Chemical ecology and biological control potential of the cockroach oothecal parasitoid Aprostocetus hagenowii (Ratzeburg)

MISSING IMAGE

Material Information

Title:
Chemical ecology and biological control potential of the cockroach oothecal parasitoid Aprostocetus hagenowii (Ratzeburg)
Physical Description:
vi, 101 leaves : ill. ; 29 cm.
Language:
English
Creator:
Suiter, Daniel Robert, 1963-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 87-100).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Daniel Robert Suiter.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002030137
notis - AKL7769
oclc - 33038019
System ID:
AA00003227:00001

Full Text









CHEMICAL ECOLOGY AND BIOLOGICAL CONTROL POTENTIAL
OF THE COCKROACH OOTHECAL PARASITOID
Aprostocetus hagenowii (Ratzeburg)












By

DANIEL ROBERT SUITER


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

anv aX











ACKNOWLEDGEMENTS

Although one name appears on the title page of this dissertation, it would not have

been possible without the resources and minds of several persons. Drs. P.G. Koehler and

R.S. Patterson allowed me the latitude to pursue this nontraditional area of pest control,

yet provided me with what I believe is a complete education. The cooperative research

program these two scientists have built is extraordinary, and I feel fortunate to have been a

part of their group for the past few years.

Special thanks go to Dr. Dave Carlson for allowing me to work out the chemical

ecology of my project in his laboratory. Drs. Dave Carlson, Bob Vander Meer, and Jim

Nation provided nearly 100 years of chemical ecology experience from which I benefitted


greatly. Drs.


Tom Walker and Richard Brenner provided insight on insect ecology, and


Dr. A. I. Khuri and Mr. Jay Harrison provided helpful recommendations in statistical data

analysis. Lloyd Davis helped me colonize the parasitoid. Behind every successful research

project is a great technician; in my case it was Michelle Hosack. Simply, the chemical

ecology aspect of my project would not have been possible without Michelle. I would like

to thank my family for keeping the faith when others did not. Most of all, I would like to

thank my dear friend and fiance, Lisa M. Ames, for her courage and patience. Advanced


degrees are all about sacrifice, and we have sacrificed plenty over the past


years.













TABLE OF CONTENTS


A CKN O W LED G EM EN TS .........................................................................................

A B STRIA CTS ............ ................ .... ........................................... ..... .... ... .. ..............

CHAPTER

I. INTRODUCTORY STATEMENT ........................................................................

II. LITERATURE REVIEW ........................ .............................................................


The Parasitoid .......................................................................................................
Kairomones. .. ..... ............ .. ................ ..


III. A SIMPLE REARING METHOD FOR THE COCKROACH OOTHECAL
PARASITOID Aprostocetus hagenowii (Ratzeburg) ..............................................


Introduction..


Materials and Methods..


Results.........


Discussion .. ................ .. .. .. .........


IV. SEASONAL INCIDENCE AND BIOLOGICAL CONTROL POTENTIAL
OF Aprostocetus hagenowli (Ratzeburg) IN TREEHOLE
MICROHABITATS ..............................................................................................


Introduction ...................................................
M materials and M methods ...................................
R e su lts.... ................... .................. .. .. ..... ......


* 4 4 4 4 4 4 4 4 4
* 4 444 444 4*~~~~~* ~4*
* 4 4 4 4 4 4 4 4 4 4 4


Discussion ............................... ..........ssion ...................................... ................... ..... ....... ............................... ..


Pane







V. A KAIROMONE FOR Aprostocetus hagenowii (Ratzeburg), A EULOPHID
PARASITOID OF AMERICAN COCKROACH (Periplaneta americana (L.))
OOTHECAE .........................................................................................................


Introduction..


Materials and Methods..
R results ..........................
D discussion ....................


111111111111(1(1(1(1(111111)111111111111 63
70
1)111111111.)1)1)111111(1(111111*((1(~~1 74


REFERENCES CITED ................... ................... ................... ..........


BIOGRAPHICAL SKETCI-I. ................... ................... ................... ..........











Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

CHEMICAL ECOLOGY AND BIOLOGICAL CONTROL POTENTIAL OF THE
COCKROACH OOTHECAL PARASITOID Aprostocetus hagenowii (Ratzeburg)

By

Daniel Robert Suiter


August 1994

Chairman: Philip G. Koehler
Major Department: Entomology and Nematology


A simplified method of mass producing a female-biased colony of


Aorostocetus


hagenowii (Ratzeburg) on Periplaneta americana (L.) oothecae was developed. Forty-five

percent of the 24 oothecae stung per day produced a clutch of parasitoids, 77% of which

were female-biased clutches; six (range 0-18) female-biased clutches emerged per day. A

significantly higher percentage of 0-1- (90%) and 2-3-wk-old (87%) oothecae produced a


female-biased clutch than 4-5-wk-old (53%) oothecae. Development in


old oothecae was significantly shorter, by


1-wk-old oothecae (35 d).


and 4-5-wk-


and 4 d, respectively, than development in 0-


The number of female (51) and male (7) progeny per clutch,


and sex ratio (14% male) did not differ significantly among the three ages of oothecae.

Oothecae were irradiated or frozen to kill cockroaches before parasitism. A

significantly higher percentage of irradiated-killed oothecae (100%) produced a female-






biased clutch than frozen-killed oothecae (47%), but per clutch, the number of female

progeny (67) was not significantly different; irradiation dose did not significantly effect


development of the parasitoid.


Weekly for 30 wk, parasitoids were released in five


cockroach-infested treeholes (20,000 females per treehole total). Parasitism of sentinel

oothecae in release treeholes was 70-100% during release weeks 6 through 15, and 50%-

80 during four additional weeks. Parasitism in control treeholes was 10-40%, but peaked


at 46% the week of July


7; winter parasitism was <10%. Parasitism of oothecae in release


treeholes remained higher for 10 of the 1


wk after parasitoid releases were terminated,


when sentinel surveys were ended.

Chemically mediated host location was investigated in Y-tube bioassays. Nearly

80% of female parasitoids responded to an ootheca after 10 minutes. Bioassay of one

ootheca equivalent from AgNO3-impregnated silica gel identified the attractant as 6,9-


heptacosadiene.


This hydrocarbon is 37 times more abundant in frass than on the ootheca;


it is the predominant hydrocarbon in frass, oothecae, and adult female cuticle. Its volatility

from oothecae was demonstrated by trapping it on Super Q adsorbent, and its identity

confirmed by comparison of its mass spectra with spectra from oothecal-6,9-

heptacosadiene obtained by column chromatography. Coinjection of the volatile with 6,9-

heptacosadiene resulted in a single GC peak at the retention time of 6,9-heptacosadiene.











CHAPTER I
INTRODUCTORY STATEMENT

In recent years the structural pest control industry has undergone a rather

significant change in pest control philosophy brought about, and paralleled by, the

environmental movement in this country. Pest control professionals have experienced a

strong increase in environmental legislation and regulation of their industry. These

stepped-up pressures are a direct result of the public's increased awareness of the potential

hazards of pesticide usage, especially in the urban environment where pesticides are

purposefully applied to areas in direct contact with human inhabitants; by their very

nature, application of pesticides in agroecosystems limits exposure to the general public.

As a result of such regulatory pressure, the structural pest control industry is adopting less

toxic, exposure-minimizing technologies for controlling urban pests. For example, the sale

of child-safe bait stations to pest control companies has skyrocketed in the past decade,

and more than doubled since 1986. The Age-of-Reduced-Exposure is also evidenced by an

increase in the number of professional pest control companies now providing once a year,

or as needed treatments.

The use of multiple control strategies to manage a pest population is often called

integrated pest management (IPM), a term becoming more and more familiar to the

structural pest control industry. IPM involves thorough inspection and diagnosis of the

pest problem, followed by recommendation and implementation of strategies to manage






2

the pest population; management decisions are often based on population dynamics and

biology of the pest. The use of toxic baits, inorganics, such as boric acid, and insect

growth regulators, such as hydroprene, methoprene and fenoxycarb, are all alternatives to

traditional pest control via application of organophosphate, carbamate, and pyrethroid

residuals, and are gaining widespread acceptance for use in IPM programs for structural

pests. Biological control of urban pests is yet another technique which warrants

investigation as a component in IPM programs for certain urban pests. Because of the


efficacious history of chemical pesticides, however, biological


control has really never


been given serious consideration, until quite recently.

There are several forms in which biological control can be achieved; the first is


classical biological control.


The target pest of a classical biological control program is


usually an accidentally introduced species, as are most pest cockroach species in the U.S.

When an organism is accidentally introduced, its primary parasites, predators and

pathogens are often left behind in its native range. In its new environment the introduced

species may near its biotic potential, limited mainly by abiotic factors and, in a few cases,


newly-acquired natural enemies in its recently expanded range.


The goal of classical


biological control is simple; reacquaint the newly-introduced pest with natural enemies

that held its numbers at equilibrium in its natural range. Classical biological control

involves the exploration of a pest's native range for natural enemies that could be

imported, quarantined, researched, and developed for biological control. Candidate

material must be returned to the area of pest infestation, placed in Quarantine and data






3

organisms; data gathering of this magnitude is costly, and often requires many years.

Another way to achieve biological control is through the augmentation of natural enemies

in the field by laboratory rearing and subsequent periodic release.

One problem arises with the concept of biological control in the urban

environment. A difference between agricultural and urban pest control involves the

concept of economic (agriculture) and aesthetic (urban) injury levels. In agriculture, a

certain level of pest infestation is tolerated, and for biological control purposes is even

desired. The level of the pest is held below the density which might cause economic loss.

When the pest level reaches some threshold, action is taken to reduce its level. In urban

entomology, where aesthetics, as opposed to crops or livestock, are being protected, the

level of infestation tolerated is often one pest organism. Urban biological control is not

being touted as a single control technique, but a technique to alleviate the reliance on

pesticides.

Successful biological control programs for urban pests need to be efficacious and

economically competitive with synthetic pesticides, or they will not survive in the

marketplace. Market competitiveness will require that several criteria be met by the


candidate product.


Those criteria will include ease of application, efficacy, safety,


production costs, and shelf life; biological control-based products must at least compete

with sprayable pesticides on all these criteria. Several corporations have demonstrated the

feasibility of developing biological control products, and have recently gone to market

with them. Bio-Path, a bait station device, is loaded with a naturally occurring, soil-based






4

infected with fungal conidia, crawl away, become diseased, and die. This product has

recently made its debut in the structural pest control market. The nematode-based product

Vector was recently released for the outdoor application to areas infested with fleas.

BioSafe is a nematode based lawn and garden insect control product which is applied to

the soil for control of cutworms, grubs, and other soil-dwelling insects. In each jug of

BioSafe, 10 million nematodes are embedded in a gelatinous sheet and placed on a small


piece of screen.


Water is added, and after a 10 minute wait the product is ready to apply.


BioSafe has a shelf life of 6 months.


The nematode used in BioSafe, Steinernema


carpocapsae, has also been shown in field tests to infect and kill Australian and

Smokybrown cockroaches in infested palm trees (D.R.S. unpublished).

The cockroach oothecal parasitoid Aprostocetus hagenowii (Ratzeburg) has some

potential for the biological control of peridomestic cockroaches. A biological control

program involving this parasitoid would combine periodic releases of it with a bait

application to cockroach harborages. Baits will undoubtedly reduce cockroach

populations, but unhatched oothecae, of which there may be hundreds, are unaffected by

the bait and remain as a source of reinfestation. The use of biological controls for urban

pests should be thought of as a preventative component in a multiple component

integrated management system. A. hagenowii may help fill part of this niche, but a

concerted effort must be put forth for its successful development.

With the continued regulation of the urban pest control industry, resulting in

inrrpacpd FPA reoictratinn rp\vntinrne it ic inevitahIl that hinlnoirsil cnntrnl nrrtipet will1






5

techniques will never replace chemical control techniques in urban pest management, but

they have a future as a component in the integrated management of some urban pests.

