran clades. Tracing early fulgoro-

morphans before the Jurassic is problematic because only tegmina occur then,

but most fulgoromorphan apomorphies are head characters. The paleontological

evidence, therefore, does not support clade Auchenorrhyncha, because of the

polyphyletic nature of Cicadomorpha (sensu Shcherbakov). Presently, it would

seem to most closely support the slightly less parsimonious 18S rDNA topologies

that indicate clade modern Cicadomorpha-+ Heteroptera. In our opinion, however,

the nucleotide-based topology is superior to very nebulous indications of. origin

for Fulgoromorpha that are revealed by fossils.

ECO-EVOLUTION OF HEMIPTERANS AND CLADOGENESIS

What selective driving forces were responsible for the major cladogenic events

that shaped the 18S rDNA topology of hemipterans? We believe that the diver-

gence, establishment and success of major evolutionary lineages (as clades) re-

quires the presence, recognition and exploitation of existing niches. At best, re-

strictive niches should hamper the evolutionary diversification of their exploitive

tive lineages. New niches (“‘neoniches’’) that develop after the establishment of

previously existing and non-competing clades, would permit multiple entry points

for would-be competing invaders from multiple, existing clades. Such homopla-

sious invasions of any neoniche should require its delineation among the poly-

phyletic neocompetitors if all are to survive as neo(sub)clades of their respective

parental clades. In time, each neoclade should genetically and morphologically

differentiate from its parental clade, both before (cladogenic) and during (anagenic)

its radiation in the neoniche.

2 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Resultant convergence among the polyphyletic neoniche invaders would be

dictated by niche-required morphology and function or other underlaying bio-

logical constraints (Wake & Larson 1987, Wake 1991). Neoniche radiations should

be recognizable by the phylogenetic distributions of their taxa among parental

clades whose basal section taxa have differing niche habitations. The eco-evolu-

tionary constraints on such a scenario are the relative chronological developments

of niches versus clades, the existence and degree of preadaptation or adaptive

constraints (Moran 1988, 1990; Wake 1991), and potentially overlaying and in-

hibiting biogeographic demarcations.

The Hemiptera illustrate these tenets, and their early cladogenesis overlays the

evolution of vascularization within plants. Clade Sternorrhyncha has intercellular

stylet-penetration of plants. Its most basal group, Psyllidae (Campbell et al. 1994),

ingest from a variety of vascular and non-vascular tissues (Ullman & McLean

1988a, b). The Psyllidae’s more derived sister-clade, Aleyrodiformes (Aphidoi-

dea+Coccoidea+Aleyrodoidea, sensu Campbell et al. 1994), ingests predomi-

nantly when their stylet tips are within phloem sieve elements (Backus 1988,

Janssen et al. 1989). In contrast, within clade Euhemiptera, Cicadomorpha and

Fulgoromorpha have intracellular stylet-penetration, with less precision than ster-

norrhynchans (Backus 1988). Clade Cicadomorpha initially evolved to feed on

xylem (Cercopidae, Cicadidae, Cicadellidae: Cidadellinae), but has radiated to

phloem (Membracidae, Cicadellidae: Deltocephalinae) and parenchyma (Cica-

dellidae: Typhlocybinae) as neoniches, presumably after the development of these

plant tissues.

Both Sternorrhyncha and Cicadomorpha have developed varying types of filter

chambers that are presumably used for osmoregulating profuse amounts of in-

gested hypotonic plant fluids. The sternorrhynchan filter (Evans 1963: type A) is

simple and anteriorly expanded; the cicadomorphan filter (Evans 1963: type B)

is complex and posteriorly expanded in association with the Malpighian tubules

(Pesson 1944, Goodchild 1966). The relatively simple sternorrhychan filter was

evolved by the appearance of the psyllids, who feed on various tissues, and was

retained (probably parsimoniously) in their phloem feeding sister-clade Aleyro-

diformes. Interestingly, psyllids are the only Sternorrhyncha that have retained

all four Malpighian tubules; Aleyrodiformes have a reduced number or none (some

aphids). The complexity of the cicadomorphan filter was probably required for

xylem feeding, because that food source is very dilute; it also was retained (again,

probably parsimoniously) among the cicadomorphan neoniche invaders (i.e.,

Membracidae, Cicadellidae: Deltocephalinae). All euhemipterans have retained

all four Malpighian tubules.

It seems probable that early fulgoromorphans initially evolved to feed on roots

and fungal hyphae, which exist in subterranean/semisubterranean (duff) niches,

much as many of their immatures do now (Wilson et al. 1994). This selection

probably happened because Sternorrhyncha and early Cicadomorpha (i.e., Ci-

cadidae, Cercopidae, Cicadellidae: Cicadellinae), respectively, already had occu-

pied intracellular and intercellular/xylem feeding niches, (before their secondary

radiations onto later neoniches). Later, with the advent of phloem, fulgoromor-

phans probably moved readily into that neoniche and radiated. The Fulgoro-

morpha, lacking the filter chambers of the coexisting clades, probably found fine

roots and fungal hyphae had relatively nutritious cells that are easily attacked; as

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA a5

a food source, both these and phloem are less dilute than liquids from xylem. As

a result, fulgoromorphans did not require development of an extraordinarily en-

larged feeding pump, the associated, enlarged clypeal housing, or the specialized

gut filter, that cicadomorphans did to handle the increased fluid load necessary

for xylem feeding.

As cladogenesis of hemipterans progressed, and earlier (“homopteran’’) clades

dominated both intra- and intercellular niches in developing vascular plant sys-

tems, their roots and soil fungi, the coleorhynchans appeared; their surviving

relictual group, peloridiids, ended up on mosses, a nonvascular plant resource

that was unoccupied by the other hemipteran clades. Probably because that niche

is not diverse, coleorhynchans did not flourish, expand and radiate as did the

sternorrhynchans, cicadomorphans and fulgoromorphans. However, their begin-

ning gular development (Hamilton 1981: fig. 23), or rather that of their immediate

ancestor in common with the heteropterans, began a change toward a prognathous

rostrum and its liberating evolutionary potential.

In contrast to the Coleorhyncha, their sister-group, Heteroptera, evolved an

alternative strategy, predation, which required the major and radical morpholog-

ical shifts witnessed in the Enicocephalomorpha (Grimaldi et al. 1993). Once the

predatory phena was accomplished, however, it opened vast new niches and

environs for suctoral feeding on animalian body fluids, which up until then only

mandibulate predators exploited. Predation as primary feeding strategy was ex-

ploited by most early heteropteran clades (Carver et al. 1991), and cladogenic

radiation (Fig. 5) occurred in both terrestrial (Enicocephalomorpha, Dipsocoro-

morpha, Cimicomorpha), and aquatic/semiaquatic environs (Gerromorpha, Neo-

morpha, Leptopodomorpha). Although some groups among the more derived

heteropteran clades reverted secondarily to phytophagy, they feed on parenchyma,

seeds and pollen (Carver et al. 1991), which are largely unexploited by the “‘ho-

mopteran” clades. Only the Pentatomomorpha, a terminal heteropteran clade

(Fig. 5) shows a major reversion to phytophagy. Wheeler et al. (1993) discuss the

phylogenetic topology of Heteroptera, and Carver et al. (1991) discuss their bi-

ology.

Our 18S rDNA findings, in conjunction with other evidence for placement of

the Coleorhyncha (Schlee 1969: 23, Popov & Shcherbakov 1991: 233, Wheeler

et al. 1993: 131-132), suggests the preceding order of hemipteran cladogenesis.

Available evidence indicates the major clades diverged by the late Permian, and

scant synapomorphies linking these clades suggest rapid divergence of morpho-

logical form occurred. Frequent homoplasy within these developing clades prob-

ably resulted from evolutionary constraints (Wake & Larson 1987, Wake 1991).

In conjunction with a relatively steady speed base substitution clock, the relatively

few 18S rDNA synapomorphies shown among these clades (Fig. 3) functionally

also indicates a relatively short time was involved during the cladogenesis. A

similar 28S rRNA topology has been found for sponges, with deep radiations

among clades that are separated by short internodes (Lafay et al. 1992).

Thus, a saltational and punctuated equilibrium mode of evolution appears to

be involved among the basal hemipteran clades, and we suspect this may have

resulted from sudden and dramatic selection pressures during the Permian, prob-

ably following one or more catastrophic truncation events. Ecomorphotypic di-

versity among every existing major group of vascular plants declined dramatically

54 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

STERNORRHYNCHA

CLYPEORRHYNCHA

' ARCHAEORRHYNCHA

Peloridiomorpha

Enicocephalomorpha

elaldiueung |

BIB|CIWBYOON

| u

Be)

6 te-]m Dipsocoromorpha

XD Ss | = Gerromorpha

t co oO : | 7 Nepomorpha

z = S ® 3 — Leptopodomorpha

= s 3 = Cimicomorpha

o Pentatomomorpha

je) s+

Figure 5. Phylogenetic relationships of proposed hemipteran suborders and existing (heteropter-

odean) infraorders. Horizontals depict the three basal clade suborders (capitals; Clypeorrhyncha =

extant Cicadomorpha; Archaeorrhyncha = Fulgoromorpha) and the more derived infraorders (lower

case; Peloridiomorpha = Coleorhyncha). Verticals depict the clade names (sensu Schuh 1979; lower

case) and the proposed suborder Prosorrhyncha (capitals; = Heteropterodea, sensu Schuh 1979, as

Coleorrhyncha-+ Heteroptera).

during the Permian (Shear 1991: 288), but rose again among the new angiosperms

during the Cretaceous. This temporally changing aspect of plant diversity parallels

the initiation of the major hemipteran clades (Permian) and their later internal

radiation (late Jurassic/Cretaceous) into modern groups.

IMPLICATIONS FOR SUBORDER NOMENCLATURE

If our 18S rDNA-based topologies are correct, the paraphyly of Auchenor-

rhyncha requires its abandonment as a cladistic subordinal taxon of Hemiptera.

Instead, recognition of four major hemipteran clades (sternorrhynchans, extant

cicadomorphans, fulgoromorphans, heteropterodeans) as suborders is clearly ap-

propriate (Fig. 5). We rely on the 18S rDNA synapomorphies of Wheeler et al.

(1993), the morphological synapomorphies of Schlee (1969), and the fossil lineage

assessment of Popov & Shcherbakov (1991: 233, as Coleorhyncha <— Ingruidae

— Scytinopteroidea — Heteroptera) for placement of the Coleorhyncha as sister-

clade to Heteroptera®. Despite anyone’s lingering uncertainty concerning the rel-

ative phylogenetic topology among the major clades, there can be no doubt of

their individual monophyly. Thus, demarcation of Hemiptera into these four

major clades, as suborders, is a conservative treatment that preserves their mor-

phological and ecological delineation.