Decades of liberal pesticide use has led to several hard lessons; environmental insult,

endangerment of human health, and irreversible pesticide resistance leading to cycles of

pesticide development are but a few. In the urban arena new pesticides (e.g. active

ingredients) are not being developed. Development of a new active ingredient can often

cost $20 to $40 million, so manufacturers are resorting to new formulations of the same

actives. Biological control is safe, can be effective, and is often implemented for a fraction

of the cost of new pesticide development.











CHAPTER II
LITERATURE REVIEW

The Parasitoid


Aprostocetus hagenowii (Ratzeburg) (Hymenoptera: Eulophidae) is a small (2

mm), gregarious parasitoid of cockroach oothecae of the family Blattidae (Roth & Willis

1954, 1960, Cameron 1955, Edmunds 1955, Kanayama et al. 1975, Narasimham 1984,


Kumarasinghe & Edirisinghe 1987


1991). Host records include the oothecae of


Periplaneta americana (L), P. australasiae (F.), P. fuliginosa (Serville), P. japonica Karny,

Blatta orientalis L., Eurycotis biolleyi Rehn, and E. floridana (Walker).


Distribution.


The distribution of A. hagenowii is world-wide; its reported


distribution is primarily from mild climates (10-30N latitude), reflecting the distribution of

its hosts. Cameron (1955) received parasitized oothecae from Trinidad, Saudi Arabia, and

India, while Burks (1943) reported this parasitoid from the eastern and southern U.S., the

West Indies, and the Hawaiian Islands. Ceianu (1986) reported A. hagenowii from B.

orientalis in Romania (45N latitude). Surveys of cockroach oothecal parasitoids in India

uncovered A. hagenowii-parasitized P. americana, P. australasiae, and P. brunnea


(Narasimham & Sankaran 1979, Kumarasinghe & Edirisinghe 1991).


Vargas and Fallas


(1974) discovered A. hagenowii parasitizing the oothecae of P. australasiae in San Jose,

Costa Rica. Kanayama et al. (1975) and Takagi (1975) found A. hagenowii-parasitized P.

fuliginosa oothecae in Japan.





7

Morphology/sexual differences. According to Edmunds (1955) and Ali and Islam

(1983) male A. hagenowii are smaller than females and possess 9- or 10-segmented

antennae which bear long setae; the antennae of females are 12-segmented, and do not

bear long setae. Barlin et al. (1981) used scanning electron microscopy to document

internal and external sensory structures on the antennae of male and female A. hagenowii.

Among the many types of sensillae, a multiporous plate sensilla (types 1 and 2, organized

in an alternating ring) in females seems the most likely candidate for host perception

through chemoreception, probably olfaction with the abundance of pores, pore tubules,


and dendrites in the antenna. Type 1 plate sensillae are more porous than type


2. Other


types of sensillae found on the female antenna were (1) nonporous, socketed hairs, (2)

multiporous, non-socketed hairs, and (3) multiporous pegs.

Mating. Mating in A. hagenowii occurred immediately after parasitoids emerged


from the ootheca (Cameron 1955, Edmunds 1955,


Vargas & Fallas 1974, Takahashi &


Sugai 1982, Jie & Wenqing 1984). Takahashi and Sugai (1982) reported that males

approached females after receiving sex pheromone, and exhibited a characteristic zig-zag

walk with intermittent wing vibrations. Males exhibited violent wing-flapping and a quick,

agitated gait while searching for females. Males then mounted females and began antennal

stimulation; females responded to this by lifting their abdomen to initiate copulation, which

took place following a backward movement of the male. Ether washes of female A.

hagenowli elicited male zig-zag patterned movements.

Oviposition behavior. Cameron (1955) reported that oviposition began soon after






8

(1974), and Narasimham (1984) described the oviposition behavior of A. hagenowii;

females mounted and drummed the ootheca with their antennae, and occasionally stopped

to groom their antennae or tap the ootheca with the tip of their abdomen (see Roth &

Willis 1954, Plate 1A). After a suitable location on the ootheca was found, the female

unsheathed her ovipositor and began to drill a hole through the ootheca wall. Edmunds

(1955) reported that parasitoid eggs hatched in 24 h, but Narasimham (1984) reported an

egg-hatching time of 48-54 h in P. americana oothecae. After oviposition, the female

sometimes fed from the oviposition wound (Roth & Willis 1954, Plate ID).

Narasimham (1984) studied the effect of superparasitism on developing broods. As

the number of females allowed to concurrently oviposit into an ootheca increased, the

number of emergent progeny per ootheca increased, as expected, but survival (i.e. number

of emergent progeny divided by the estimated number of eggs deposited into an ootheca

by a female) and developmental time decreased. In addition, the sex ratio increased (i.e.

greater proportion of males). Emergent female parasitoids from superparasitized oothecae

were also smaller, and their lifespan was shortened in comparison to female progeny from

oothecae that were not superparasitized.


Although A. hagenowii parasitized oothecae of all ages, ca.


14-d-old oothecae


were preferred. Roth and Willis (1954) demonstrated parasitism of P. americana oothecae


up to 29-d-old,


but not oothecae 35-d-old (i.e. just before hatch); there was some


preference for 12 to 14-d-old oothecae. Narasimham (1984) demonstrated parasitoid

nreferencr.e fnr 11 tn 1 5 d-nld P americana nnthecae. hut 30-d-old onthecae were also






9

P. americana oothecae to A. hagenowii; there were few differences in the proportion of

oothecae parasitized, but development (i.e. oviposited egg to emergent adult) in 14-d-old

oothecae was significantly shorter than in the four other ages. Kumarasinghe and

Edirisinghe (1987) reported A. hagenowii oviposition in 0 to 34-d-old P. americana

oothecae, but development only in 0 to 28-d-old oothecae.

Oviposition frequency. During her lifetime, a female will parasitize only one or two

P. americana oothecae (Narasimham 1984, Hagenbuch et al. 1988, Heitmans et al. 1992).

Larger females contained more eggs than smaller females. Large females distributed their

eggs in two oothecae, but smaller females tended to parasitize only one ootheca (Heitmans


et al. 1992). A majority of eggs were deposited during the first


d postemergence (Roth


& Willis 1954). Jie and Wenqing (1984) reported peak parasitoid activity during the first


d postemergence.


Development. Developmental time (i.e. oviposited egg to emergent adult) of A.

hagenowii is correlated (negative), as expected, with temperature; typical developmental

time is ca. 30 days at 30C. At 16-18C, Cameron (1955) reported a developmental time

of ca. 90 d in P. americana oothecae, but at 23C development was completed in 24 d.

Narasimham (1984) reported parasitoid development in P. americana oothecae of 61,


45.4, and 33.5 d at 15,


and 30C, respectively. In mass-rearing, Hagenbuch et al.


(1988) reported a developmental time in P. americana oothecae of 36 d at 27C.

Cameron (1955), Edmunds (1955), and Narasimham (1984) provided detailed

descriptions of immature (eggs and larvae) development (including sketches) in A.







(Edmunds 1955).


There were 3 larval instars. The second and third instars lasted 4-5 and


10-22 d, respectively; the prepupal stage lasted 1-2 d, and at 18-21 C adult emergence

occurred after ca. 10-22 d as exarate pupae. Roth and Willis (1954) provided close-up

photographs of developing parasitoids within cockroach oothecae. Edmunds (1955)

reported that larval feeding was completed in 18-25 d, and that there was an inverse

relationship between brood size and developmental time. Edmunds (1955), Roth and

Willis (1954), and Takagi (1975) demonstrated an inverse relationship between brood size

and parasitoid size, probably due to limiting resources within the ootheca.


ratio.


sex ratio of A. hagenowii is strongly female-biased. Roth and Willis


(1954) and Cameron (1955) reported


sex ratios of ca. 12-33 and 10% male in P.


americana oothecae, respectively. Harlan and Kramer (1981) reported a


sex ratio from P.


americana and B. orientalis oothecae of 25% male, and Vargas and Fallas (1974) reported

a sex ratio from P. australasiae oothecae of 38% male, and from E. biolleyi 32% male. In


P. fuliginosa oothecae, sex r

1975, Jie & Wenqing 1984).


atios of 17 and 29% male were reported (Kanayama et al.


ratio was influenced by host size and species


(Narasimham 1984). Edmunds (1955) reported that unmated female parasitoids produced


all-male


clutches.


Rearing/diet/longevity. Unfed A. hagenowii adults typically live less than 10 d;


adult lifespan can be extended


, however, by provision of a carbohydrate diet to adult


parasitoids. Male A. hagenowii provided raisins as a food source lived from 9-23 d (mean

15); females lived 19-45 d (mean 25). Unfed males lived 1.7 d, while unfed females lived








survival of 6.33 (range 4-11) and 14.5 d (12-37 d), respectively. Parasitoids survived at

18C with water and sugar for 2-6 wk (Cameron 1955). Jie and Wenqing (1984) reported

adult longevities of 3-47 and 2-40 d for females and males, respectively.

Narasimham (1984) tested several diets (honey, flower mucilage, and sugar) for

their ability to lengthen adult life. A honey diet was superior, as adult females and males


lived a mean of 9 and


d, respectively, which was 6.2 and 3.5 d longer than starved


individuals. Hagenbuch et al. (1988) described a rearing protocol for A. hagenowii and its

host, P. americana oothecae. Parasitoids in their study were fed a honey diet, and lived 2-4


d longer than starved parasitoids.


Without exception, in all studies, males lived a shorter


time than females.

Parasitism in the field. Surveys of naturally-deposited oothecae suggest that A.

hagenowii is of considerable importance in the natural control of Blattid cockroach pests.

In Trinidad, Saudi Arabia, and India 15-57% of field-collected P. americana and P.


australasiae oothecae contained A. hagenowii (Cameron 1955,


Narasimham & Sankaran


1979). In another survey, parasitism by A. hagenowii in India revealed rates as high as


and 100% for P. americana and P. australasiae, respectively (Kumarasinghe & Edirisinghe

1991). In San Jose, Costa Rica, 33 and 52% of P. australasiae oothecae from two

locations were parasitized by A. hagenowii (Vargas & Fallas 1974). In China, 0-93% of P.

fuliginosa oothecae were parasitized by A. hagenowii (Jie & Wenqing 1984); parasitoid

activity began in April, and peak field populations were found from May to August. In

Japan, parasitism among oothecae collected from pig-pens and animal houses was between





12

(Kanayama et al. 1975). In bird houses, 50% of P. americana and P. australasiae oothecae

were parasitized (Hagenbuch et al. 1988).

Field tests. Purposeful release of A. hagenowii for control of Blattid cockroach


species has had varying levels of success.


Typically, researchers evaluate their release


program by monitoring sentinel ootheca parasitism. Roth and Willis (1954) released A.

hagenowii in rooms with sentinel P. americana oothecae distributed throughout. A release


rate of 10 female parasitoids per sentinel ootheca resulted in ca.


75-85% parasitism. In


experimental kitchens artificially infested with P. americana adults, Hagenbuch et al.


(1989) released A. hagenowii and achieved 96% parasitism of viable oothecae after


7 wk.


During the 8-11 wk period, parasitism rose to 98%, and at 12-17 wk was still above 95%.

Parasitism of ceiling-deposited oothecae was significantly lower than in other areas. In

their study, parasitoid release chambers accidentally allowed parasitoids to escape and

enter control chambers holding P. americana oothecae; here parasitism was 46% after 7-

11 wk, and 79.4% after 12-17 wk. Narasimham (1984) reported a preference of this

parasitoid for warm, humid microclimates much like that of its hosts. In the same study,

parasitoid activity was greatest at low humidities and high temperatures (35C optimum).

Kairomones


Kairomones are interspecific semiochemicals (allelochemics) which when

produced, acquired by, or released as a result of the activities of an organism, which, when

it contacts an individual of another species in the natural context, evokes in the receiver a

behavioral or physiological reaction adaptively favorable to the receiver but not to the





13

Greany et al. (1984) provided excellent overviews of kairomone-dependent host selection

by parasitoids, and list the following steps: (a) habitat location, including general habitat

location, and potential host community (microhabitat) location, (b) host location, (c) host

acceptance, (d) host suitability, and (e) host regulation. The "steps" involved in the host

selection process are meant to be a framework only, and are only sometimes mutually

exclusive; depending on the case, they may or may not all be present, and are often only

vaguely separable. Host location and acceptance are the crux of the current research; the

remainder of this review shall center on the abundance of primary literature involved in

location and acceptance of hosts by parasitoids.