Three new suborder names are proposed here, however, because several po-

tential obfuscations confuse the application of the currently available names. First,

there is a polyphyletic, paleontological use (e.g., Shcherbakov 1984) of Cicado-

morpha, that differs from the one that is monophyletic covering extant taxa only

(e.g., Carver et al. 1991). Second, there are varying definitions of Heteroptera,

which may (e.g., Carver et al. 1991) or may not (e.g., Henry & Froeschner 1988)

8 Recent molecular evidence based on 18S rDNA sequences (ex Wheeler et al. 1993) shows resolute

synapomorphic sites supporting Coleorhyncha+ Heteroptera monophylly; to be discussed elsewhere

(Campbell et al., unpublished data).

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA 55

include Coleorhyncha, versus Heteropterodea (e.g., Schuh 1979, Wheeler et al.

1993) or its alternative, initial spelling Heteropteroidea (Schlee 1969). Third, a

problem exists regarding the implied relative hierarchical status of Cicadomorpha

and Fulgoromorpha in contrast to heteropteran infraorders, which also end in

suffix ““-morpha”’ (e.g., Schuh 1979).

Standardizing on suffix ‘“‘-rrhyncha” to denote suborder, we retain Sternor-

rhyncha, and propose as hemipteran suborders: (a) Clypeorrhyncha [Gr. “‘shield-

nose’’], for the monophyletic extant cicadomorphan taxa, (b) Archaeorrhyncha

[Gr. “‘ancient-nose’’], for Fulgoromorpha, and (c) Prosorrhyncha [Gr. “‘front-”’ or

“‘forward-nose’’|, for clade Coleorhyncha+ Heteroptera. We believe these names

provide a much needed alleviation of confusion over the boundaries, hierarchical

status and monophyly of these groups. Their application toward that end is feasible

because the ICZN code does not require priority-basis recognition of subordinal

names. In view of our 18S rDNA findings, the clade name Neohemiptera is also

proposed for the clade Fulgoromorpha+ Heteropterodea in Schuh’s (1979) system

(our clade Archaeorrhyncha+ Prosorrhyncha).

It is appropriate, under this system, to refer to Coleorhyncha as Peloridiomor-

pha, indicating its infraordinal level within suborder Prosorrhyncha. Continued

use of Coleorhyncha would imply its subordinal status, and necessarily that of

Heteroptera, negating Prosorrhyncha. In contrast, use of Heteroptera can imply

a greater clade division within suborder Prosorrhyncha (i.e., Hemiptera: Prosor-

rhyncha: Heteroptera), as the sister-group to Peloridiomorpha. Continued use of

Fulgoromorpha and Cicadomorpha, however, would confuse their infra- and

subordinal status. Moreover, use of Cicadomorpha confuses its paleontological

versus extant taxonomic meaning; therefore, any such use should be in a non-

cladistic fashion only, to indicate the extinct, polyphyletic Mesozoic taxa that

may be relatives of the modern, monophyletic Clypeorrhyncha, but that lack all

the latter’s defining synapomorphies.

The logic for continuation of “-morpha”’ suffixed infraorders, and proposed

adoption of “‘-rrhyncha”’ suffixed suborders, for Hemiptera is independent of, but

related to, another question that should be asked. Because Hemiptera is mono-

phyletic, and heteropterists generally use Heteroptera for “their group,” perhaps

it is time to recognize and address a common, often expressed resentment by

many “homopterists,”’ for whatever rationale, towards incorporation of those

groups under the name Hemiptera. Unfortunately, Fabricius’ (1775) neutral or-

dinal name, Ryngota, later modified to Rhyngota (Fabricius 1803) and then Rhyn-

chota (Burmeister 1835), the latter championed by Hamilton (1981, 1983) and

others (e.g., Dworakowska 1988), has been largely ignored for hemipterans. Adop-

tion of Rhynchota may be appropriate as an admittedly political, but pragmatic,

attempt at appeasing and unifying all “hemipterists” under one ordinal banner.

ACKNOWLEDGMENT

We thank E. Backus (University of Missouri), R. Denno (University of Mary-

land), R. Foottit and K. G. A. Hamilton (Canada Agriculture, Ottawa), D. Hendrix

(USDA, Phoenix, Arizona), C. Schaefer (University of Connecticut), D. Schlee

(Staatliches Museum fiir Naturkunde, Stuttgart, Germany), and T. Wood (Uni-

versity of Delaware) for helpful discussions. Robert V. Dowell served as editor,

and three anonymous reviewers’ comments were addressed and incorporated.

56 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

LITERATURE CITED

Backus, E. A. 1988. Sensory systems and behaviors which mediate hemipteran plant-feeding: a

taxonomic overview. J. Insect Physiol., 34: 151-165.

Baker, G. T. & A. Chandrapatya. 1993. Antennal sensilla of the Chinese lantern fly, Pyrops candelaria

L. (Homoptera: Fulgoridae). Proc. Entomol. Soc. Wash., 95: 245-252.

Betts, C. R. 1986a. The comparative morphology of the wings and axillae of selected Heteroptera.

J. Zool., London, Ser. B, 1 (Part 2): 255-282.

Betts, C. R. 1986b. Functioning of the wings and axillary sclerites of Heteroptera in flight. J. Zool.,

London, Ser. B, 1 (Part 2): 283-302.

Betts, C. R. 1986c. The kinematics of Heteroptera in free flight. J. Zool., London, Ser. B, 1 (Part

2): 303-316.

Bode, A. 1953. Die Insektenfauna des ostniedersachsischen Oberen Lias. Palaeontographica, Stuttgart

(A): 103: 1-375.

Borror, D. J. & D.M. DeLong. 1971. An introduction to the study of insects (3rd ed.). Holt, Rinehart

and Winston, New York.

Bourgoin, T. 1985. Morphologie antennaire des Tettigometridae (Hemiptera: Fulgoromorpha). Nou-

velle Rev. Entomol. (N.S.), 2: 11-20.

Bremer, K. 1990. Combinable component consensus. Cladistics, 6: 369-372.

Burmeister, H. C. C. 1835. Handbuch der Entomologie. Tome 2. T. Enslin, Berlin.

Campbell, B. C., J. D. Steffen-Campbell & R. J. Gill. 1994. Evolutionary origin of whiteflies (He-

miptera: Sternorrhyncha: Aleyrodidae) inferred from 18S rDNA sequences. Insect Molecular

Biology, 3: 73-88.

Carmean, D., L. S. Kimsey & M. L. Berbee. 1992. 18S rDNA sequences and the holometabolus

insects. Molecular Phylogenetics and Evolution, 1:270-278.

Carver, M., G. F. Ross & T. E. Woodward. 1991. Hemiptera (bugs, leafhoppers, cicada, aphids,

scale insects etc.). pp. 429-509. Jn Naumann, I. D., P. B. Crane, J. F. Lawrence, E. S. Neilsen,

J. P. Spradbery, R. W. Taylor, M. J. Whitten & M. J. Littlejohn (eds.). The insects of Australia,

a textbook for students and research workers. Vol. 1 (2nd ed.). Melbourne University Press,

Melbourne, Australia.

Chapman, R. F. 1971. The insects: structure and function (2nd ed.). American Elsevier Publishing

Company, Inc., New York.

Christen, R., A. Ratto, A. Baroin, R. Perasso, K. G. Grell & A. Adoutte. 1991. An analysis of the

origin of metazoans, using comparisons of partial sequences of the 28S RNA, reveals an early

emergence of triploblasts. EMBO Journal, 10: 499-503.

Claridge, M. F. 1985. Acoustic behavior of leafhoppers and planthoppers: species problems and

speciation. pp. 103-125. Jn Nault, L. R. & J. G. Rodriguez (eds.). The leafhoppers and plant-

hoppers. John Wiley & Sons, New York.

Claridge, M. F. & P. W. F. de Vrijer. 1994. Reproductive behavior: the role of acoustic signals in

species recognition and speciation. pp. 216-233. Jn Denno, R. F. & T. J. Perfect (eds.). Plant-

hoppers, their ecology and management. Chapman & Hall, New York.

Cobben, R.H. 1968. Evolutionary trends in Heteroptera. Part I. Eggs, architecture of the shell, gross

embryology, and eclosion. Mededeling 151, Laboratory of Entomology, Agricultural University,

Wageningen. Centre for Agricultural Publishing and Documentation, Wageningen, Netherlands.

Cobben, R. H. 1978. Evolutionary trends in Heteroptera. Part IJ. Mouthpart structures and feeding

strategies. Mededelingen Landbouwhogeschool Wageningen 78 (5). J. Veenman & Zonen B.V.,

Wageningen, Netherlands.

Crespi, B. J. 1992. Natural selection and morphometrics. pp. 55-64. In Sorensen, J. T. & R. Foottit

(eds.). Ordination in the study of morphology, evolution and systematics of insects: applications

and quantitative genetic rationals. Elsevier Science Publishers, Amsterdam, the Netherlands.

Cwikla, P. S. & P. H. Freytag. 1983. External morphology of Xestocephalus subtessellatus (Homop-

tera: Cicadellidae: Xestocephalinae). Ann. Entomol. Soc. Am., 76: 641-651.

D’Urso, V. & S. Ippolito. 1994. Wing-coupling apparatus of Auchenorrhyncha (Insecta: Homoptera).

Int. J. Insect Morphol. & Embryol., 23: 211-224.

Dworakowska, I. 1988. Main veins of the wings of Auchenorrhyncha (Insecta: Rhynchota: Heme-

lytrata). Entomol. Abh. (Dresden), 52: 63-108.

Emel’yanov, A. F. 1977. Homology of wing structures in the Auchenorrhyncha and Polyneoptera.

Trudy vses. Entomol. Obshch., 58: 3-48. (in Russian)

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA ae

Emel’yanov, A. F. 1987. The phylogeny of the Cicadina (Homoptera, Cicadina) based on comparative

morphological data. Trudy vses. Entomol. Obshch., 69: 19-109. (in Russian)

Evans, J. W. 1946. A natural classification of leaf-hoppers (Homoptera, Jassoidea). Part 2: Aeta-

lionidae, Hylicidae, Eurymelidae. Trans. Royal Entomol. Soc., 97: 39-54.

Evans, J. W. 1956. Palaeozoic and Mesozoic Hemiptera (Insecta). Aust. J. Zool., 4: 165-258.

Evans, J. W. 1963. The phylogeny of the Homoptera. Ann. Rev. Entomol., 8: 77-94.