Host location, or finding, has been reviewed by several authors (Vinson 1976,


Weseloh 1981), and is rather difficult to define.


Weseloh (1981) defined chemical host


location as the discovery, or perception, of a host from a distance due to short- (primarily)

and long-range (rare) volatile compounds produced by the host, its activity, or one of its

products. The cues responsible for discovery of the host can be physical (auditory, visual),

chemical, or combinations of each.

Weseloh (1981) arbitrarily divided host location into long- (volatile) and short-

(low volatility and tactile compounds) range chemoreception, and provided examples of

parasitoids that locate hosts in these manners. Long-range cues serve to attract the

parasitoid from a greater distance than short-range cues. Long-range host location cues

are not as well-documented as short-range, relatively nonvolatile cues. A classic long-


range host location cue used by parasitoids for host location is host


sex pheromone






14

Although a wealth of literature involving the tritrophic interactions of a parasitoid and the

long-range, volatile synomones produced by its host-plant complex exists, such

interactions are not considered host location, but rather host microhabitat location, and

will not be dealt with in this review. In short, parasitoids are often attracted to host

microhabitats by volatile plant chemicals released as a result of herbivore damage (Elzen et

al. 1983, 1984, 1986, Baehrecke et al. 1990, Keller 1990, Turlings et al. 1990, 1991a,b,


McAuslane et al.


1991a,b).


Host location via detection of short-range kairomones has been well-documented


and provides most of the host location literature.


To utilize short-range kairomones as


host location cues, parasitoids must come in close proximity to these low to nonvolatile

compounds; the compounds must often be contacted. Commonly, these kairomones are

found on or emanating from host-associated products or the host itself. Short-range

kairomones commonly bring about a change in the behavior of the parasitoid when they

are contacted, often releasing unique searching behaviors in the parasitoid that result in

increased, intensive host seeking in a localized area.

After location of a host, the parasitoid will examine it, and either accept or reject it

based on certain chemical and physical criteria. Arthur (1981) and Vinson (1976) reviewed

chemical and physical stimuli utilized by parasitoids to examine hosts. Host examination

leading to acceptance or rejection commonly involves antenna drumming behaviors

mediated by nonvolatile compounds. Nonchemical examination cues leading to host

acceptance or rejection include host size, shape, texture and color (Vinson 1976, Arthur






15

acceptance/rejection is based on physical contact of the parasitoid with the host, thus

chemical cues are almost always large, nonvolatile compounds.

In relation to the current research, there are a few host-related products of

particular interest. Host frass is one of the most common materials utilized by parasitoids


to locate hosts (Jones et al. 1971, Lewis & Jones 1971,


Vinson et al. 1976, Henson et al.


1977

1985


, Nordlund & Sauls 1981,


Weseloh 1981, Lewis & Nordlund 1984, Dmoch et al.


, Nordlund & Lewis 1985, Takabayashi et al. 1985,


Van Leerdam et al. 1985,


Clement et al.


1986, Drost et al. 1986, Nemoto et al. 1987a,b, Eller et al. 1988, Herard et


al. 1988a,b, Lewis & Tumlinson 1988, Auger et al. 1989, Fukushima et al. 1989,


Takabayashi & Takahashi 1989, McAuslane et al.


1990, Lewis et al. 1991,


Ma et al.


1992). Both volatile and contact host location kairomones have been identified from host

frass. Parasitoids characteristically antennate host frass and alter their search behavior

when they contact the frass.

Chemical products of host glands (e.g. accessory, mandibular) are also often used


by parasitoids to locate hosts (Vinson et al. 1975, Nordlund et al.

Corbet 1982, Lewis & Nordlund 1984, Mudd et al. 1984, Van Di


1977a,b, Mudd &


riesche & Hulbert 1984,


Muzaffar & Inayatullah 1986,


Vinson & Piper 1986, Cave et al. 1987


Nordlund et al.


1987). Egg parasitoids, especially members of the Trichogrammatidae, are known to use


host location kairomones from the


hosts (Vinson 1975, Nordlund et al. 1977a,b,


Leonard et al.


1987


, Renou et al. 1989, Shu & Jones 1989). A variety of other host and


host-related products have been reported to serve as host location and acceptance






16

1985, Hagvar & Hofsvang 1989), saliva (Jones et al. 1971, Lewis & Jones 1971, Clement


et al. 1986,


Ma et al. 1992), silk & cocoons (Dmoch et al. 1985,


Weseloh 1987


Ma et al.


1992), larval trail pheromones (Howard & Flinn 1990), epideictic pheromones (Prokopy


& Webster 1978), pupa/larva/exuvium/cocoon (Leonard et al. 1975,


Vinson et al. 1976,


Dmoch et al. 1985, Takabayashi et al. 1985, Takabayashi & Takahashi 1986, Herard et al.

1988a,b, Carde & Lee 1989, Takahashi et al. 1990, Ma et al. 1992), and wing scales


(Lewis et al.


1972, Jones et al. 1973,


Vinson 1975, Nordlund et al.


1976, Gueldner et al.


1984, Noldus


van Lenteren 1985a,b, Chiri & Legner 1986, Zaborski et al.


1987


Thomson & Stinner 1988, 1990, Shu


Jones 1989, Shu et al. 1990).











CHAPTER III
A SIMPLE REARING METHOD FOR THE COCKROACH OOTHECAL
PARASITOID Aprostocetus hagenowii (Ratzeburg)

Introduction


Aprostocetus hagenowii (Ratzeburg) is a small, gregarious, Eulophid parasitoid of

oothecae from cockroaches in the family Blattidae. Its biology is well documented (Roth

& Willis 1954, 1960, Cameron 1955, Edmunds 1955, Narasimham 1984). Host records

include oothecae of these cockroaches: Periplaneta americana (L.) (American), P.

australasiae (F.) (Australian), P. fuliginosa (Serville) (smokybrown), P. japonica Karny

(Japanese), P_. brunnea Burmeister (brown), Blatta orientalis L. (Oriental), Eurycotis

floridana (Walker) (Florida woods), and E. biolleyi Rehn. A. hagenowii may prefer P.

americana oothecae. Narasimham (1984) reported that A. hagenowii reared in P.

australasiae oothecae for three generations subsequently preferred P. americana oothecae.

Female parasitoids emerge from the ootheca with their full complement of eggs,

immediately sibmate, and can then parasitize a host; females typically deposited their eggs

in one or two oothecae (Heitmans et al. 1992). Even though A. hagenowii will parasitize


and develop in all ages of P. americana oothecae, up to ca. 28-d-old,


1 to 15-d-old


oothecae were preferred (Narasimham 1984). Hagenbuch et al. (1988) exposed five ages

(7-, 14-, 21-, 28-, and 35-d-old) of P. americana oothecae to A. hagenowii; there were





18

few differences in the proportion of oothecae parasitized, but those in 14-d-old oothecae

emerged in a significantly shorter time.

Developmental time and number of emergent progeny per ootheca varies

considerably. Narasimham (1984) reported parasitoid development, from oviposition to


adult emergence, in P. americana oothecae, in 61,


45.4, and 33.5 d at 15,


and 30C


respectively; Hagenbuch et al. (1988) reported a developmental time of 36 d at 270C.

Cameron (1955) reported 30-40 parasitoids from P. americana oothecae, and Harlan and

Kramer (1981) reported ca. 30 parasitoids emerged from P. americana oothecae. In mass-

rearing, P. americana oothecae produced ca. 56 parasitoids per ootheca (Hagenbuch et al.

1988).


sex ratio of A. hagenowii is strongly female-biased. Cameron (1955) reported


sex ratios of ca. 10 to 20% male, and Roth and Willis (1954) and Harlan and Kramer


(1981) reported


sex ratios of 11-33% and 25%, respectively. Edmunds (1955) reported


that nonmated female parasitoids deposited all-male clutches. Narasimham (1984)


reported that


sex ratio was influenced by host size and species.


Narasimham (1984) and Edmunds (1955) studied the effect of superparasitism by

A. hagenowii. As the number of females allowed to concurrently oviposit into an ootheca

increased, the number of emergent progeny per ootheca increased, as expected, but

survival (i.e. number of emergent progeny divided by the estimated number of eggs

deposited into an ootheca by a female) and developmental time decreased; indeed, there

was an inverse relationship between the number of emergent progeny and the





19

the number of concurrently ovipositing females. Emergent female parasitoids from

superparasitized oothecae were also smaller, and their lifespan was shortened in

comparison to female progeny from oothecae that were not superparasitized.

It is not known whether this parasitoid takes carbohydrate in the field. However,

carbohydrate diets (raisins, honey, sugar, flower mucilage) prolonged A. hagenowii's life

by ca. 2-6 d (Cameron 1955, Edmunds 1955, Narasimham 1984, Hagenbuch et al. 1988);

unfed parasitoids typically lived 2-4 d. A. hagenowii also fed from the sting wound after

oviposition (Roth & Willis 1954, plate ID). Because A. hagenowii was most active and


deposited the majority of its


eggs


during the first


2-3 d post-emergence (Roth & Willis


1954, Jie & Wenqing 1984), the practicality of prolonging this parasitoids life with an

artificial carbohydrate diet is questionable. The purpose of this study was to develop a


large-scale,


sex ratio-efficient (i.e. female-biased) rearing method for A. hagenowii.


Materials and Methods


Parasitoid. A. hagenowii was colonized in field-collected P. americana oothecae in

the spring of 1993 from 10-20 adult female parasitoids collected at the Swine Research

Unit on the University of Florida campus, Gainesville, Florida. A sample of parasitoids

was sent to the U.S. National Museum in Washington, D.C., and identified as A.

hagenowii by Dr. Mike Schauff. The colony was supplemented throughout the following

year with parasitoids from sentinel P. americana oothecae placed in treehole microhabitats

to monitor the natural population dynamic of A. hagenowii (Chapter IV).

Host production. P. americana oothecae were mass-reared in ootheca production





20

each of eight galvanized steel tubs (59 liter) was removed and replaced with screen (0.65


by 0.65 cm mesh).


The screened-bottom of each tub was then placed into an ootheca


collection tray (43 by 43 by 9 cm); the floor of each tray was lined with brown

construction paper onto which oothecae fell. The top rim of each tub had a 10-cm-wide


inward projecting lip.


The inner walls of the tub and the inside ceiling of the lip were


coated with Teflon to keep cockroaches from escaping.

Cockroaches were obtained as needed from a P. americana-infested sewage

treatment plant on the University of Florida campus (see Cornwell 1968 Fig. 115). One

hundred to 300 mixed sex adult cockroaches were placed in each tub and provided water


from a chick waterer, seven polyvinylchloride harborages (


cm i.d. by 12 cm long), and


laboratory rat chow (23% crude protein; Purina, #5001, St. Louis, MO). Dead

cockroaches were replaced, water and food replenished, and oothecae collected weekly;

time spent in colony maintenance was noted.

Parasitoid production. Daily, a cohort (some multiple of 6) of oothecae of varying,

known age was placed individually into 18-ml snap cap vials (Thornton Plastics, Salt Lake


City) each containing a single


or 3-d-old female parasitoid emerged from a known


number of laboratory- and field-stung, sentinel oothecae (see Chapter IV). Cohorts were

stored in plastic boxes (20 by 32 by 10 cm) and held at 27-30C, in ambient light and

humidity. Each d for 45 d, the number of female- and male-biased parasitoid clutches

which emerged were recorded for each cohort, and summed for all cohorts. After 45 d,

the number of oothecae producing anything other than parasitoids (i.e. (1) cockroaches or







containing


newly emerged progeny were placed daily, with snap caps intact, into a


Plexiglass cage (30 by 30 by 30 cm; Hagenbuch et al. 1988) and held for 24 h to ensure

sibmating. After 24 h the snap cap was removed from each vial, allowing parasitoids to


Water and food were provided from water and honey-water soaked cotton balls


placed individually into souffle cups. Parasitoids were held for an additional


d, and then


killed by holding the cage at 4C for 24 h.

Parasitoid development in aged oothecae. The development of A. hagenowii in 1

to 8-, 15 to 21-, and 30 to 36-d-old (fixed qualitative factor, 3 levels) pristine (i.e. not

dented, and without attached detritus) P. americana oothecae was investigated in a

randomized complete block design (RCBD) with seven clutches (extraneous factor) of

parasitoids serving as blocks; parasitoids from each clutch were the progeny of a single

female. A RCBD was used because females within a clutch (i.e. sisters) are more similar

genetically than females between clutches; interclutch genetic variability among females

was assumed beforehand to contribute to variability in response variables.

For each clutch, ten oothecae of each age were randomly selected and placed

individually into 18-ml snap cap vials containing a single, randomly selected 2-d-old, unfed

female parasitoid. Oothecae were incubated at 30C under natural lighting until

cockroaches or parasitoids emerged, at which time cockroaches were discarded, but

parasitoids were stored in 70% isopropyl alcohol until they could be sexed and counted.

Response variables from oothecae that produced a female-biased clutch of parasitoids

were (1) number of emergent female progeny per clutch, (2) number of emergent male





22

initiation until the day of progeny emergence. If neither cockroaches nor parasitoids had

emerged after 45 d, oothecae were counted and discarded. For each age, across all

clutches, was recorded the fate of each ootheca. Each ootheca could have produced (1) a

female-biased clutch of parasitoids, or (2) a male-biased clutch of parasitoids, or (3)

cockroaches, or (4) neither parasitoids nor cockroaches.

Descriptive statistics. Data for parasitoid and host colonies were summarized with

the descriptive statistics mean, standard deviation, mode, highest value, 90, 75, 50

(=median), and 25th percentiles, and the lowest value. A percentile is the percentage of

values in the observed data set which are lower than the data value reported.

Categorical data analysis. Differences in proportions among the three ages of

oothecae (rows) for each of the four responses (columns), ((1) number of oothecae from

which emerged a female-biased clutch, or (2) number from which emerged a male-biased

clutch, or (3) number from which emerged cockroaches, or (4) number from which

emerged neither parasitoids nor cockroaches) were analyzed in a 3 by 4 contingency table

with a Chi-Square Homogeneity of Proportions test (Ott 1988, p. 253) since row marginal

totals were fixed (n=70) in advance. Proportions among ootheca ages (i.e. within each

column) were separated by noninclusion of zero in a 95% confidence interval constructed

on the difference in proportions between oothecal ages (i.e. pairwise comparisons of


ootheca age proportions within each column) (Hochberg & Tamhane 1987

formula).


Equal variance assumption and analysis of variance.


The assumption of equality of





23

oothecal ages in the RCBD experiment was tested with Levene's test (Levene 1960,

Conover et al. 1981) for the following six response variables: (1) number of female

progeny per clutch, (2) number of male progeny per clutch, (3) the arcsine of the square

root of the sex ratio, (4) developmental time, (5) the square root of developmental time,

and (6) the log of developmental time. In Levene's test, the mean of each sample is

subtracted from each data point within the sample, and an analysis of variance (PROC

GLM in SAS) conducted on the absolute values of these differences (SAS Institute 1988).

In the RCBD experiment, means were computed for each ootheca age across all clutches.


The corresponding SAS MODEL for Levene's test was Y= age.


The prechosen alpha level


of the test was 0.10. Levene's test indicated non-significantly different variances among

ootheca ages for the number of female (F=0.00, df=2,161, P=0.9986) and male (F=0.99,

df=2,161, P=.3746) progeny per clutch, and the arcsine of the square root of the sex ratio

(F=1.05, df=2,161, P=.3517). Subsequently, these variables were analyzed by analysis of


variance with the corresponding SAS MODEL for a RCBD,


Y= age clutch


both


hypotheses were tested with the mean square for error at alpha=.05.

Nonparametric analysis. Developmental time (F=15.28, df=-2,161, P=.0001), the

square root of developmental time (F= 15.19, df=2,161, P=.0001), and the log of

developmental time (F=14.65, df=2,161, P=.0001) grossly violated the assumption of


equal variance.


Therefore, developmental time was alternatively analyzed with a Friedman-


type test (Mack & Skillings 1980, Skillings & Mack 1981); in this test, a Kruskal-Wallis


procedure (PROC NPARIWAY in SAS) is conducted on each block, yielding a





24

blocks and compared to a critical X2 value at the summed degrees of freedom and

prechosen alpha level (.05) of the test. Median developmental times were separated by

pairwise comparisons using the same technique at alpha=.01.

Results

Approximately 24, 2-wk-old oothecae were stung per d (Table 3.1). Although the

largest number of oothecae stung per d was 72, the interquartile range (i.e. 75th

percentile-25th percentile) was 6 oothecae. Likewise, the oldest oothecae stung were ca.

34-d-old, but the interquartile range was ca. 10 d. Daily cohorts of oothecae were stung


from


2- and 3-d-old female parasitoids that had hatched from ca. 12 laboratory-reared


oothecae plus a few field-stung oothecae; parasitoids from field-stung oothecae were not

available daily, as indicated by a mode of zero, but the colony was infused with field genes

whenever possible.

From the 176 daily cohorts of oothecae from which data were gathered, 54.9% of


them either produced cockroaches or nothing; female-biased clutches accounted for


77%


of the oothecae which produced a clutch of parasitoids (Table 3.2). Production of male-

biased clutches, the result of nonmated females, was ca. 10% from each cohort, while ca.

20% of oothecae in each cohort hatched neither parasitoids nor cockroaches.

For all cohorts, there were nearly 6 (mean=5.6) female-biased clutches produced


per d compared to only ca.


male-biased clutches; the largest number of female-biased


clutches produced per d was 18, but the most common (=mode) was 5. About 30 minutes

per d was obligated to parasitoid colony maintenance (Table 3.3).





25

Weekly, several hundred (mean=464, median=410) P. americana oothecae were

harvested for parasitoid colony maintenance, and laboratory and field experiments; the

largest number harvested in a wk was 1,070 (Table 3.3). Typical host colony working time

was 2-3 h, once a wk.

In the RCBD experiment, a significantly greater percentage of 1 to 8- and 15 to

21-d-old oothecae produced a female-biased clutch than the 30 to 36-d-old oothecae (Fig.


3.1).


This was a result of A. hagenowii's reduced ability to develop in 30 to 36-d-old


oothecae, as evidenced by a significantly greater percentage of 30 to 36-d-old oothecae

producing cockroaches than the younger oothecae.

Analysis of variance among the oothecal ages indicated no difference in the

number of females (P=.0997), males (P=.5850), and sex ratio (P=.3405) per clutch

(Tables 3.4 and 3.5). By pooling all data (n=164 clutches), the mean (95% confidence


interval) number of female and male progeny per clutch, and


(48.7-53.5), 9.0 males (7.5-10.4), and


sex ratio, was 51.1 females


.139 male (.121-.158), respectively. Parasitoid


developmental time, however, differed significantly among the 3 ages of oothecae

(P<.005); developmental in 15 to 21- and 30 to 36-d-old oothecae was 3 and 4 d shorter,


respectively, than developmental time in


to 8-d-old oothecae (Table


Discussion


The only published attempt at large scale production of A. hagenowii was by

Hagenbuch et al. (1988). One of the most important factors limiting large scale production

of this parasitoid is production of host oothecae. Hagenbuch et al. (1988) produced ca.






26

in the present study, weekly oothecal production averaged 464 (range 51-1,070) (Table

3.3). Indeed, oothecal production in this study was limited by colony space, rather than

numbers of cockroaches.

Research on artificial diets for rearing larval parasitoids is greatly needed. Because

of the physical, visual and chemical complexity in the host selection process by parasitoids

(Vinson 1976), as well as the physiological and immunological constraints within the host

after oviposition, studies on artificial diets necessarily involve a thorough understanding of

the chemical nature of the host selection process. Of primary importance is the

identification of internal and external host acceptance and oviposition stimulants (Li

1990), as well as nutritional balance in the artificial diet. Because of these constraints, it is

doubtful that an artificial diet for this parasitoid will be developed in the near future.

The reproductive strategy of A. hagenowii is such that presentation of a single

foundress with a single host results in a strongly female-biased sex ratio (Table 3.5). In


this study, production of female-biased clutches was 5.6 per d (Table 3.3),


females per wk from ca.


or 1,900-2,100


168 oothecae. Additionally, ca. 86.1% of emerging parasitoids


were female. Hagenbuch et al. produced ca. 6,160 adult parasitoids per wk from ca.


oothecae, but due to their rearing procedure, where superparasitism was common, only

53% were female.

In superparasitism, the number of progeny per ootheca and sex ratio increases as

the number of foundresses per ootheca increases (Edmunds 1955, Narasimham 1984),

resulting in more parasitoids per clutch, but a higher proportion of males. Heitmans et al.






27

female progeny decreases with clutch size. Daughters from a single foundress are larger,

and live longer, than the female progeny from superparasitism. Also, egg supply in females

is strongly correlated (positive) with body size. Large females often distribute their eggs in

two hosts, with the second clutch always smaller and with a smaller probability of survival

than the first; small females typically parasitize a single host. A single, large parasitoid will

not deposit all her eggs in one host when only one host is available (Heitmans et al. 1992).

If two hosts are available she will parasitize both, but the first clutch will be larger than the

second. Collectively, the progeny from parasitism of two oothecae is greater than

parasitism of only a single ootheca.

Several problems with parasitoid rearing involving superparasitism are evident

from the viewpoint of a practical biological control program. First, superparasitism results

in an inordinate number of useless male progeny (47% male) and fewer female progeny

than the single foundress method (13.9% male). Second, female progeny from


superparasitism do not live as long as females from a single foundress.


This has


implications for augmentative field release, because small females cannot search for hosts

for as long as larger females. And last, small females have fewer eggs and typically only


parasitize one host, while larger fema


parasitize two hosts.


The literature reporting general biological parameters (e.g. clutch size, sex ratio,

developmental time) of A. hagenowii has rarely (excluding Heitmans et al. 1992) taken

into consideration interclutch genetic variability among females when using them as

mothers for stinging experiments. In the present study, the RCBD experiment was used to






28

significant clutch effect for the number of female (P=.0001) and male (P=.0001) progeny

per clutch, and the sex ratio (P=.0001) (Table 3.4). In general, the results of this study

(Table 3.5 and Fig. 3.1) agree quite well with those of other researchers (Roth & Willis


1954, Cameron 1955,


Harlan & Kramer 1981


, Narasimham 1984, Hagenbuch et al. 1988).


A. hagenowii is a parasitoid that has potential as a tool in the integrated

management of peridomestic cockroaches. In uninhabited yet sensitive environments,

augmentative releases of the parasitoid coupled with the use of toxic baits might eradicate

existing populations of cockroaches. However, such a program would require a scheduled,


dependable, and large supply of oothecal parasitoids.


The present study has shown that a


sex ratio-efficient (i.e. female-biased) colony of this parasitoid can be maintained.


























































O E
TO0


r- c)

O O

O1


C.)




Ca


Ca


0)
-c
H
.0


v







30




4) o
0>
cI)



*A -



4) -c,





en e dr n '.

CCl
IcII


-"o-o
4- 1w N 0
a 4-


4- *0

ot 0
CI* -
4- InIc
Q)- ON d f
Cl Q tfl Clru --
0I C. I I

C.) Ca




a. 0
0


a en enJ -

N'.

(4. ClU Cl -S I

































'tta
o O\


,en
oN


c


E
0
C-

O
<*-<
L.


42
0
U

-o

* f
>1
SSW
ce

T3
O
-o

4)
I-
C)


* -
4.)
O



I-
0
C


C<
i-wmE


at'


ur
" ~i
O
;re ~5:
z &
3 "
-I'cd


r



















41a


ur~


7b
-*4b


Female-biased


Male-biased


Cockroaches


Neither


Response


Fig. 3.1


Percentage of oothecae from which emerged (1) a female-biased clutch of


parasitoids, or (2) a male-biased clutch of parasitoids, or (3) cockroaches, or (4)
neither parasitoids nor cockroaches in the RCBD experiment. Chi-square Homogeneity


of Proportions test (Ott 1988, p. 253): XZ=46.5,


df=6. P<.001


For each response, bars


separated by different letters are significantly different. Proportions within each
response were separated by non-inclusion of zero in a 95% confidence interval
constructed on the difference in proportions between ootheca ages (i.e. pairwise
comparisons of age proportions within each response) (Hochberg & Tamhane 1987
p. 275 for formula).


m 1 to 8-d-old oothecae I 15-21 U 30-36


3a6a
3,,*.