Evans, J. W. 1964. The periods of origin and diversification of the superfamilies of the Homoptera-

Auchenorhyncha [sic] (Insecta) as determined by a study of the wings of Palaeozoic and Mesozoic

fossils. Proc. Linn. Soc., London, 175: 171-181.

Fabricius, J. C. 1775. Systema entomologiae, sistens insectorum classes, ordines, genera, species,

adjectis synonymis, locis, descriptionibus, observationibus. Flensburgi et Lipsiae, Korte.

Fabricius, J.C. 1803. Systema Rhyngotorum secundum ordines, genera, species adjectis synonymis,

locis, descriptionibus, observationibus. Carolum Reichard, Brunsvigae.

Felsenstein, J. 1982. Numerical methods for inferring evolutionary trees. Quart. Rev. Biol., 57: 379-

404.

Felsenstein, J. 1983. Methods for inferring phylogenies: a statistical view. Jn Felsenstein, J. (ed.).

NATO Advanced Study Institute, Series G, Number 1. Springer-Verlag, New York.

Felsenstein, J. 1988. Phylogenies and quantitative characters. Ann. Rev. Ecol. & Syst., 19: 445-471.

Felsenstein, J. & H. Kishino. 1993. Is there something wrong with the bootstrap on phylogenies? A

reply to Hillis and Bull. Syst. Biol., 42: 193-200.

Goodchild, A. J. P. 1966. Evolution of the alimentary canal in the Hemiptera. Biol. Rev., 41: 97-

140.

Grimaldi, D., C. Michalski & K. Schmidt. 1993. Amber fossil Enicocephalidae (Heteroptera) from

the Lower Cretaceous of Lebanon and Oligo-Miocene of the Dominican Republic, with bio-

geographic analysis of Enicocephalus. Amer. Mus. Novitates, 3071.

Hamilton, K. G. A. 1981. Morphology and evolution of the rhynchotan head (Insecta: Hemiptera,

Homoptera). Canad. Entomol., 113: 953-974.

Hamilton, K. G. A. 1983. Classification, morphology and phylogeny of the family Cicadellidae

(Rhynchota: Homoptera). pp. 15-37. In Knight, W. J., N. C. Pant, T. S. Robertson & M. R.

Wilson (eds.). 1st international workshop on leafhoppers and planthoppers of economic im-

portance. Commonwealth Institute of Entomology, London.

Hamilton, K.G. A. 1990. Chapter 6. Homoptera. pp. 82-122. Jn Grimaldi, D. A. Insects from the

Santana formation, Lower Cretaceous, of Brazil. Bull. Am. Mus. Nat. Hist., 195.

Hendriks, L. H., R. De Baere, C. Van Broeckhoven & R. De Wachter. 1988. Primary and secondary

structure of the 18S ribosomal RNA of the insect species Tenebrio molitor. FEBS Lett., 232:

115-120.

Hennig, W. 1981. Insect phylogeny. [Pont, A. C., ed. and translator]. John Wiley & Sons, New York. |

Henry, T. J. & R. C. Froeschner. 1988. Introduction. pp. ix—xix. Jn Henry, T. J. & R. C. Froeschner

(eds.). Catalog of the Heteroptera, or true bugs, of Canada and the continental United States.

E. J. Brill, Leiden, the Netherlands.

Hillis, D. M. & J.J. Bull. 1993. Anempirical test of bootstrapping as a method for assessing confidence

in phylogenetic analysis. Syst. Biol., 42: 182-192.

Hillis, D. M., J. P. Huelsenbeck & C. W. Cunningham. 1994. Application and accuracy of molecular

phylogenies. Science, 264 (29 Apr 1994): 671-677.

Howse, P. E. & M. F. Claridge. 1970. The fine structure of Johnston’s organ of the leafhopper,

Oncopsis flavicollis. J. Insect Physiol., 16: 1665-1675.

Imms, A.D. 1940. On the antennal musculature in insects and other arthropods. Quar. J. Microscop.

Sci., 81: 273-320.

Janssen, J. A. M., W. F. Tjallingii & J. C. McLean. 1989. Electrical recording and ultrastructure of

stylet penetration by the greenhouse whitefly. Entomol. Exp. Appl., 52: 69-81.

Kristensen, N. P. 1975. The phylogeny of hexapod “orders.” A critical review of recent accounts.

Z. Zool. Syst. Evolforsch, 13: 1-44.

Kristensen, N. P. 1991. Phylogeny of extant hexapods. pp. 125-140. Jn Naumann, I. D., P. B. Crane,

J. F. Lawrence, E. S. Neilsen, J. P. Spradbery, R. W. Taylor, M. J. Whitten & M. J. Littlejohn

(eds.). The insects of Australia, a textbook for students and research workers. Vol. 1 (2nd ed.).

Melbourne University Press, Melbourne, Australia.

Kukalova-Peck, J. 1983. Origin of the insect wing and wing articulation from the arthropodan leg.

J. Canad. Zool., 61: 1618-1668.

58 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Kukalova-Peck, J. 1991. Fossil history and the evolution of hexapod structures. pp. 141-179. In

Naumann, I. D., P. B. Crane, J. F. Lawrence, E. S. Neilsen, J. P. Spradbery, R. W. Taylor, M.

J. Whitten & M. J. Littlejohn (eds.). The insects of Australia, a textbook for students and

research workers. Vol. 1 (2nd ed.). Melbourne University Press, Melbourne, Australia.

Kukalova-Peck, J. & C. Brauckmann. 1992. Most paleozoic Protorthoptera are ancestral hemipter-

oids: major wing braces as clues to a new phylogeny of Neoptera (Insecta). Can. J. Zool., 70:

2452-2473.

Kwon, O. K. Ogino & H. Ishikawa. 1991. The longest 18S ribosomal RNA ever known: nucleotide

sequence and presumed secondary structure of the 18S rRNA of the pea aphid, Acyrthosiphon

pisum. Eur. J. Biochem., 202: 827-833.

Lafay, B., N. Boury-Esnault, J. Vacelet & R. Christen. 1992. An analysis of partial 28S ribosomal

RNA sequences suggests early radiations of sponges. BioSystems, 28 (1992): 139-151.

Lanyon, S. M. 1993. Phylogenetic frameworks: towards a firmer foundation for the comparative

approach. Biol. J. Linnean Soc., London, 49: 45-61.

Latreille,P. A. 1910. Ordre III: Hemipteres. Hemiptera. Section Premiere. Heteropteres. Heteroptera.

pp. 250-251, 254-261, 421, 433-434. Jn Considerations generales sur l’ordre naturel des an-

imaux composant les classes des crustaces, des arachnides, et des insectes, avec un tableau

methodique de leurs genres, disposes en families. Schoell, Paris.

Lewontin, R. C. 1989. Inferring the number of evolutionary events from DNA coding sequence

differences. Molecular Biol. Evol., 6: 15-32.

Linnaeus, C. 1758. Systema naturae per regna tria naturae, secundum classes, ordines, genera, species,

cum characteribus, differentiis, synonymis, locis. Editio decima, reformata. Laurentii Salvii,

Homiae.

Maddison, W. P. & D.R. Maddison. 1992. MacClade, analysis of phylogeny and character evolution,

version 3.01. Sinauer Associates, Inc., Sunderland, Massachusetts.

Marshall, A. T. & C. T. Lewis. 1971. Structured variation in the antennal sense organs of fulgoroid

Homoptera (Insecta). Zool. J. Linn. Soc., 50: 181-184.

Michevich, M. F. & S. J. Weller. 1990. Evolutionary character analysis: tracing character change on

a cladogram. Cladistics, 6: 137-176.

Mishler, B. D., K. Bremer, C. J. Humpheries & S. P. Churchill. 1988. The use of nucleic acid

sequence data in phylogenetic reconstruction. Taxon, 37: 391-395.

Mitomi, M., T. Ichikawa & H. Okamoto. 1984. Morphology of the vibration-producing organ in

adult rice brown planthopper, Nilaparvata lugens (Stal) (Homoptera: Delphacidae). Appl. En-

tomol. Zool., 19: 407-417.

Moran, N. A. 1988. The evolution of host alternation in aphids: evidence that specialization is a

dead end. Amer. Nat., 132: 681-706.

Moran, N. A. 1990. Aphid life cycles: two evolutionary steps. Amer. Nat., 136: 135-138.

Ossiannilsson, F. 1949. Insect drummers, a study on the morphology and function of the sound-

producing organ of Swedish Homoptera Auchenorrhyncha with notes on their sound production.

Opusc. Entomol., Suppl. 10.

Patterson, C. 1989. Phylogenetic relationships of major groups: conclusions and prospects. pp. 471-

488. In Fernholm, B., K. Bremer & H. Jornvall (eds.). The hierarchy of life. Elsevier, Amster-

dam.

Pesson, P. 1944. Contribution a l’etude morphologique et fonctionnelle de la tete de l’appareil buccal

et du tube digestif des femelles de Coccides. Monogr. Stat. Lab. Recherche Agron., Paris.

Popov. Y. 1981. Historical development and some questions on the general classification of the

Hemiptera. Rostria (suppl), 33: 85-99.

Popov, Y. & D. Shcherbakov. 1988. Origin of Coleorrhyncha based upon fossil evidence. Abstracts

and authors index. p. 8. Proc. XVIII Int. Cong. Entomol., 1988, Vancouver, BC, Canada.

(abstract)

Popov, Y. & D. Shcherbakov. 1991. Mesozoic Peloridiodea and their ancestors (Insecta: Hemiptera,

Coleorrhyncha). Geologica et Palaeontologica, 25: 215-235.

Pringle, J. W. S. 1954. A physiological analysis of cicada song. J. Exp. Biol., 31: 525-556.

Pringle, J. W. S. 1957. The structure and evolution of the organs of sound production in cicadas.

Proc. Linn. Soc., London, 167: 144-159.

Rohdendorf, B. B. (ed.). 1962. Foundations of palaeontology: handbook for palaeontology and

geology of the USSR. Acad. Sci. USSR, Moscow.

1995 SORENSEN ET AL.: NON-MONOPHYLY OF AUCHENORRHYNCHA ao

Rohdendorf, B. B., E. E. Bekker-Migdisova, O. M. Sharov & A. G. Sharov. 1961. Palaeozoic insects

of the Kuznetsk Basin. Tr. Paleontol. Instit. Akad. Nauk SSSR, 85.

Schlee, D. 1969. Morphologie und symbiose; ihre beweiskraft fur die verwandtschaftsbezichungen

der Coleorrhyncha (Insecta, Hemiptera). Phylogenetische studien an. Hemiptera IV: Heterop-

teroidea (Heteroptera + Coleorrhyncha) als monophyletische gruppe. Stuttgarter Beitrage zur

Naturkunde, Staatlichen Museum fur Naturkunde. Stuttgart, 210: 1-27.

Schneider, D. 1964. Insect antennae. Ann. Rev. Entomol., 9: 103-122.

Schuh, R. T. 1979. [Review of] Evolutionary trends in Heteroptera. Part I]. Mouthpart-structures

and feeding strategies, by R. H. Cobben. Syst. Zool., 28: 653-656.

Schuh, R. T. 1986. The influence of cladistics on heteropteran classification. Ann. Rev. Entomol.,

31: 67-93.

Shaw, K. C. & O. V. Carlson. 1979. Morphology of the tymbal organ of the potato leafhopper,

Empoasca fabae Harris (Homoptera: Cicadellidae). J. Kansas Entomol. Soc., 52: 701-711.

Shcherbakov, D. 1981. Diagnostics of the families of the Auchenorrhyncha (Homoptera) on the

basis of wings. I. Fore wings. Entomol. Review, 60 (4): 64-81.

Shcherbakov, D. 1982. Diagnostics of the families of the Auchenorrhyncha (Homoptera) on the

basis of wings. II. Hind wings. Entomol. Review, 61 (3): 70-78.

Shcherbakov, D. 1984. Systematics and phylogeny of Permian Cicadomorpha (Cimicida and Ci-

cadina). Paleont. Zh. 1984 (2): 89-101 [in Russian, translation in Paleont. J., 1984 (2): 87-97].

Shcherbakov, D. 1988. Origin and evolution of Auchenorrhyncha based upon fossil evidence. Ab-

stracts and authors index. p. 8. Proc. X VIII Int. Cong. Entomol., 1988, Vancouver, BC, Canada.

(abstract)

Shear, W. A. 1991. The early development of terrestrial ecosystems. Nature, 351: 283-289.

Sidow, A./C. Stewart. Parsimony or Statistics? (Scientific Correspondence). Nature, 367 (6 Jan. 1994):

26-27.

Smith, J. W. & G. P. Georghiou. 1972. Morphology of the tymbal organ of the beet leafhopper,

Circulifer tenellus. Ann. Entomol. Soc. Am., 65: 221-226.

Smothers, F. J., C. D. Von Dohlen, L. H. Smith Jr. & R. D. Spall. 1994. Molecular evidence that

the myxozoan protists are metazoans. Science, 265 (16 Sep 1994): 1719-1721.

Sorensen, J.T. 1990. Taxonomic partitioning of discrete-state phylogenies: relationships of the aphid

subtribes Eulachnina and Schizolachnina (Homoptera: Aphididae: Lachninae). Ann. Entomol.

Soc. Am., 83: 394—408.

Sorensen, J. T. 1992. The use of discriminant function analysis for estimation of phylogeny: parti-

tioning, perspective and problems. pp. 65-93. Jn Sorensen, J. T. & R. Foottit (eds.). Ordination

in the study of morphology, evolution and systematics of insects: applications and quantitative

genetic rationals. Elsevier Science Publishers, Amsterdam,. the Netherlands.

Steel, M. A., M. D. Hendy & D. Penny. 1993. Parsimony can be consistent! Syst. Biol., 42: 581-

587.

Swofford. 1993. PAUP: phylogenetic analysis using parsimony (Version 3.1.1). [A computer pro-

gram.] Illinois Natural History Survey, Champaign, Illinois.

Ullman, D. E. & D. L. McLean. 1988a. The probing behavior of the summer-form pear psylla.

Entomol. Exp. Appl., 47: 115-125.

Ullman, D. E. & D. L. McLean. 1988b. Feeding behavior of the winter-form pear psylla, Psylla

pyricola (Homoptera: Psyllidae), on reproductive and transitory host plants. Environ. Entomol.,

17: 675-678.

Vawter, L. & W. M. Brown. 1993. Rates and patterns of base change in the small subunit ribosomal

RNA gene. Genetics, 134: 597-608.

Wake, D. B. 1991. Homoplasy: the result of natural selection, or evidence of design limitation.

Amer. Nat., 138: 543-567.

Wake, D. B. & A. Larson. 1987. Multidimensional analysis of an evolving lineage. Science, 238 (2

Oct 1987): 42-48.

Wheeler, W. C., R. T. Schuh & R. Bang. 1993. Cladistic relationships among higher groups of

Heteroptera: congruence between morphological and molecular data sets. Entomol. Scand., 24:

121-137.

Whitfield, J. B. & W. R. M. Mason. 1994. Mendesellinae, a new subfamily of braconid wasps

(Hymenoptera, Braconidae) with a review of relationships within the microgastroid assemblage.

Syst. Entomol., 19: 61-76.

60 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Wilson, S. W., C. Mitter, R. F. Denno & M. R. Wilson. 1994. Evolutionary patterns of host plant

use by delphacid planthoppers and their relatives. pp. 7-113. In Denno, R. F. & T. J. Perfect

(eds.). Planthoppers, their ecology and management. Chapman & Hall, New York.

Wootton, R. J. 1965. Evidence for tracheal capture in the early Heteroptera. pp. 65-67. Proc. XII

Int. Cong. Entomol., 1964, London.

Wootton, R. J. 1979. Function, homology and terminology in insect wings. Syst. Entomol., 4:

81-93.

Wootton, R. J. 1981. Palaeozoic insects. Ann. Rev. Entomol., 26: 319-344.

Wootton, R. J. & C. R. Betts. 1986. Homology and function in the wings of Heteroptera. Syst.

Entomol., 11: 389-400.

Zhang, Z. T. & L. Y. Chen. 1987. Analysis of courtship signals in planthoppers (Nilaparvata lugens

and Sogatella furcifera) and leafhopper (Nephotettex cincticeps). Kexue Tongbao (Sci. Bull.),

20: 1583-1586.

Zhang, Z. T., W. Z. Korg, J. L. Gao & T. H. Shao. 1988. Acoustic signal-producing organ of brown

planthopper. Int. Rice Res. Newsl., 13: 38-39.

Note added in final galley: The homoplasy (in literature) of sites 79 (dipterans, ex Carmean et

al. 1992) and 454 (various heteropteran lineages, ex Wheeler et al. 1993) was discovered after

acceptance of this manuscript, and was addressed in initial galley, along with insertion of

footnotes 7 (p. 41) and 8 (p. 54). We have since tested the effect of removal of these sites on

generation of the most parsimonious topology for the modified SET 4 matrix. The absence of 454,

leaving only 19 nucleotides, created six MLT's (TL 28, CI 0.714) rather that the single SET 4 MLT.

The additional absence of 79, leaving 18 nucleotides, produced the identical six topologies (TL

27, CI 0.704). These are:

(A) {{{{{CICAD , MIRID }, CERCO }, MEMBR }, DELPH }, STERN }

(B) {{{{{CICAD , MIRID }, MEMBR }, CERCO }, DELPH }, STERN }

(D) {{{{{MEMBR , MIRID }, CERCO }, CICAD }, DELPH }, STERN }

(E) {{{{{CICAD , CERCO }, MEMBR }, MIRID }, DELPH }, STERN }

(F) {{{{MEMBR , CERCO }, CICAD }, { DELPH , MIRID }} , STERN }

These MLTs all negate clade Auchenorrhyncha. MLT F is identical with the original SET 4

MLT, espousing clade Neohemiptera. MLT E places Heteroptera as sister clade to clade

Clypeorrhyncha. MLT's A-C negate clade Clypeorrhyncha, placing the heteropteran variously

among its members. The 50% majority rule consensus tree, with compatible groupings, from

these MLTs is the same as MLT C, as:

{{{{CICAD , MIRID } 50, { MEMBR , CERCO } 33 } 83, DELPH } 100, STERN }

However, our further analyses (see footnote 7) using additional taxa (Campbell et al.,

unpublished data), to be published elsewhere, together with significant morphological

synapomorphies that we do not consider to be selection-induced homoplasies, indicate the

monophylly of Clypeorrhyncha. Thus, our suborder proposal remains unaffected.

PAN-PACIFIC ENTOMOLOGIST

71(1): 61-63, (1995)

Scientific Note

AQUATIC MACROINVERTEBRATE RESPONSE TO

SHORT-TERM HABITAT LOSS IN EXPERIMENTAL

POOLS IN THAILAND

Biotic interactions can cause community structures to change temporally so

that new communities in recently disturbed habitats differ from established ones

in similar undisturbed habitats (Sousa, W. P. 1979. Ecology, 60: 1225-1239). In

this study, I contrasted aquatic macroinvertebrate communities in experimental

pools that dried temporarily and were then reflooded to those in similar habitats

that remained flooded.

This study was conducted using facilities at Chiang Mai University in northern

Thailand. The conditions in natural rain-filled pools were simulated using 10

cement tanks, each 80 cm in diameter, that were flooded by rainfall in September

1990. To provide sediments for benthic invertebrates, I added 0.5 liter of soil to

each tank. Tanks were paired spatially so that both tanks in each pair received

similar amounts of shading and organic matter from falling tree leaves. In most

cases, much of the water column in tanks eventually became filled with leaves.

From September through December 1991, rainfall kept experimental pools flood-

ed and aquatic invertebrate communities developed naturally within them.

With the onset of Thailand’s winter dry season in December, the pools began

to dry. When water depths in tank pairs had declined to approximately 2 cm, I

randomly selected one member of each pair to be refilled to a depth of 4 cm by

adding 10 liters of tap water per tank; water was left standing for 24 h prior to

introduction to allow chlorine to dissipate. Water in all of the tanks continued to

evaporate, and the unfilled tanks in each of the five pairs dried completely during

late January. An examination of the dry leaves and sediments in these tanks

revealed no live aquatic invertebrates. These tanks were allowed to remain dry

for 48 h.

I then refilled these disturbed tanks to a depth of 4 cm by adding 20 liters of

water per tank. Concurrently, 10 liters of water was added to the non-dried tank

of each pair. I continued to add equal amounts of water to paired tanks, as needed,

to maintain depths of at least 2 cm; in mid-March, all tanks were allowed to dry.

Thus, in this experiment, identical volumes of water were added to paired tanks

but one half of each pair was disturbed by drying for two days in January whereas

the other retained water from September through March.