33







Table 3.4. Analysis of variance for the number of female and male progeny


per clutch, and the arcsine of the square root of


sex ratio (RCBD


experiment).

Dependent Sum of Mean
Variable Source3 Squares Square F Value P Value

# Females Age 970.537 485.269 2.34 0.0997

Clutch 6645.085 1107.514 5.34 0.0001

Error 32144.152 207.382

Total 39759.774

# Males Age 69.557 34.779 0.54 0.5850

Clutch 4674.659 779.110 12.05 0.0001

Error 10020.729 64.650

Total 14764.945

Arcsine Age 0.03921 0.01960 1.08 0.3405

Clutch 1.27035 0.21173 11.72 0.0001

Error 2.80085 0.01807

Total 4.11041


aFor each analysis of variance, the degrees of freedom for Age, Clutch,


and Total are


Error,


6, 155, and 163, respectively


































































































3
Z
In
m
d,


o3J0.
0.?
~~cn


'tG)
.
len
il4


. o











CHAPTER IV
SEASONAL INCIDENCE AND BIOLOGICAL CONTROL
POTENTIAL OF Aprostocetus hagenowii (Ratzeburg)
IN TREEHOLE MICROHABITATS

Introduction


The Hymenopteran natural enemies of cockroaches were recently reviewed

(LeBeck 1991), and several show potential as candidates for biological control. One of

those natural enemies, Aprostocetus hagenowii (Ratzeburg), is a small, gregarious

Eulophid parasitoid of cockroach oothecae of the Blattidae (Roth & Willis 1954, Cameron

1955, Edmunds 1955). Piper et al. (1978) surveyed Hymenopteran oothecal parasitoids in

several Texas and Louisiana cities inhabited by Periplaneta americana (L.) (American

cockroach), P. fuliginosa (Serville) (smokybrown cockroach), P. brunnea Burmeister


(brown cockroach), and Blatta orientalis L. (Oriental cockroach). About 25% of the


oothecae collected were parasitized by A. hagenowii and Evania appendigaster (L.);

99.4% of parasitized oothecae were parasitized by A. hagenowii. This parasitoid was

present at 14 of the 17 Texas and all 4 Louisiana sites. P. americana oothecae suffered the

highest rate of parasitism (46%), even though they accounted for only 8.6% of the

oothecae sampled. P. fuliginosa made up 87% of the oothecae sampled, but suffered ca.

half the rate of parasitism of P. americana oothecae.

Fleet and Frankie (1975) surveyed rates of P. fuliginosa ootheca parasitism at


several cockroach-infested homes in Texas; rates of ootheca parasitization were






36

84.2% for oothecae collected in the outdoor urban environment. All surveys conducted in

their study were in the summer or fall of the year.

Several authors investigated the effect of indoor releases of A. hagenowii on the


rate of parasitization of sentinel oothecae.


Roth and Willis (1954) obtained 83%


parasitism of sentinel oothecae in a room (5 m by 5 m by 4 m high) when parasitoids were

released at a rate of 10 females per sentinel ootheca. Hagenbuch et al. (1989) released ca.


600 parasitoids per wk in experimental chambers (2.4 m by 2.4 m by


artificially infested with adult P. americana; after


2.8 m high)


7 wk, parasitism of naturally-deposited


oothecae was >95%. The resulting sex ratio was 1 male to 0.71 females, indicating severe

superparasitism when the release rate was 8.3 female parasitoids per host. In indoor

residences, Piper and Frankie (1978) attained maximum rates of P. americana ootheca

parasitization of 57-62% with releases of 8 and 12 female parasitoids per ootheca. In


plumbing chases, Pawson and Gold (1993) obtained a parasitism rate of 30.6% (range


39%) of sentinel P. americana oothecae with as few as 300 (275-415) female parasitoids

released per wk from parasitized P. americana oothecae; 300 (229-355) parasitoids

released every 2 wk resulted in ca. 18.2% sentinel ootheca parasitism (range 2-44%). A.

hagenowii found sentinel oothecae placed in all locations within the plumbing chase;

sentinel oothecae placed 16 m from the release site were parasitized.

The four studies described above indicate that A. hagenowii indeed has some

potential as a biological control organism of peridomestic cockroaches. However, none of

i-\^^c' .aoti;oio ;m/10+0 ti-ia afarlt nfO '2iMi i tnnnrl ralaaca nC A bcny aThnn ii;n nn contmnpil






37

identifiable, limited harborages such as trash piles, under wooden decks, and in trees and


treeholes (Fleet et al. 1978, Appel & Rust 1985,


Brenner 1988, Brenner & Pierce 1991).


The release of oothecal parasitoids in these habitats seems most logical. The objective of

this study was to evaluate the potential of prolonged, inundative release of A. hagenowii

for biological control of peridomestic cockroaches in treehole microhabitats.

Materials and Methods


Parasitoid development in irradiated and frozen oothecae. A disadvantage of

releasing parasitoids from laboratory-parasitized oothecae is the likelihood of further,

unintentional cockroach infestation of the habitat. About 35% of the time (see Table 3.2)

cockroaches emerge from an ootheca placed in a vial with a female parasitoid. Therefore,

the development of A. hagenowii in 13 to 19-d-old irradiated-killed, frozen-killed, and

unkilled (fixed qualitative factor, 3 levels) P. americana oothecae was investigated in a

randomized complete block design (RCBD) with three clutches (extraneous factor) of

parasitoids serving as blocks; parasitoids from each clutch were the progeny of a single

female. A RCBD was used because females within a clutch (i.e. sisters) are more similar

genetically than females between clutches; interclutch genetic variability among females

was assumed beforehand to contribute to variability in response variables.

Oothecae were treated by freezing (-16C) for 24 h, or irradiating (2.5 kR) with a

Cesium-137 irradiator (Radiation Machinery Corporation, Parsippany, NJ); control


oothecae were not treated.


To monitor cockroach mortality from treatment, an additional


group of oothecae (n=15 for both treatments and control) were treated the same but not






38

oothecae from each treatment and the control were randomly selected and placed

individually into 18-ml snap cap vials containing a single, randomly selected 2-d-old, unfed

female parasitoid. Oothecae were incubated at 30C under natural lighting until

cockroaches or parasitoids emerged, at which time cockroaches were discarded, but

parasitoids were stored in 70% isopropyl alcohol until they could be sexed and counted.

Response variables from oothecae that produced a female-biased clutch of parasitoids

were (1) number of emergent female progeny per clutch, (2) number of emergent male

progeny per clutch, and (3) developmental time, defined as the time (d) from experiment

initiation until the day of progeny emergence. If neither cockroaches nor parasitoids had

emerged after 45 d, oothecae were counted and discarded. For each treatment, across all

clutches, was recorded the fate of each ootheca. Each ootheca could have produced (1) a

female-biased clutch of parasitoids, or (2) a male-biased clutch of parasitoids, or (3)

cockroaches, or (4) neither parasitoids nor cockroaches.

In a second experiment, the effect of ootheca age (hereafter age) and irradiation

dose (hereafter dose) on the production of A. hagenowii was investigated in a 3 by 4

factorial experiment with blocking employed (see Ott 1988, p. 696 for design); 8 clutches


(hereafter clutch) served as blocks.


The fixed qualitative factor age had 3 levels, 3 to 11-,


12 tol8-, and 19 to 25-d-old, while the fixed factor dose had 4 quantitative levels,


20, and 40 kR of radiation. For each clutch,


oothecae from each age/dose combination


were randomly selected and placed individually into 18-mi snap cap vials containing a


oinnli 'LrLIA 1nf0A Fa n _nnorhr4A T k^ ni ,n n n a ;3 +


~nnl niiwr~nr nF nnC\broa ;n t~n Cn~fnr;nl






39

Categorical data analysis. For the RCBD experiment, differences in proportions

among the three treatments (rows) for each of the four responses (columns) ((1) number

of oothecae from which emerged a female-biased clutch, or (2) number from which

emerged a male-biased clutch, or (3) number from which emerged cockroaches, or (4)

number from which emerged neither parasitoids nor cockroaches) were analyzed in a 3 by

4 contingency table with a Chi-square Homogeneity of Proportions test (Ott 1988, p.253)

since row marginal totals were fixed (n=15) in advance. Proportions among the treatments

(i.e. within each column) were separated by non-inclusion of zero in a 95% confidence

interval constructed on the difference in proportions between treatments (i.e. pairwise

comparisons of treatment proportions within each column) (Hochberg & Tamhane 1987,

p. 275 for formula).

Equal variance assumption testing and analysis of variance. The assumption of


equality of variances (Ho:


Variances are equal versus HA: At least two variances are


unequal) among treatments in the RCBD experiment was tested with Levene's test

(Levene 1960, Conover et al. 1981) for the following four response variables: (1) number

of female progeny per clutch, (2) number of male progeny per clutch, (3) the arcsine of the


square root of the


sex ratio, and (4) developmental time. In Levene's test, the mean of


each sample is subtracted from each data point within the sample, and an analysis of

variance (PROC GLM in SAS) conducted on the absolute values of these differences

(SAS Institute 1988). In the RCBD experiment, means were computed for each treatment


n2rnAc all rli tr'hic


Thn rnrrcncnnnrflnn CAQ C hflfllIT frr T ax/,nati fct ,aC' V---- frotJiran*+





40

different variances among treatments for the number of female (F=0.14, df=2,32,

P=.8656) and male (F=0.73, df=2,32, P=.4906) progeny per clutch, arcsine of the square

root of the sex ratio (F=0.20, df=2,32, P=.8214), and developmental time (F=0.86,

df=2,32, P=.4325). Subsequently, these variables were analyzed by analysis of variance


with the corresponding SAS MODEL for a RCBD,


Y= treatment clutch; both hypotheses


were tested with the mean square for error at alpha=.05

For the factorial design, Levene's test was conducted on the means of each factor

level combination, across all clutches, by analysis of the main effects and their interaction;


the SAS MODEL statement for Levene's test was Y= dose age dose*age.


The number of


females (F=1 .02, df=l 1,113, P (total model)=.4356) and the log of the number of males


(F=1.16, dfll1,


16, P (total model)=.3213) were found to have variances that did not


differ significantly, and were therefore analyzed by analysis of variance with the SAS


MODEL Y


dose age dose*age clutch (see Ott 1988, p.696 for analysis).


Parasitoid release.


Weekly for 30 wk (beginning June


1993) parasitoids were


released in five individual Blattidae-infested (P. americana, P. australasiae, and P.

fuliginosa) treeholes by placement of laboratory-parasitized oothecae into a parasitoid


release chamber (Fig.


4. 1) mounted near (<60 cm) the entrance of each treehole. Before


laboratory-parasitization, 1 to 14-d-old oothecae were killed by a 24 h freeze (-16C) (for

each release wk 1 through 12) or a dose of irradiation (2.5 kR) (for each release wk 13

through 30). After killing, oothecae were immediately (<2 h) placed individually into 18-

ml snap cap vials containing a single, 2- or 3-d-old female parasitoid, and incubated at





41

The release of parasitoids was accomplished by placing the laboratory-parasitized

oothecae, hereafter termed "seeds", in the male side of a plumbers pipe adapter (Fig. 4.1 E;


Carlon Co., E943 H;


cm high (2 cm threaded),


cm i.d.); to the bottom of the male


side was attached a screen (3 by 3 mm mesh) to secure the seeds, and to allow parasitoids


to escape after emergence.


The female half (Fig.


4. ID; 5.5 cm high (2 cm threaded inside),


5 cm i.d.), also screened, was fastened to the outside of a larger section of pipe (Fig. 4.1C;

6 cm o.d., 11 cm from apex of dome to bottom) open on the bottom but with a closed,


dome-shaped apex onto which lay a piece of plywood (Fig.