In early February, 10 days after January refloodings of dried tanks, macroin-

vertebrates in each of the five tank pairs were sampled. A 10-cm by 6-cm net

(0.1-mm mesh) was swept across a randomly selected transect spanning each

tank’s 80-cm diameter; the distal edge of the net was scraped along the bottom

so that benthic invertebrates were collected. This technique had the desirable

feature of not being destructive to the habitats but probably under-sampled benthic

and fast-swimming organisms. Thus, for the sample collected at the experiment’s

end in mid-March, I vigorously stirred the water and sediments for 1 min and

62 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Table 1. Numbers of macroinvertebrates collected from experimental tanks that had been disturbed

by drying and reflooding in January 1991 compared to numbers in paired habitats that did not dry

(undisturbed). The February sweep-net samples were collected 10 days after disturbed habitats were

reflooded. For March samples, collected six weeks after refloodings, water and sediments were stirred

vigorously before sweep netting to increase capture efficiency for benthic organisms.

February samples

: : March samples

Disturbed habitats poe 62 ee St

(number per sweep _- Undisturbed habitats Disturbed habitats Undisturbed habitats

Taxa [SE]) (number per sweep [SE]) (number per sweep [SE]) (number per sweep [SE])

Ephemeroptera

Baetidae 6 (2) 4 (2) 23 (32) 15 (21)

Diptera

Culicidae 59 (14) 6 (5)? 35 (11) 1 (0)

Chironomidae 12 (6) 7 (1) 92 (45) 76 (6)

Ephydridae 8 (2) 3 (1) <1 <l

Ostracoda 3 (2) 179 (113) 116 (116) 163 (69)

@ Numbers differed significantly (paired t-test; P < 0.05) between treatments.

swept the net through the slurry, selecting transects as above. The numbers of

macroinvertebrates collected per treatment were contrasted using paired f-tests.

In February samples (Table 1), numbers of seed shrimp (Ostracoda) in the

disturbed habitats were < 2% of those in undisturbed habitats (3 + 2 ostracods/

sweep vs. 179 + 113 ostracods/sweep; paired t-test, P < 0.01). Ostracod im-

matures and adults are noted for their ability to tolerate drought if sediments

remain humid (Wiggins, G. B., R. J. Mackay & I. M. Smith. 1980. Archiv f.

Hydrobiol. Supple., 58: 97-206), so their reductions from January dryings suggests

the disturbance from drought was severe. However, numbers of Culex quinque-

fasciatus Say mosquito larvae (Diptera: Culicidae) were 10 times greater in habitats

that had dried in January and were reflooded for only 10 days than in habitats

that had retained water since September (59 + 14 larvae/sweep {1 SE} vs. 6 +

5 larvae/sweep; paired t-test, P < 0.01). In addition, analyses (paired t-tests, P

> 0.05) failed to detect density differences between disturbed and undisturbed

habitats for the other common invertebrates such as midge larvae (Diptera: Chi-

ronomidae) (12 + 6 larvae/sweep vs. 7 + | larvae/sweep), brine fly larvae (Diptera:

Ephydridae) (8 + 2 larvae/sweep vs. 3 + 1 larvae/sweep), and mayfly nymphs

(Ephemeroptera: Baetidae) (6 + 2 nymphs/sweep vs. 4 + 2 nymphs/sweep).

Clearly, rates of recolonization and development by these insect species were

rapid. I occasionally saw predators such as dragonfly nymphs (Odonata: Libel-

lulidae) and backswimmer adults (Hemiptera: Notonectidae) in undisturbed tanks

but these organisms were rare and were not collected in sweep samples.

In March, samples (Table 1), collected six weeks after January refloodings,

mosquito larvae were still abundant in the habitats that dried in January, and

significantly more larvae were collected there than in habitats that had not dried

(35 + 11 larvae/sweep vs. 1 + O larva/sweep; paired t-test, P < 0.01). By March,

ostracod numbers in the habitats that dried in January had rebounded so that

numbers were no longer significantly different in disturbed vs. undisturbed hab-

itats (116 + 116 ostracods/sweep vs. 163 + 69 ostracods/sweep; P > 0.05).

1995 SCIENTIFIC NOTE 63

Although the capture efficiency for mosquitoes and ostracods appeared to be

similar for the two sampling variations used in February and March, the stirring

of the sediments prior to sweep netting clearly increased the capture rates of

benthic midges and mayflies in March. I collected 92 + 45 midge larvae/sample

in disturbed habitats, and 76 + 6 larvae/sample in undisturbed ones. For mayflies,

23 + 32 nymphs/sample were collected in disturbed habitats vs. 15 + 21 nymphs/

sample in undisturbed habitats. However, as in the February sample, midge and

mayfly numbers during March did not differ significantly between treatments (P

> 0.05). Brine fly larvae were rare during March.

Although disturbance from drying and reflooding habitats is known to benefit

Aedes spp. and other mosquitoes that have desiccation resistant eggs (Wiggins et

al. 1980), the temporary drawdowns in this study also increased densities of Culex

mosquitoes, which do not have desiccation resistant eggs. The greater numbers

of mosquito larvae in the drought disturbed habitats may have been related to

differences in water chemistry between treatments, which may have influenced

Oviposition rates or larval survivals. Water pH was lower in disturbed habitats

than in undisturbed habitats (8.1 + 0.4 vs. 8.5 + 0.2, respectively; P < 0.05,

paired t-test). The water in the tanks that dried was in all cases also visibly darker

brown than in corresponding tanks that had remained flooded, suggesting greater

concentrations of humic materials. Although shallow, organic-rich habitats with

few predators are considered to be prime habitats for mosquito larvae (Laird, M.

1988. The natural history of larval mosquito habitats. Academic Press, New

York), this axiom applied only to habitats that temporarily became dry and not

to habitats that remained flooded.

Aside from the strong responses to drying by mosquitoes, the remaining mac-

roinvertebrates showed little response to the disturbance. Keys were not available

to determine species compositions of the immature insects (except mosquitoes),

so the failure to detect response at the community level may be in part a product

of low taxonomic resolution. However, all common species occurred in both

treatments. Alternatively, a combination of rapid rates of recolonization and

growth in the warm tropical conditions may have limited the impact of disturbance

in these pools. Although January dryings appeared to kill all invertebrates in the

disturbed habitats, the fast-maturing insects present in the undisturbed habitats

during January probably also disappeared shortly thereafter via emergence. Sub-

sequent recolonization of both habitat types was rapid, and colonizers (except for

mosquitoes) failed to differentiate between disturbed and undisturbed habitats.

Thus, disturbance may be less important to community development in habitats

dominated by fast-growing immature insects than in habitats dominated by other

types of organisms (e.g., Power, M. A. & A. J. Stewart. 1987. Amer. Midl. Nat-

uralist, 117: 333-345).

Acknowledgment. —I appreciated the cooperation in this study of the Depart-

ment of Entomology at Chiang Mai University, Thailand and editorial comments

on an earlier draft of this paper by Vincent H. Resh.

Darold P. Batzer, 218 Wellman Hall, Department of Entomological Sciences,

University of California, Berkeley, California 94720.

PAN-PACIFIC ENTOMOLOGIST

71(1): 64-65, (1995)

Scientific Note

LALAPA LUSA PATE (HYMENOPTERA: TIPHIIDAE):

NEW LOCALITIES AND NEW FLORAL

ASSOCIATIONS IN THE PACIFIC NORTHWEST

Lalapa lusa Pate (Hymenoptera: Tiphiidae) is a distinctive wasp and the sole

North American representative of the subfamily Anthoboscinae. As such, it has

attracted the attention of aculeate Hymenoptera workers. Lalapa lusa is known

from Idaho, Washington, Oregon and California (Krombein, K. V. 1979. Cat.

Hymen. Amer. N of Mex., 2: 1268) where it occurs in arid and semiarid regions.

Yet, it remains rare in collections and its biology is unknown.

The only previous Idaho record was a single female collected at Hollister, on

21 Aug 1930, which was designated as the holotype (Pate, V. S. L. 1947. J. N.Y.

Entomol. Soc., 55: 115-145). There were two previously known collections of L.

lusa from Washington: at Whiskey Dick Canyon, (M. Wasbauer, personal com-

munication), and at Burke, which came from the M. T. James Collection, De-

partment of Entomology, Washington State University.

We collected twelve additional specimens of L. /usa from Idaho and Washington

(see records). Collections were made by sweeping low vegetation, the blossoms

of rabbitbrush, Chrysothamnus nauseosus (Pallas) Britton (Asteraceae), and Can-

ada thistle, Cirsium arvense (L.) Scopoli (Asteraceae). Specimens were also ob-

tained from flowers of white sweet clover, Melilotus alba Desrousseaux (Fabaceae),

and from yellow pan traps.

All four floral associations for L. /usa are new. Lalapa lusa was previously

known wild buckwheat flowers, Eriogonum sp. (Polygonaceae), in Riverside Co.,

California, and feeding on the exudate of Disholcaspis sp. galls (Hymenoptera:

Cynipidae) on oak, Quercus sp. (Fagaceae) (M. Wasbauer, personal communi-

cation).

Lalapa lusa are 6-7 mm long, mostly matte black with a red apex of the

abdomen. The body is largely covered with long, white to tan setae that are

particularly dense on the apical abdominal segments. These characters allow dis-

crimination of L. /usa from most other tiphiids in the field. This is important

because it was found sharing inflorescences of Canada thistle and white sweet

clover with a far more abundant Paratiphia sp. (prob. neomexicana Cameron)

(Hymenoptera: Tiphiidae). Under the microscope Lalapa lusa can be distin-

guished from other tiphiids by examining the middle and hind tibiae which are

flattened and bear several rows of short, thick, blunt, peg-like tubercles. The wasps

are active from mid summer to early fall. Pan trap collections indicate that they

spend a good deal of time flying low over the ground, as would be typical for a

fossorial wasp. Thus, pan traps or flight intercept traps may be the best means of

sampling for L. /usa. Subsequent observations in the area may then reveal bio-

logical secrets of another fascinating wasp and yield data important to our un-

derstanding of the Tiphioidea.

Records.—IDAHO. NEZ PERCE Co.: Hatwai Crk, 8 km (5 mi) E of Lewiston, 31 Jul 1984, W. J.

1995 SCIENTIFIC NOTE 65

Turner, Chrysothamnus nauseosus (Pallas) Britton (Rabbitbrush, Asteraceae) (1); Hell’s Gate State

Park, 6.4 km (4 mi) E of Lewiston, 14 Jul 1982, T. D. Miller, Cirsium arvense (L.) Scopoli (Canada

thistle, Asteraceae) (1); same loc., but 28 Jul 1982, J. B. Johnson, sweeping low vegetation, (1); same,

but 14 Jul 1983, Melilotus alba Desrousseaux flowers (white sweet clover, Fabaceae) (2); same, but

22 Aug/6 Sep 1984, T. D. Miller, yellow pan trap (3); same, but 10/20 Aug 1985 (2). OWYHEE Co.:

Murphy, 18 Jul 1982, J. B. Johnson, sweeping low vegetation (1). TWIN FALLS Co.: Hollister, 21

Aug 1930 (holotype). WASHINGTON. BENTON Co.: 8 km (5 mi) N of Richmond, 2 Oct 1982, J.

B. Johnson, sweeping low vegetation (1). GRANT Co.: Burke, 9 Aug 1950, E. Klostermeyer, sweeping

Salsola kali L. (Russian thistle, Chenopodiaceae). KITTITAS Co.: Whiskey Dick Cyn, 12.8 km (8

mi) N of Vantage.

Acknowledgment. —Published with the approval of the director of the Idaho

Agricultural Experiment Station, as Research Paper 93733.

James B. Johnson!, Terry D. Miller?, William J. Turner?, '!Department of Plant,

Soil and Entomological Sciences, University of Idaho, Moscow, Idaho 83844-

2339; 2Department of Entomology, Washington State University, Pullman, Wash-

ington 99164-6432; >Department of Entomology, Washington State University,

Pullman, Washington 99164-6432.

PAN-PACIFIC ENTOMOLOGIST

71(1): 65-66, (1995)

Scientific Note

NEW RECORDS OF TRICHOSTERESIS FOERSTER FROM

THE WESTERN UNITED STATES

(HYMENOPTERA: MEGASPILIDAE)

Only two species of Trichosteresis Foerster are recorded from North America—

T. floridana Ashmead from Jacksonville, Florida, and 7. vitripennis Whittaker

from Chilliwack, British Columbia (Muesebeck, C. F. W. 1979. Catalog of Hy-

menoptera of North America: 1191). Their hosts are not known, but European

species have been reared from syrphid pupae. However Dessart (Dessart, P. 1974.

Ann. Soc. Entomol. Fr., (N.S.) 10:395-448) believes Trichosteresis is monotypic,

and all other species, are synonyms of 7. glabra (Boheman).

In 1993, I found T. vitripennis in two widely-separated locations in California,

suggesting that it occurs throughout the state. In April, I collected a female on

the roof of my car that had just come through a car wash in the City of San

Bernardino (see records). In June, I swept two females from the foliage of Quercus

agrifolia Nee near the North Oakland Sports Center in Oakland. Determinations

were made using a generic key to world Ceraphronoidea (Alekseyev, V. N. 1978.

Entom. Rev., 57: 449-453) and the species description (Whittaker, O. 1930. Proc.

Entom. Soc. Wash., 32: 72-73); no comparisons with the types were made. These

discoveries caused me to reexamine the material in the Essig Museum, where I

found four additional Trichosteresis specimens. One was a T. vitripennis female,

which I had swept from Medicago sativa L. at U.C. Berkeley’s experiment station

in Albany. The other three were females that did not match the descriptions of

either 7. floridana or T. vitripennis; these were collected at the Needle Rocks at

the north end of Pyramid Lake, Nevada (collector unknown).

66 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

Trichosteresis is easily distinguished from other megaspilids by its forewing

characteristics (lacking a marginal fringe and the disc mostly bare or with micro-

scopic hairs). The three species from North America also have the radial vein

shorter than the breadth of the pterostigma. The mesonotum of T. vitripennis has

a complete and deeply impressed longitudinal median line, and two anterior

longitudinal furrows between the median line and notauli; in 7. floridanus the

median line is incomplete posteriorly and the two longitudinal furrows absent.

The species from Nevada closely resembles 7. vitripennis, but its first funicular

segment is about the same length as the pedicel; in 7. vitripennis, this segment 1s

1.5-2 times longer than the pedicel.

All seven specimens are deposited in the collection at the Laboratory of Bio-

logical Control in Albany, part of the Essig Museum at the University of California

at Berkeley.

Records. — Trichosteresis vitripennis. CALIFORNIA. ALAMEDA Co.: Oakland, 3 and 10 Jun 1993,

R. Zuparko, Quercus agrifolia Nee, 2 females; Albany, 3 Oct 1992, R. Zuparko, Medicago sativa L.,

1 female. SAN BERNARDINO Co.: San Bernardino, 12 Apr 1993, R. Zuparko, 1 female.

Trichosteresis sp.. NEVADA. WASHOE Co.: Needle Rocks at N end of Pyramid Lake, 15 Sep

1983.

Robert L. Zuparko, Laboratory of Biological Control, University of California,

Berkeley, 1050 San Pablo Avenue, Albany, California 94706.

PAN-PACIFIC ENTOMOLOGIST

71(1): 66-68, (1995)

Scientific Note

INITIATION OF MATING ACTIVITY AT THE TREE

CANOPY LEVEL AMONG OVERWINTERING

MONARCH BUTTERFLIES IN CALIFORNIA

Overwintering monarch butterflies, Danaus plexippus (L.) (Danaidae, Lepi-

doptera) congregate west of the Rocky Mountains in clusters, ranging from a few

-hundred to several thousand individuals, in certain groves along the California

coastline. This aggregation behavior, the result of migration to escape the rigors

of winter, is believed to lessen bird predation (Fadem, C. M. & A. Shapiro. 1979.

Pan-Pacif Entomol., 55: 309-310) and concentrates the sexes in a locale that

increases their chances of mating in the spring. Mating activity at a wintering

grove may occur sporadically throughout the winter, but begins in earnest by early

or mid-February and lasts approximately two weeks. Females usually mate with

several males before their spring dispersal to oviposition sites (Hill, H. F., A. M.

Wenner & P. H. Wells. 1976. Amer. Mid. Nat., 95: 10-19). The males capture

females in flight (Hill et al. 1976) or “nudge” them toward the ground (Pliske, T.

E. 1975. Ann. Ent. Soc. Amer., 68: 143-151) before mating. Similar behavior has

been observed in the laboratory (Rothschild, M. 1978. Antenna, 2: 38-39) and

in summer populations (Zalucki, M. P. 1982. J. Aust. Ent. Soc., 21: 241-246).

1995 SCIENTIFIC NOTE 67

Table 1. The average number of canopy and ground butterfly pairs recorded at two-h intervals

(08:00 h to 14:00 h) on. 6 separate days in February, 1993. The mean and SE are presented for each

observational period.

Time interval

08:00 h 10:00 h 12:00 h 14:00 h

Canopy attempts

Male-male 4.3+ 1.9 12.0 42.2 6.21] 1.4 + 0.6

Male-female 1.2 + 0.6 $3.2 0:7 3.5 + 0.6 0.8 + 0.4

Ground pairs

Male-male 1.2 + 0.6 2.2°0,7 L214 0.2 + 0.4

Male-female Lee ies 5.3: 4.0.7 3.3 + 0.8 1.6 + 0.9

During February, 1993, at a winter site in Pismo Beach North State Park, San -

Luis Obispo County, California, (35°07'46” latitude; 120°37'53” longitude), I ob-

served, using a spotting scope (20 x 45, Zoom Bushnell Spacemaster) and bin-

oculars (10 x 50 Bushnell), mating activity occurring at the canopy level of large

blue gum, (Eucalyptus globulus Labillardiere) and Monterey cypress, (Cupressus

macrocarpa Gordon) trees. Males captured potential mates sunning on the foliage

within butterfly clusters and/or on the foliage of neighboring trees. The male

would fly above a butterfly (either a female or another male) that was resting on

the foliage with outstretched wings and pounce and grasp the thorax of the in-

dividual with its legs. The pair frequently displayed rapid wing fluttering and other

interactions typical of copulating pairs (Pliske 1975). If male-female paired, the

struggling butterflies, after 3 to 10 sec, would slowly fall from the foliage to the

ground and complete or abandon their mating activity. On two occasions, I

observed a male capture and successfully couple with a female at the canopy level.

Both matings occurred on the flat and shelf-like foliage of Monterey cypress. If

two males, the interplay between them lasted < 5 sec, but was occasionally longer,

especially toward the end of the overwintering season. In such cases, the pair fell

from the canopy foliage and separated either before, or shortly after, landing on

the ground.

I documented the initial and latter phase of mating activity by recording the

number of canopy and ground butterfly pairs (male/female & male/male) at two-

h intervals, starting at 08:00 h and ending when mating activity ceased (14:00 h).

At the beginning of each interval, the number of ground pairs was counted beneath

each cluster tree, before recording the number of canopy mating attempts observed

during a thirty min period. At the end of thirty min, the number of paired

butterflies on the ground was again recorded. This procedure was repeated on 6

separate days in February. The average number of canopy and ground pairing

attempts for each two-h interval showed a similar pattern of mating activity at

the foliage level and on the ground (Table 1). Pairing attempts began slowly in

the morning, reached a peak at 10:00 h and slowly declined as daytime temper-

atures cooled. There were more male/male than male/female encounters at the

canopy level, indicating that D. plexippus seek mates visually during the early

stages of the mating sequences, resulting in many missed pairings. This relation-

ship was exactly opposite on the ground, however, where more male/female than

male/male couples were observed. Once physical contact of courtship is initiated,

68 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

the males are able to discriminate between the sexes (via visual or short distance

chemical communication) resulting in longer interplay and frequently to a suc-

cessful union.

I believe that the initiation of mating sequences on the canopy is common

among overwintering monarch butterflies and that it has been overlooked by

earlier investigators. Once observed, it can be recognized easily; I examined and

noticed this mating activity on a kodachrome slide I took in the winter of 1990-

1991 and recently, in a picture of Mexican monarch butterflies on roosting trees

published with a article in Natural History (Larsen, T. 1993. Nat. His., (6): 30-

39). The capture of mates in flight, at least for California overwintering monarch

butterfly populations, may not be as frequent as those initiated at the canopy

level. I observed only one in-flight capture of a female by a male during many

hours of field observations. This single event was observed when the female slowed

her flight while trying to land on foliage, and was then captured by a male. The

capture of stationary “mates”’ concentrated in a small area (foliage of roosting

trees), after months of overwintering, saves time and energy, and provides max-

imum opportunities for mating.

Acknowledgment. —I thank Dennis Frey, Harry Kaya and Aryan Roest for their

critical review and comments and Christopher Nagano for asking the question

“where” mating was occurring.

Kingston L. H. Leong, Department of Biological Sciences, California Polytechnic

State University, San Luis Obispo, California 93407.