4.1B; 1 cm thick, 21.5 by 21.5


cm) which had a hole (7 mm diam.) drilled through its diagonal center. A nylon cord (Fig.

4.1 A) was guided through the plywood and a similar hole drilled in the apex of the


domed-top of the large pipe (Fig.


4. IC); a triple-knot was then tied inside the dome to


allow the entire apparatus to hang from the cord. A 30 cm spike was driven ca.


4-5 cm


into the tree, and to its end was tied the loose end of the cord so that the entire unit was

suspended above the ground and away from the tree.

After remaining in the field for 3 or 6 wk, seeds were returned to the lab, and the

number from which parasitoids had emerged was counted. Parasitoid emergence was

determined by the presence of an exit hole (0.5-1 mm diam.), and confirmed by opening all

seeds. Presence of a parasitoid emergence hole was not always indicative of total

parasitoid emergence; sometimes only a few parasitoids would emerge, leaving most


inside, where they would eventually die.


When this occurred, the seed was counted as


having not produced parasitoids. Also, sometimes a parasitoid emergence hole was not






42

parasitoids; the inside of a seed from which parasitoids had emerged was clean, and devoid

of embryonic material. By opening each seed and observing from the inside out, the

parasitoid exit hole often became evident.

Estimates of the number of female parasitoids released per treehole for the 30 wk

test were based on a mean of 53.2 female A. hagenowii per clutch (Table 4.5) from

irradiated 3 tol 1-d-old oothecae. This value was used to calculate the maximum number

of female parasitoids which could have emerged, per treehole microhabitat, had all


oothecae producing parasitoids produced a female-biased clutch.


This, however, is often


not the case; a percentage of oothecae will produce an all-male clutch. Therefore, the

number of female parasitoids which could have been available had only 80, 60, and 40% of

oothecae producing parasitoids produced a female-biased clutch (i.e. 20, 40, and 60% of

the oothecae produced an all-male clutch) was also calculated.


Parasitism of sentinel oothecae.


To monitor the activity and biological control


potential of A. hagenowii at natural and inflated (i.e. inundative release) populations, the

parasitism of sentinel P. americana oothecae was followed weekly for one year near (<30

cm) the entrance of 25 Blattidae-infested treeholes on the University of Florida campus


beginning on April 13, 1993.


Table 4.


provides, for each treehole, the wk sentinel


oothecae were used, the number of sentinel oothecae used, and the percent total recovery


of sentinel oothecae. Parasitism was monitored by placing 3,


ootheca-holding device (OHD; Fig.


<14-d-old oothecae into an


4.2). Oothecae were replaced after 1 wk. Exposed


oothecae were returned to the lab, and held in 18-ml snap cap vials at 28-30C for (1) 45





43

of each ootheca was recorded as having produced parasitoids, cockroaches, or neither.

The oothecae from one OHD were held together in the same vial. There were 1-3 OHDs

placed per treehole.

Each OHD was stabilized by a 171 gm pyramid-shaped, lead fishing weight (Fig.

4.2K); attached to the weight was ca. 30 cm of 60 lb. test monofilament fishing line (Fig.

4.2B). The monofilament line was guided through the inverted plastic cap from a 20 ml


liquid scintillation vial (Fig.


4.2J; 1.3 cm high, 2.7 cm diam.; Wheaton Scientific, Millville,


NJ) which had a 3 mm diameter hole drilled through its center top. Next was guided a disc


of thin, pliable plastic (Fig.

(Sweetheart Plastics, Wilm


4.2H;


uington, M/


cm diam.) cut from the top of a 237 ml cup

\). The disc served as a landing and crawling


platform for parasitoids. Onto the disc was p

0.6 cm high) from a plastic souffle cup (Fig.


inverted, the round top (4.8 cm diam.,


4.2G; 29.6 ml, Solo Cup Company, Urbana,


IL) which served as the holder of the sentinel oothecae (Fig.


vial cap (Fig.


4.21). Another scintillation


4.2F) was then placed, dorsum up, into the souffle cup top to serve as a


spacer for the roof of the OHD. On top of the cap was placed the dome-shaped bottom


(i.e. roof) (Fig. 4.2E; 6.5 cm diam.) from a 237 ml plastic cup (Sweetheart Plastics,

Wilmington, MA). The hole in the apex of the dome was sealed, to prevent water seepage


into the ootheca-holding well, with a 2 by


cm patch of rubber from a bicycle tube (Fig.


4.2D) secured by the weight of a 43 gm barrel-shaped,


lead fishing weight (Fig.


4.2C).


The loose end of the monofilament line was then secured to the end of a steel spike (Fig.


4.2A; 13 cm long) driven ca. 3 cm into the tree.


The OHD was constructed so that all the






44

Results


In the RCBD experiment, a significantly greater percentage of irradiated-killed and

control oothecae produced a female-biased clutch than oothecae killed by freezing; this


was the result of A. hagenowii's


inability to develop in some frozen-killed oothecae (Fig.


4.3). The proportion of oothecae from which a female-biased clutch emerged for

irradiated-killed and control treatments was not significantly different.

Analysis of variance among treatments in the RCBD experiment indicated no

difference in the number of female (P=.4016) and male (P=.8181) progeny per clutch,


arcsine of the square root of the


sex ratio (P=.7460),


and developmental time (P=.5569)


(Tables 4.2 and 4.3).


Therefore, by pooling all data (n=35 clutches), the mean (95%


confidence interval) number of female and male progeny per clutch,


sex ratio, and


developmental time was 68.5 females (64.7-72.3), 7.1 males (5.7-8.5), 0.091 male (.080-

.103), and 28.7 d (28.2-29.2).

For each factorial analysis of variance, only age was significant (Table 4.4).

Therefore, the effect of age on the number of emergent females and the log of the number

of emergent males per clutch was analyzed, over all doses, by pairwise comparison of

adjusted treatment means (LSMEANS age / Pdiff; in SAS) for clutches (Table 4.5).

Significantly more female parasitoids emerged from 3 tol 1- and 12 to l8-d-old oothecae

than from 19 to 25-d-old oothecae. Although the number of emergent females from 3 to

11- and 12 to 18-d-old oothecae was not significantly different at alpha=.05, the P value


( 0747) was small enonuh to indicate the likelihood of a significant difference.


The number






45

During the 30 wk release program there was a total of 1,399 seeds placed in each


of the


release treeholes, with the number ranging from 10 to


11 per treehole per wk


(median=47


, mode=30), depending on availability. For the first 12 wk, seeds were killed


by freezing (-16C) before being laboratory-stung, incubated, and distributed among the


five release treeholes; 19.6+3.0% (mean+SE) (n=60; 12 wk


treeholes) of the frozen-


killed seeds placed per wk per treehole produced a clutch of parasitoids. For the remaining


18 wk seeds were killed by irradiation (2.5 kR);


26.2+3.


7% (n=90; 18 wk x 5 treeholes) of


the irradiated-killed seeds placed per wk per treehole produced a clutch of parasitoids.


Over the entire 30 wk release period, in treeholes 1,


and 10 there were


and 356 seeds, respectively, from which parasitoids emerged, or


27.1,


25.2,


26.7,


27.3,


25.5%


of the total number of seeds (1,399) placed per treehole.


These percentages resulted in comparable per treehole estimates of the number of female

parasitoids released (Table 4.6). Assuming all emergent clutches were female-biased with

a mean of 53.2 females per clutch, the maximum number of female parasitoids released


(20,322) was in treehole


, while the lowest (18,780) was in treehole


In treeholes where parasitoids were released, sentinel ootheca parasitism for the 10


wk period July


28 through September


inclusive, (6th-15th wk of release) was 70-


100%, and 50-80% on July


, October 20, November 10, and December 1, representing


the 3rd, 18th, 21st, and 24th wk of release, respectively (Fig.


4.4). Sentinel ootheca


parasitism in control treeholes reached its peak on July 7, when it was 46%. Control


parasitism was between ca. 10 and 40% during 26 of the


wk from April 6 through


_


1





46

decreased during the fall and into the winter months. Parasitism rose above 10% only


twice (12% and


7%) during the 20 wk period from November 17 through March 30


(1994), and was 0% on six of those wk.

Parasitism of sentinel oothecae in treeholes where parasitoids were released was

higher than in control treeholes each wk after the parasitoid release program was initiated

on June 23. Furthermore, parasitism of sentinel oothecae in release treeholes was higher


than in the control treeholes for


10 of the


1 wk after the parasitoid release program was


terminated.


Discussion


A principal requirement for the inundative field-release of parasitoids from

laboratory-stung hosts is to prevent further infestation from unsuccessful laboratory-

parasitism. Morgan et al. (1986) prevented Musca domestic L. adults from emerging by

irradiating 2-d-old pupae with a 20 kR dose of gamma irradiation. Furthermore, exposing


2-d-old pupae to Spalangia endius within


h after they had been irradiated with


resulted in 95% successful parasitism. Irradiated pupae could also be stored at refrigerator

temperatures for several wk and remain acceptable to female parasitoids. Parasitism of


irradiated pupae stored at 15C in paper cartons was ca. 80-90% for each of the first


of storage, but was <36% for each of the next


wk. Storage of irradiated pupae (15C) in


small heat-sealed glass containers in a plain air or nitrogen atmosphere provided 2-3 more

wk of shelf-life, as successful parasitism remained above 70% through the seventh

(nitrogen atmosphere) and eighth (air atmosphere) wk of storage. Parasitism of pupae held





47

In the present study, >2.5 kR irradiation or 24 h at -160C killed 100% of exposed

P. americana oothecae. Per ootheca production of female A. hagenowii from irradiated-

and frozen-killed oothecae was not significantly different (Table 4.3), but the proportion

of irradiated-killed oothecae that yielded a female-biased clutch was twice that of frozen-


killed oothecae (Fig.


4.3). From a practical standpoint for field-release, the advantage of


parasitoid development in killed oothecae is obvious. Release of parasitoids in the field

from laboratory-parasitized oothecae can be achieved without the threat of accidental

infestation from unsuccessful laboratory parasitization of viable oothecae. Furthermore,

radiation dose did not affect the number of female parasitoids produced per ootheca

(P=.7761) in the factorial experiment, but the age of the ootheca did (P=.0012) (Table

4.4). In Chapter III, the randomized block design showed a borderline difference in the

number of females per clutch (P=.0997) (Table 3.4) among 3 ages of oothecae, again

suggesting a possible age effect on the number of females per clutch. Differential female

production could be the result of several factors. First, it is known (Heitmans et al. 1992)

that the probability of offspring survival increases with clutch size in A. hagenowii. If

mothers deposit a greater number of eggs per oviposition in younger oothecae, this might

account for the larger number of emergent females in younger hosts. One way to

document this would be to dissect and count the number of eggs remaining in the ovaries


of females that had oviposited in different aged oothecae.


The combination of egg counts


with emergent progeny might elucidate the difference. Also, there may be an age-related

physical or physiological constraint of the host on the parasitoid affecting the number of






48

Parasitism of sentinel oothecae in release treeholes was higher than in the control


treeholes for 10 of the


1 wk after the parasitoid release program was terminated. The


final date of parasitoid release was January 19, 1994 and, as stated earlier, seeds remained

in the field for 6 wk to allow parasitoids to emerge, as development of A. hagenowii is


slowed by cool temperatures (Narasimham 1984); ca. 26% of the


seeds, or about 5


seeds per treehole, produced parasitoids during the final 6 wk period. In laboratory

studies, females of the Gainesville, Florida strain of A. hagenowii survived 3.4 and 7.3 d

when unfed and honey-fed, respectively (Hagenbuch et al. 1988). Conservatively then, if

all parasitoids placed in the field on January 19 emerged during the last wk (February 23

to March 2), all would have died by March 9. Parasitism of sentinel oothecae from March

9 forward then, would be the result of naturally occurring parasitoid populations. In


treeholes where parasitoids were released (Fig.


4.4), parasitism of sentinel oothecae


remained higher than in control treeholes for the 4 wk period from March 9 through


March 30.