PAN-PACIFIC ENTOMOLOGIST

71(1): 68-71, (1995)

Scientific Note

OBSERVATIONS OF THE FORAGING PATTERNS OF

ANDRENA (DIANDRENA) BLENNOSPERMATIS THORP

(HYMENOPTERA: ANDRENIDAE)

Pollinator foraging movements can determine pollen transfer among flowers

and thus may affect pollen and gene flow within and among plant populations

(Handel, S. N. 1983. pp. 163-211. In Real, L. (ed.). Pollination biology. Academic

Press Inc.). From a landscape perspective, pollinator foraging patterns form a

spatial link among available floral resources that is bounded by the particular

species’ flight and floral preferences. Foraging studies that emphasize the latter

can also help identify the spatial requirements of insect pollinators. Therefore,

quantification of pollinator foraging patterns can yield information pertinent to

floral ecology and evolution as well as to landscape utilization by insect pollinators.

Unfortunately, many insect pollinator foraging patterns are not well documented,

and that is especially true for solitary bee species.

1995 SCIENTIFIC NOTE 69

WHITE

100m

80m

TORANGE}

Figure 1. Schematic diagram of Blennosperma nanum nanum patches observed. The approximate

shapes of patches, and distances between patches are as indicated. The approximate sizes of the white,

orange, and yellow patches are 54 m?, 560 m?, and 960 m7, respectively.

We studied the foraging patterns of the native, solitary bee Andrena (Diandrena)

blennospermatis Thorp, which is oligolectic on Blennosperma nanum nanum

(Hook.) Blake, a vernal pool plant (Thorp, R. W. 1969. Univ. Calif. Publ. En-

tomol., 52: 1-146). Specifically, we sought to determine whether individual fe-

males of A. blennospermatis forage between discrete patches of B. n. nanum. Our

primary purpose was to evaluate the spatial utilization of discrete floral patches

by female A. blennospermatis during one flight season. Consequently, in this study,

we are interested in the spatial, rather than temporal, patterns of female A. blen-

nospermatis foraging.

We conducted this study at Jepson Prairie Preserve, near Dixon, Solano Co.,

California from 6-18 Mar 1993. This period encompassed the peak of the adult

flight season of this bee in 1993. To observe interpatch foraging, we marked and

subsequently observed marked female bees in three separate patches of B. n.

nanum (Fig. 1) on each of nine nonconsecutive days (March 6-8, 10-11, 13-14,

16, 18) between 10:00-13:00 h. However, the period of time we spent marking

and observing each day varied to some degree, depending upon weather conditions

and flight activity. Except for the first date, one person marked and observed bees

70 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

in all patches on each day. On the first date, a total of six people simultaneously

marked and observed bees, two per floral patch. The number of bees marked per

patch per date varied (1-11 bees) yielding a total of 64 marked bees.

We marked female bees by placing a dot of enamel paint on the thorax using

a plastic dental toothpick. Each floral patch was assigned a unique color: white,

yellow, or orange. Female bees netted within a floral patch were marked with that

patch’s color and immediately released. To assess the extent of interpatch foraging

by A. blennospermatis females, we recorded observations of foraging females in

which the female’s color was different from the assigned patch color. Because we

were concerned with the spatial patterns of floral utilization, rather than the

temporal aspects of those foraging patterns, we kept each patch color constant

throughout the study. This system of marking, however, did not allow us to

distinguish individuals marked the same color unless they were sighted simul-

taneously. Voucher specimens are deposited in the Bohart Museum, University

of California, Davis.

Our observations indicate that limited interpatch foraging by female A: blen-

nospermatis occurs. Of 28 females marked in the white patch, only 2 (7%) were

observed foraging simultaneously in the orange patch (Fig. 1). Of 25 females

marked in the orange patch, only 1 (4%) was observed foraging in the white patch.

None of the 11 females marked in the yellow patch were observed foraging in the

other two patches. These observations suggest that individual females occasionally

forage between B. n. nanum patches 25 m apart (Fig. 1), but rarely forage, if at

all, between patches 80-100 m apart. In contrast, we commonly observed intra-

patch foraging by females that we were able to visually follow directly after mark-

ing. We cannot determine whether these instances of interpatch foraging represent

movements within a single foraging bout, within a single day, or between days.

It is unclear how the floral patches (Fig. 1) are spatially related to the nest sites

of the marked A. blennospermatis females. Although there were several areas near

the white and orange patches that contained scattered nests of A. blennospermatis,

Andrena (Tylandrena) sp. and other Andrena spp., we did not identify the indi-

vidual nests of the marked females. However, we did identify the nests of two A.

blennospermatis females whom we later individually marked. We observed one

bee foraging in the orange patch that was approximately 50 m away from her

nest. We observed the other female foraging at B. n. nanum patches 25 m or

closer to her nest.

Our observations of limited interpatch foraging and common intrapatch for-

aging by A. blennospermatis females are consistent with other studies of andrenid

bee foraging behavior. These studies (Danforth, B. N. 1989. J. Kansas Entomol.

Soc., 62: 59-79; Thorp 1969, 1990. pp. 109-122. In Ikeda, D. H. and R. A.

Schlising (eds.). Vernal pool plants—their habitat and biology. Studies from the

Herbarium No. 8, California State University, Chico.) suggest that the foraging

areas of several andrenid bee species are spatially very limited. Thorp (1990)

found that at Jepson Prairie Preserve, most females of Andrena (Hesperandrena)

limnanthis exhibited limited foraging areas and were observed 10 m or closer to

the original marking site. Our observations suggest that A. blennospermatis females

tend to forage within a particular B. n. nanum patch. Consequently, it is likely

that females largely transfer pollen within a B. n. nanum patch rather than between

patches (Fig. 1). Because A. blennospermatis females are one of the most common

1995 SCIENTIFIC NOTE 71

visitors to B. n. nanum flowers (Thorp 1969, 1990), they may have the potential

to strongly influence pollen flow within B. n. nanum patches.

Acknowledgment. —We thank K. Schick, J. Schick, D. Carmean, S. Heydon, C.

Heydon, and J. Barthell for assistance.

Joan M. Leong, Robert P. Randolph,! and Robbin W. Thorp, Department of

Entomology and Ecology Graduate Group, University of California, Davis, Cali-

fornia 95616; 'Current address: 310 S. Orange Ave. #33 Lodi, California 95240.

PAN-PACIFIC ENTOMOLOGIST

71(1): 71-74, (1995)

Scientific Note

LONG-TERM CHANGES IN OBSCURA GROUP

DROSOPHILA SPECIES COMPOSITION AT

MATHER, CALIFORNIA!

Since the 1940s, Dobzhansky and his collaborators have collected obscura group

Drosophila species from Mather, California, located at 1375 m on the western

slope of the Sierra Nevada. Mather’s Transition Zone association and moderate

climate have made it one of the most heavily-collected areas in the world for

Drosophila. Fifty years of accumulated data give us the rare opportunity to study

long-term changes in Drosophila species’ relative abundances. Such studies can

also identify effects of climatic changes on species frequencies. This study docu-

ments evolution in obscura group Drosophila species composition at Mather,

shows that the changes are associated with climate, and provides a baseline for

future investigations.

The genetics of the D. obscura group have been studied extensively, but their

ecology is largely unknown. The ranges of the four native California species vary

latitudinally and altitudinally. In zones of geographic overlap, Drosophila pseu-

doobscura Frolova and D. azteca Sturtevant & Dobzhansky are more frequent in

warmer and drier areas and at lower altitudes (Dobzhansky, T. & J. Powell. 1975.

pp. 537-587. In R. King (ed.). Invertebrates of genetic interest. Plenum Press,

New York.). Drosophila persimilis Dobzhansky & Epling and D. miranda Dob-

zhansky predominate at higher elevations and northern latitudes. Based on their

zoogeography, D. azteca and D. pseudoobscura are perceived as the more dry/hot

adapted of the four species.

Dobzhansky (1973. Evolution, 27: 565-575) noted an increase in the frequency

of D. persimilis relative to D. pseudoobscura at Mather. Figure 1A shows this

trend, supplemented with data from later collections by the persons mentioned

in the acknowledgment. However, such relative abundances can be deceptive

when the rest of the species group is ignored. Figure 1B shows the frequencies of

! Author’s page charges were partially offset by a grant from the C. P. Alexander Fund, PCES.

(per)/(per+pseudo)

(per or mir)/(totel obscura group)

(azteca)/(total obscura group)

100

THE PAN-PACIFIC ENTOMOLOGIST

r2=0.117

1940 1950 1960 1970 1980 1990 2000

Collection

O persimilis r2=0.117

O miranda r2=0.185

1970 1975 1980 1985 1990 1995

Collection

PZ 0.622

1970 1975 1980 1985 1990 1995

Collection

Vol. 71(1)

1995 SCIENTIFIC NOTE 73

Table 1. Multiple regression equation of arcsine transformed species frequency with cm rainfall

and year: month collection (e.g., June 1974 = 74.5).

Species Intercept Rainfall Year : month r n P

D. pseudoobscura — 13.56 —0.91 0.20 0.85 7 0.021

D. azteca —4.60 —0.05 0.08 0.68 12 0.006

D. persimilis and D. miranda relative to the frequencies of all obscura group

species for the last twenty years of collections. Except for one anomalous year,

their relative frequencies remained constant. However, D. azteca increased in

frequency dramatically (see Fig. 1C), and D. pseudoobscura’s frequency dropped

from 54% in 1974 to 2% in 1993.

Temperature data for Mather are not available, but I obtained rainfall measures

for the 1973-1993 period from the National Climatic Data Center. Multiple

regressions of D. pseudoobscura and D. azteca frequencies on collection date and

cm rainfall for the preceding month were significant (see Table 1). The frequencies

of both species regressed positive to collection date and negative to rainfall, but

the different coefficient magnitudes caused the opposite trends observed. The

negative regressions to rainfall corroborate the claim that these species prefer drier

environments.

Although rainfall partially explains these species’ relative frequencies, the as-

sociation with collection date suggests an unknown secular trend. The simplest

explanation is that D. azteca is replacing D. pseudoobscura, but there are many

alternative hypotheses. Is there something gradually accumulating or being de-

pleted in the environment causing the changes in frequencies? Do these changes

reflect differences in the intrinsic potential for these species to increase in numbers?

Many questions remain unanswered.

The processes occurring at Mather can be better understood with density studies.

Without an estimate of the number of flies present each year, we cannot say

whether any species are increasing or declining in absolute abundance. Absolute

numbers available for previous collections at Mather are unrevealing due to dif-

ferent collecting methods. Further, previous studies at Mather have shown that

the densities of these species vary considerably, even at the same time of year

over different years (Taylor, C. & J. Powell. 1983. pp. 29-59. Jn Ashburner, M..,

H. Carson & J. Thompson (eds.). The genetics and biology of Drosophila (Vol.

3). Academic Press, New York). Future studies should combine relative species

abundances with density measurements to ascertain how the observed relative

trends relate to changes in population sizes.

Acknowledgment. — Financial support was provided by a Genetics Training Grant

from the Department of Health and Human Services, a Grant-in-Aid of Research

—_—

Figure 1. Relative frequencies of Drosophila species collected at Mather. (A) shows the percentage

of D. persimilis flies relative to the total number of D. pseudoobscura plus D. persimilis flies caught.

(B) shows the percentage of D. persimilis and D. miranda flies relative to the total number of obscura

group flies captured. (C) shows the percentage of D. azteca flies relative to the total number of obscura

group flies collected.

74 THE PAN-PACIFIC ENTOMOLOGIST Vol. 71(1)

from Sigma Xi, and a Hinds Fund grant from the University of Chicago. J. Coyne,

J. Kelly, M. Wayne, and T. Wootton provided comments on the manuscript, and

A. Bronikowski provided statistical help. Thanks to F. Mestres, A. Prevosti, C.

Taylor, and especially, J. Powell for providing me with data from their collections

at Mather and to the Carnegie Institute of Washington for allowing the use of

their Mather facility for Drosophila collections.

Mohamed A. Noor, Department of Ecology and Evolution, University of Chi-

cago, 1101 East 57th Street, Chicago, Illinois 60637.

New Journal Submission Address,

New Laboratory Building :

Plant Pest Diagnostics Center

Effective 1 Jan 1995, the following address change should be noted for

all new submissions to The Pan-Pacific Entomologist:

Dr. John T. Sorensen

Editor - Pan-Pacific Entomologist

California Dept. of Food & Agriculture,

Plant Pest Diagnostic Center,

3294 Meadowview Road,

Sacramento, California 95832-1448

All correspondence to the PCES Treasurer or Secretary, including all

financial inquires, should continue to be addressed to their California

Academy of Sciences addresses. Correspondence directed to the Associate

Editor, Dr. Robert V. Dowell, should be addressed to him at: CDFA,

Pest Detection/Emergency Projects, 1220 "N" Street, Sacramento, CA 95814.

This address change is caused by the movement of CDFA's Insect

Taxonomy Laboratory to a newly constructed biology laboratory building,

the Plant Pest Diagnostics Center (PPDC). The new facility is in southern

Sacramento at the CDFA meadowview operations complex, which it shares

with the department's Chemistry Laboratories, greenhouses and facilities

for emergency projects and biological control.

The new 53,000 sq ft PPDC building houses separate laboratories for

Insect Biosystematics, Botany, Seeds, Plant Pathology and Nematology.

The Insect Biosystematics section has a 1.5 million specimen, compactor-

based insect collection, a nucleotide laboratory and a critical point drying

facility. Additionally, all PPDC laboratories share a photographic

darkroom, scanning and transmission electron microscope facilities and

a 25,000 volume, compactor-based library. The PPDC has a staff of 46,

including 22 scientists. Its insect biosystematists are: Fred Andrews,

Karen Corwin, Tom Eichlin, Eric Fisher, Ray Gill, Alan Hardy, Terry

Seeno, Ron Somerby, and John Sorensen.

PAN-PACIFIC ENTOMOLOGIST

Information for Contributors

See volume 66(1): 1-8, January 1990, for detailed general format information and the issues thereafter for examples; see below for

discussion of this journal’s specific formats for taxonomic manuscripts and locality data for specimens. Manuscripts must be in English,

but foreign language summaries are permitted. Manuscripts not meeting the format guidelines may be returned. Please maintain a copy

of the article on a word-processor because revisions are usually necessary before acceptance, pending review and copy-editing.

Format. — Type manuscripts in a legible serif font IN DOUBLE OR TRIPLE SPACE with 1.5 in margins on one side of 8.5 x 11 in,

nonerasable, high quality paper. THREE (3) COPIES of each manuscript must be submitted, EACH INCLUDING REDUCTIONS

OF ANY FIGURES TO THE 8.5 x 11 IN PAGE. Number pages as: title page (page 1), abstract and key words page (page 2), text

pages (pages 3+), acknowledgment page, literature cited pages, footnote page, tables, figure caption page; place original figures last.

List the corresponding author’s name, address including ZIP code, and phone number on the title page in the upper right corner. The

title must include the taxon’s designation, where appropriate, as: (Order: Family). The ABSTRACT must not exceed 250 words; use

five to seven words or concise phrases as KEY WORDS. Number FOOTNOTES sequentially and list on a separate page.

Text. — Demarcate MAJOR HEADINGS as centered headings and MINOR HEADINGS as left indented paragraphs with lead phrases

underlined and followed by a period and two hypens. CITATION FORMATS are: Coswell (1986), (Asher 1987a, Franks & Ebbet

1988, Dorly et al. 1989), (Burton in press) and (R. F. Tray, personal communication). For multiple papers by the same author use:

(Weber 1932, 1936, 1941; Sebb 1950, 1952). For more detailed reference use: (Smith 1983: 149-153, Price 1985: fig. 7a, Nothwith

1987: table 3).

Taxonomy. — Systematics manuscripts have special requirements outlined in volume 69(2): 194-198; if you do not have access to that

volume, request a copy of the taxonomy/data format from the editor before submitting manuscripts for which these formats are

applicable. These requirements include SEPARATE PARAGRAPHS FOR DIAGNOSES, TYPES AND MATERIAL EXAMINED

(INCLUDING A SPECIFIC FORMAT), and a specific order for paragraphs in descriptions. List the unabbreviated taxonomic author

of each species after its first mention.

Data Formats. — All specimen data must be cited in the journal’s locality data format. See volume 69(2), pages 196-198 for these

format requirements; if you do not have access to that volume, request a copy of the taxonomy/data format from the editor before

submitting manuscripts for which these formats are applicable.

Literature Cited. — Format examples are:

Anderson, T. W. 1984. An introduction to multivariate statistical analysis (2nd ed). John Wiley & Sons, New York.

Blackman, R. L., P. A. Brown & V. F. Eastop. 1987. Problems in pest aphid taxonomy: can chromosomes plus morphometrics

provide some answers? pp. 233-238. Jn Holman, J., J. Pelikan, A. G. F. Dixon & L. Weismann (eds.). Population structure, genetics

and taxonomy of aphids and Thysanoptera. Proc. international symposium held at Smolenice Czechoslovakia, Sept. 9-14, 1985.

SPB Academic Publishing, The Hague, The Netherlands.

Ferrari, J. A. & K. S. Rai. 1989. Phenotypic correlates of genome size variation in Aedes albopictus. Evolution, 42: 895-899.

Sorensen, J. T. (in press). Three new species of Essigella (Homoptera: Aphididae). Pan-Pacif. Entomol.

Illustrations. — Illustrations must be of high quality and large enough to ultimately reduce to 117 x 181 mm while maintaining label

letter sizes of at least 1 mm; this reduction must also allow for space below the illustrations for the typeset figure captions. Authors

are strongly encouraged to provide illustrations no larger than 8.5 x 11 in for easy handling. Number figures in the order presented.

Mount all illustrations. Label illustrations on the back noting: (1) figure number, (2) direction of top, (3) author’s name, (4) title of

the manuscript, and (5) journal. FIGURE CAPTIONS must be on a separate, numbered page; do not attach captions to the figures.

Tables. — Keep tables to a minimum and do not reduce them. Table must be DOUBLE-SPACED THROUGHOUT and continued

on additional sheets of paper as necessary. Designate footnotes within tables by alphabetic letter.

Scientific Notes. — Notes use an abbreviated format and lack: an abstract, key words, footnotes, section headings and a Literature Cited

section. Minimal references are listed in the text in the format: (Bohart, R. M. 1989. Pan-Pacific. Entomol., 65: 156-161.). A short

acknowledgment is permitted as a minor headed paragraph. Authors and affiliations are listed in the last, left indented paragraph of

the note with the affiliation underscored.

Page Charges. — PCES members are charged $35.00 per page, for the first 20 (cumulative) pages per volume and full galley costs for

pages thereafter. Nonmembers should contact the Treasurer for current nonmember page charge rates. Page charges do not include

reprint costs, or charges for author changes to manuscripts after they are sent to the printer. Contributing authors will be sent a page

charge fee notice with acknowledgment of initial receipt of manuscripts.

THE PAN-PACIFIC ENTOMOLOGIST

Volume 71 January 1995 Number 1

Contents

LINSLEY, E. G. & J. CHEMSAK— Obituary: Celeste Green, scientific illustrator, 1913-1994 ___.

GUILBERT, E., M. BAYLAC & J. NAJT—Canopy arthropod diversity in a New Caledonian

primary forest sampled by fogging

PUNZO, F.—Feeding and prey preparation in the solpugid, Eremorhax magnus Hancock

(Solpugida: Eremobatidae)

CROSLAND, M. W. J.—Nest and colony structure in the primitive ant, Harpegnathos venator

(Smith) (Hymenoptera: Formicidae)

HORTON, D. R. & T. M. LEWIS—Tethered flight characteristics of male and female pear

psylla (Homoptera: Psyllidae): comparison of pre-reproductive and reproductive

insects

SORENSEN, J. T., B. C. CAMPBELL, R. J. GILL & J. D. STEFFEN-CAMPBELL—Non-

monophyly of Auchenorrhyncha (““Homoptera’’), based upon 18S rDNA phylogeny: eco-

eco-evolutionary and cladistic implications within pre-Heteropterodea Hemiptera (s.1.)

and a proposal for new monophyletic suborders

SCIENTIFIC NOTES

BATZER, D. P.—Aquatic macroinvertebrate response to short-term habitat loss in experi-

mental pools in Thailand

JOHNSON, J. B., T. D. MILLER & W. J. TURNER—Lalapa lusa Pate (Hymenoptera: Ti-

phiidae): new localities and new floral associations in the Pacific Northwest

ZUPARKO, R. L.—New records of Trichosteresis Foerster from the western United States

(Hymenoptera: Megaspilidae)

LEONG, K. L. H.—Initiation of mating activity at the tree canopy level among overwintering

monarch butterflies in California

LEONG, J. M., R. P. RANDOLPH & R. W. THORP— Observations of the foraging patterns

of Andrena (Diandrena) blennospermatis Thorp (Hymenoptera: Andrenidae)

NOOR, M. A.—Long-term changes in obscura group Drosophila species composition at Mather,

California

Announcement: new journal submission address, new laboratory building: Plant Pest Diag-

nostics Center