This is a good indication that parasitoids from naturally occurring Blattid


oothecae, parasitized during the release program, served as inoculum for sentinel oothecae

during that 4 wk period. Unfortunately, sentinel ootheca parasitism monitoring was

terminated on April 6.

The biological control of urban pests has recently received a great deal of

attention. Increased legislation resulting in the banning of certain active ingredients, the

high cost of organic pesticide development, pest resistance, and the lack of active

inrPrClintc xiAth n mnpe nf art inn arp all drrivino the stnc.tlural nest control industry to






49

(Hagenbuch et al. 1989, Pawson & Gold 1993 ), the potential of A. hagenowii as a

biological control agent has been documented. A method has been developed whereby

oothecae can be killed, laboratory-stung, and distributed in the field for mass-release of A.

hagenowii in biological control programs of Blattid cockroach species.















Screen
\


Side View


C


.---'
t-. ,---


Top


View







51


Table 4.1. For 25 treeholes, the weeks sentinel ootheca parasitism was monitored, the total number of
sentinel oothecae used, and the overall percent recovery of sentinel oothecae from each treehole.

Weeks sentinel # Sentinel Recovery (%) of
Treehole oothecae were used oothecae used sentinel oothecaeb


11-21


SParasitoids were released weekly for


30 wk, from Jun


e 23, 1993 to January 19, 1994.







Tree


Fig. 4.2. Ootheca-holding device for use in monitoring sentinel ootheca
parasitism in treehole microhabitats. A, 13 cm spike; B, 60 lb. test monofilament;
C, 43 gm barrel weight; D, rubber tubing; E, roof; F, spacer cap; G, inverted
souffl6 cup top; H, plastic disc; I, ootheca; J, inverted cap; K, 171 gm pyramid
weight All cnmnnnentQ nfthfe devire are a2t1l2l i7pe































Female-biased


Male-biased Cockroaches


Neither


Response


Fig. 4.3.


Percentage of oothecae from which emerged (1) a female-biased clutch of


parasitoids, or (2) a male-biased clutch of parasitoids, or (3) cockroaches, or (4) neither
parasitoids nor cockroaches in the RCBD experiment. Chi-square Homogeneity of


Proportions test (Ott 1988, p


X2=2


df-=6, P<.005. For each response, bars


separated by different letters are significantly different. Proportions within each response
were separated by non-inclusion of zero in a 95% confidence interval constructed on the
difference in proportions between treatments (i.e. pairwise comparisons of treatment


proportions within each response) (Hochberg & Tamhane, p.


for formula).






54





Table 4.2. Analysis of variance for the number of female and male


progeny per clutch, arcsine of the square root of


sex ratio, and


developmental time (RCBD experiment).

Dependent Sum of Mean
Variable Sourcea Squares Square F Value P Value


# Females


Treatment


180.378


90.189


0.4016


Clutch


1684.


0.0010


Error


Total


2876.195


4741 143


# Males


Treatment


0.8181


Clutch


83.981


91.990


0.0054


Error


Total


441.950


631.886


Arcsine


Treatment


0.00183


0.000


0.7460


Clutch


0.03032


0.01516


0.0144


Error


0.0928


0.00309


Total


.12497


Time


Treatment


Clutch


0.60


10.845


0. 1457


Error


Total


79.14


93.143


a For each analy


of variance, the degrees of freedom for Treatment, Clutch,


Error, and Total are 2, 2, 30, and 34, respectively.








55








~in enVWO

Cur C.., +I +1 +1 +1
-o~ ~ c C
-k C )M r'



4- 4- -I -~ CJ
C -I




ut Cu Cu C

4) In 4
-o~~ A 4). C' '. C
-o Cu C)
U)b : r d e
$ )~ene
-) .- I
-t3 '.4 0
I- l \ \ 0
..cta In 0



EEn
N 0=


vl~~C .-)

cC3




C-c


C)C



9~~~~~ I+ + I
ed C>

oo -o

C c c
I op bi) __



it4C II~
H
-~cfe





56







Table 4.4. Analysis of variance for the number of female progeny per clutch, and


the log of the number of males per clutch (3


x 4 factorial experiment).


Dependent Sum of Mean
Variable Sourcea Squares Square F Value P Value

# Females Dose 223.937 74.646 0.37 0.7761

Age 2915.469 1457.734 7.19 0.0012

Dose*Age 641.144 106.857 0.53 0.7867

Clutch 15735.926 2247.989 11.09 0.0001

Error 22098.241 202.736

Total 41900.305

Log (# males) Dose 0.34585 0.11528 1.88 0.1371

Age 0.52112 0.26056 4.25 0.0167

Dose*Age 0.39835 0.06639 1.08 0.3770

Clutch 1.33946 0.19135 3.12 0.0048

Error 6.68069 0.06129

Total 9.64060


aFor each analysis


of variance, the degrees of freedom for Dose, Age, Dose*Age,


Clutch, Error, and Total are 3,


2, 6,


7, 109, and 127.






57







Table 4.5. Means separation by pairwise comparison
of adjusted treatment (i.e. age) means (LSMEANS in
SAS) for clutches for the number of females per
clutch and the log of the number of males per clutch.

Number (mean) per clutch

Age (d) n Femalesa Malesb

3-11 48 53.2 a 5.3 a

12-18 44 47.9 a 3.9 b

19-25 36 41.2b 3.8b


a Pairwise comparison of LSMEAN


vs 12-18,


P=.074


vs 19-25,


P=.0002; 12-18


vs 19-25,


P=.0386.


b Pairwise comparison of LSMEANS: 3-1


vs 12-18,


P=.0113; 3-11


vs 19-25, P=.0085; 12-1


19-25,


P=.8277.


















Table 4.6.


Estimated numbers of A. hagenowii females released in 5


treeholes over 30 wk.


# Females released per given treehole


% Female-biased


clutches a


20,163


18,780


19.844


20,322


18,939


6.130


12,098

8,065


1.268


7,512


1.906


7,937


12.193


8,129


1.364


7.576


a Percentage of the total number of parasitoid-producing oothecae from
which a female-biased clutch emerged. Female production estimates are
based on a mean of 53.2 female parasitoids per clutch (Table 4.5). For
example, for treehole 1 there were 379 oothecae from which parasitoids
emerged. If 100% of clutches were female-biased, then ca. 20,163 (379
x 53.2) female parasitoids were released during the 30 wk period. If only
80% of the 379 oothecae produced a female-biased clutch, then 16,130
females were released, assuming the remaining 20% of clutches were all-
male.




























































-a'
-


3 3
S M
6
O
C
zt O


*d
P,
"d
a,
o

9,







0/


91/

91,I;
9Z/

6 I/ I,
91,I I,


OI,.
1 / 1.

LVOL-
9/01.
816
8/6

SVS

/..IL
47/9




6/9
929/
^L/9
8; /1
I,/11


0 0 0
0 00 cD











CHAPTER V
A KAIROMONE FOR Aprostocetus hagenowii (Ratzeburg), A
EULOPHID PARASITOID OF
AMERICAN COCKROACH (Periplaneta americana (L.)) OOTHECAE

Introduction


Host-location, or finding, by insect parasitoids has been reviewed by several


authors (Vinson 1976,


Weseloh 1981).


Weseloh (1981) defined chemical host-location as


the discovery, or perception, of a host from a distance due to short- and long-range

volatile compounds (i.e. kairomones) produced by the host, its activity, or one of its

products. Host-location by short-range kairomones has been well documented, providing

most of the host-location literature. Commonly, short-range kairomones are found on or

emanating from the host or one of its products. Eggs (Leonard et al. 1987), pupae (Carde

& Lee 1989), larvae (Vinson et al. 1975), wing scales (Jones et al. 1973, Shu et al. 1990),


and frass (Jones et al. 1971, Nemoto et al. 1987a,b, Auger et al.


1989, Fukushima et al.


1989) contain compounds which elicited host-finding and host-acceptance behaviors from


parasitoids.


Vinson et al. (1975) identified several methyl-branched, saturated C32,


and C34 hydrocarbons from larvae of Heliothis virescens which elicited host-searching

behaviors from Cardiochiles nigriceps. Jones et al. (1973) and Shu et al. (1990) showed

that tricosane and methyl-branched hydrocarbons found in moth scales attract the egg


parasitoids Trichogramma evanescens and T. nubilale,


respectively. Frass also contains


chemical attractants for parasitoids; both volatile and contact host-location kairomones






62

have been identified from host frass. Jones et al. (1971) identified 13-methylhentriacontane

in the frass of H. zea; this hydrocarbon triggered short-range, host-seeking behaviors by

the parasitoid Microplitis croceipes. Parasitoids characteristically antennate host frass and

alter their search behavior when they contact it.

After locating a host, a parasitoid will examine it, and accept or reject it based on

chemical and physical criteria. Host examination leading to acceptance or rejection

commonly involves antennal drumming behaviors mediated by tactile compounds. For

example, Aprostocetus hagenowii (Ratzeburg), a Eulophid parasitoid of oothecae of

cockroaches in the family Blattidae, exhibited typical host-recognition behaviors

(drumming, tapping, drilling) on glass beads treated with calcium oxalate from colleterial

glands and mucopolysaccharides from salivary glands of Periplaneta americana (L.), the

American cockroach (Vinson & Piper 1986). Comperia merceti (Compere) uses the

ootheca-fixing cement from the cockroach Supella longipalpa as a host-acceptance

kairomone (Van Driesche & Hulbert 1984).

The biology of A. hagenowii has been well documented (Roth & Willis 1954,

1960, Cameron 1955, Edmunds 1955, Narasimham 1984). Its preferred host is oothecae

of P. americana, but it will also parasitize the oothecae of P. australasiae (F.) (Australian),

P. fuliginosa (Serville) (smokybrown), P. japonica Karny (Japanese), P. brunnea

Burmeister (brown), Blatta orientalis L. (Oriental), Eurycotis floridana (Walker) (Florida

woods), and E. biolleyi Rehn.

T Tfnnn oro-nonrrvnnrn Crr/rn dlr i nhrnrarartnrlc ota anA r" an ;mmnAiatali,






63

male progeny per clutch after a 31-35 d developmental period at 30C (Chapter III). The

sex ratio of this parasitoid is strongly female-biased (86%). The broad host range,

reproductive strategy, and life cycle of A. hagenowii make it an excellent candidate for the

biological control of Blattidae. Until now, little was known concerning chemically-

mediated host-location by A. hagenowii. In the present study, I investigated a host-

location kairomone, for A. hagenowii, from P. americana oothecae and frass.


Materia


and Methods


Parasitoid. A. hagenowii used in this study were 1-d-old (29 + 1


h), unfed


females that had emerged from 3-5 P. americana oothecae. The day preceding bioassay,

parasitoids were released together into a clean Plexiglass cage (Hagenbuch et al. 1988),


and the empty oothecae removed to avoid possible habituation to the kairomone.


inside of cages were wiped once every 6 wk with ca.


10% isopropyl alcohol, and air dried


before reuse. See Chapter III for details concerning A. hagenowii and P. americana

colonization and colony maintenance. In short, daily 24, 14-d-old oothecae from field-

collected P. americana were placed individually into 18-ml plastic vials containing a single

2- or 3-d-old female parasitoid and incubated at 28-30C.


Olfactometer.


The Y-tube olfactometer described by Vander Meer et al. (1988,


includes sketch) for the isolation of Solenopsis invicta Buren trail pheromone was used in

this study. In short, the barbed ends of a glass, Y-shaped hose connector (0.6 cm i.d;


Kontes,


Vineland, NJ) were replaced with glass tubing. The resulting common arm was


7 5 cm Each lateral arm extended 5 cm before nenetrating ca 1 5 cm into the male end of





64

4 cm out of the female half of the joint, resulting in a 16 cm insect trap comprised of the

joint; total lateral arm length was 25 cm.

A chamber, constructed from the top half of a 50 ml Nalgene squirt bottle (3 cm


, 1 cm i.d. mouth), was built to allow the unprovoked, natural migration of parasitoids


into the common arm during a bioassay. Into the bottom of the chamber was fitted

stainless steel screen (150 mesh; Small Parts, Miami, FL) to prevent parasitoids from

escaping and to allow the free flow of air from the common arm. Parasitoids were

transferred from the Plexiglass holding cage into the chamber through its mouth, which

was then fitted to the end of the common arm.

Air for all bioassays was breathable quality compressed air which was humidified

(ca. 75% R.H.) over 100-200 ml of saturated NaCI (99.5%; Sigma, St. Louis, MO) before

being regulated through two flowmeters (65 mm variable area, stainless steel float, 333

ml/minute maximum airflow; Cole Parmer, Niles, IL) mounted to the inside, back wall of a


white box (1.25 cm thick plywood; 58 cm high by


open front and top.


74 cm wide by 69 cm deep) with an


The olfactometer was placed on the center floor and connected to the


outlet tube from each flowmeter; a


cm space between the floor and the bottom of the


back wall allowed the flowmeter outlet tubes to connect to the ends of the lateral arms of

the olfactometer.

The air tube connecting the compressed air tank and the 0.5 L flask housing the

saturated NaCI solution was corrugated Teflon FEP tubing (6.4 mm i.d., 250 max PSI;

Cole Parmer, Niles, IL) with 2.54 cm non-corrugated ends. An identical tube connected






65

flowmeter inlet; each inlet was equipped with a Teflon, straight male pipe adapter (6 mm


o.d.).


The outlet from each flowmeter was also equipped with a Teflon, straight male pipe


adapter (10 mm o.d.); corrugated Teflon FEP tubing (10 mm i.d.) connected each

flowmeter outlet with the corresponding lateral arm of the olfactometer.

Ootheca bioassay development. The bioassay was developed by placing a single,


pristine (i.e. not dented, and without attached detritus),


<10-d-old P. americana ootheca in


the distad end of a lateral arm of the olfactometer and recording the number of parasitoids,

of 10 placed in the chamber, that chose the ootheca or the sterile air (i.e. the lateral arm

opposite the ootheca), and those that did not make a choice, after 10 minutes at each of


four airflow rates (30,


60, 120, and 240 ml/minute/flowmeter); the procedure was


repeated but with the ootheca in the other lateral arm of the olfactometer. Each run was

conducted with a unique cohort of 10 parasitoids.

A successful "choice" was the event whereby a parasitoid migrated up a lateral arm


and into the olfactometer trap created by the 24/40 ground glass joint (see sketch,


Vander


Meer et al. 1988); at this point parasitoids could not retreat. A bioassay replicate was the

summation ofparasitoids from both lateral arms for each response, at each airflow rate.


There were 6, 9,


bioassay replicates, or 120, 180, 100 and 100 parasitoids


bioassayed, for the 30, 60, 120, and 240 ml/minute/flowmeter airflow rates, respectively.

After each run of 10 parasitoids, the olfactometer was disassembled, rinsed with acetone

(99.9%; Baxter, Muskegon, MI), and fan-dried for 3 to 5 minutes.





66

The response ofparasitoids to sterile air (i.e. no ootheca) from both lateral arms

was investigated at the highest airflow rate. All bioassays were conducted between 0900

and 1700 h, and 28-310C.

Frass bioassay. After the bioassay for oothecae had been developed, 50, 100, and


200 mg samples (n=2, 3, and


mixed


bioassay replicates, respectively) of 1 to 5-d-old frass from


sex adult and large nymphal P. americana (see Chapter III for cockroach origin,


diet, and rearing) was also bioassayed for its attractiveness to A. hagenowii.


Isolation of kairomone from oothecae. On two occasions 30,


<10-d-old P.


americana oothecae were immersed in successive 10,


and 2 minute hexane (99.9%;


Baxter, Muskegon, MI) washes to remove cuticular lipids. On both occasions, the 4

washes were combined, reduced under a gentle stream of nitrogen to ca. 0.5 ml, and

chromatographed on a silica gel (70-230 mesh; Sigma, St. Louis, MO) packed disposable

Pasteur pipette column (5 cm long bed), wetted with hexane, by gravity elution with

sequential 6-ml aliquots of 0, 1, 5, 10, and 30% anhydrous ether (99.9%; Aldrich,


Madison,


WI) in hexane (Carroll 1961). Each fraction was dried under nitrogen and


immediately (<1 minute) reconstituted in hexane to provide one ootheca equivalent (OE)

per 10 ul. For each fraction, an OE was pipetted onto a paper wick (3 by 14 mm;


Northfork, Tumwater,


WA) and bioassayed. For each OE tested, a paper wick control


wetted with 10 ul hexane was placed in the opposite lateral arm of the olfactometer. As

before, a bioassay replicate was the response of 20 parasitoids (i.e. 10 bioassayed for one

OE in each lateral arm).








For the first separation, there were 4,


4, 3, 3, and 3 bioassay replicates for the 0, 1,


10, and 30% ether in hexane fractions, respectively; for the second separation there


were 3, 3,


bioassay replicates. Each bioassay replicate was expressed as a


percentage of the activity of one ootheca bioassay replicate on the day fractions were

bioassayed (i.e. on the same day, the number of parasitoids responding to a fraction

divided by the number responding to the ootheca). The activities were combined for each


fraction, and reported as mean+SE.


When more parasitoids responded to a fraction than


the ootheca, the response was treated as >100%.

After biological activity was discovered in the hydrocarbon-containing hexane


fraction above, oothecal hydrocarbons were separated and bioassayed.


Thrice, ootheca


lipids were extracted and total hydrocarbons collected by column elution with a 6-mi

aliquot of 1% ether in hexane (Brill & Bertsch 1985). The 1% fraction was concentrated

and chromatographed on 10% AgNO3-impregnated silica gel (200+ mesh, 3 cm long bed;


Aldrich, Milwaukee,


WI) by elution with 6-mi aliquots of 0,


and 10% ether in hexane to


separate saturated and unsaturated hydrocarbons


(Carlson & Service 1980). The resulting


fractions were concentrated and chromatographed again through a 1 to


gel by elution with


cm bed of silica


ml of hexane to remove potentially disruptive silver. Fractions


were dried, immediately reconstituted in hexane to provide one OE per 10 ul, and

bioassayed as before.


of the separations, there was


bioassay replicate for each of the 0,


10% ether in hexane fractions; for the third separation, there were 4,


4, and 7 bioassay






68

Hydrocarbon analysis from oothecae and frass. Total hydrocarbon from oothecae

and frass were analyzed by gas chromatography (GC), as were ootheca hydrocarbon

fractions from AgNO3-impregnated silica gel. Frass hydrocarbons were obtained by

immersion of 200 to 600 mg in hexane for 24 h. Frass pellets were then ground with a

glass rod, rinsed with 3-5 additional aliquots of hexane, concentrated, and

chromatographed on silica gel by elution with 1% ether in hexane (Brill & Bertsch 1985).

Samples were dried and immediately reconstituted in hexane to provide one mg equivalent

of frass per ul. Additionally, the quantity of 6,9-heptacosadiene was determined for

oothecae and frass (n=4 independent separations for each, 1 quantification per separation)

by coinjection of 0. 1 ug hexacosane standard and a known equivalent of total hydrocarbon

from oothecae or frass.

Bioassay of adult female-derived 6,9-heptacosadiene. Fifty adult female P.

americana of unknown age were placed in a jar (4 L) and washed with successive 100 ml


aliquots of hexane for 10 and 15 minutes.


The two washes were combined,


rotoevaporated to ca.


ml, and chromatographed on silica gel (1 cm i.d. by


cm long


bed) by elution with 80 ml 1% ether in hexane; ca.


female-equivalents were saved for


total hydrocarbon analysis. Hydrocarbons from the remaining ca.


45 female-equivalents


were separated on 10% AgNO3-impregnated silica gel (1 cm i.d. by 15 cm long bed) by

elution with 60-ml aliquots of 0, 2, and 10% ether in hexane. Silver was removed as


before, fractions dried by rotoevaporation, reconstituted in


mis hexane, and stored at


m~. iron








To further purify 6,9-heptacosadiene for bioassay,


female equivalents from the


10% ether fraction above were chromatographed again on a Pasteur pipette packed with

10% AgNO3-impregnated silica gel as before, and analyzed with GC to obtain purity. The

amount ofalkadiene was then quantified, formulated in hexane to provide ca. one OE per


0 ul, and bioassayed (n=4 replicate bioassays).


Volatile collection from oothecae.


Volatile hydrocarbons were collected on 20-30


mg Super Q adsorbent (Alltech, State College, PA) packed into the constriction of a

disposable Pasteur pipette (VOLTRAP); the adsorbent was held in place on either end by

glass wool. After packing, the adsorbent was cleaned by elution with 4-5 mis ofhexane,

and blown to dryness with a stream of nitrogen. The ootheca-holding chamber (OHC) was


a glass tube (1.22 m by 0.6 cm i.d.) into which 175 pristine,


to 7-d-old oothecae were


placed. At the airflow outlet of the OHC was attached a sleeve of heat-shrinkable Teflon


tubing that extended ca.


VOLTRAP was fitted into the


cm past the end of the OHC. The airflow inlet of the


cm extension so that it became part of the OHC.


Charcoal gravel (#202292, GC hydrocarbon trap grade; Chromtec, Lake Worth, FL) was


centrally packed between two,


cm long plugs of highly activated charcoal (Lewcott,


Millbury, MA) in a glass tube (22 cm by 0.6 cm i.d.; CHARTRAP) to provide 13 cm of

charcoal filter. The airflow inlet of the CHARTRAP was fitted into the Teflon FEP airflow


outlet tube from an airflow meter (see "Olfactometer" section). To the

the CHARTRAP was attached a sleeve of Teflon tubing that extended

of the CHARTRAP. The airflow inlet of the OHC was fitted into the 2


airflow outlet of


cm past the end


: cm Teflon sleeve






70

(breathable quality) at an airflow rate of 7.212 liters/h, and 28-30C. The VOLTRAP was

eluted with ca. 0.5 ml hexane, concentrated to 100 ul, and analyzed by GC.

Gas chromatography. Because of the simplistic hydrocarbon composition of P.

americana, hydrocarbons in this study were identified by chromatogram comparison with

published identification of P. americana cuticular and hemolymph-derived hydrocarbons

(Baker et al. 1963, Beatty & Gilby 1969, Jackson 1972), and retention time comparisons

with n-alkane standards.


Hydrocarbon samples (in 1 or


ul hexane) were analyzed with a Tracor model 540


gas chromatograph equipped with a cool on-column injector, and a flame ionization


detector held at 340C.


The capillary column was non-polar fused silica (DB-1; 30 m by


0.32 mm i.d., 0.25 micron thickness; J & W Scientific, Folsom, CA). Hydrogen was the


carrier gas at a linear flow rate of 25 cm/sec.


The temperature program began with a


minute hold at 600C, then increased by 200C per minute to 2400C with no hold time, then

increased at 50C per minute to 320C and was held for 30 minutes. Data were recorded

with Turbochrom 3 (PE Nelson, Cupertino, CA) analytical software with a PE Nelson 900

Series Interface.

Results

A. hagenowii was attracted to volatiles emanating from P. americana oothecae and

frass. After 10 minutes, a significantly greater number of parasitoids (7.67) had responded

to the ootheca at the highest airflow rate than at any of the lower airflow rates (Fig. 5.1A).

Furthermore, at this airflow rate parasitoids invariably chose the ootheca over sterile air;





71

(i.e. number choosing ootheca divided by number choosing ootheca plus number choosing

air) of the time (Fig. 5.1B). To determine the response of A. hagenowii to air movement,

parasitoids were subjected to sterile air at the highest airflow rate, since this rate provided

the best response to an ootheca. Parasitoids did not respond to sterile air; of 100

parasitoids (n=5 bioassay replicates) tested, only 1 migrated up the Y-tube. P. americana

frass was also attractive to A. hagenowii; 50, 100, and 200 mg of 1 to 5-d-old frass were


, 124.4, and 96.4% as active as an ootheca (i.e. on the same day, the number responding


to frass divided by the number responding to an ootheca).


Once a suitable bioassay had been developed, the response of


hagenowii to


oothecal cuticular lipids was investigated. Bioassay of one OE of the fractions from silica

gel clearly demonstrated that the majority of biological activity was in the hexane fraction

(i.e. hydrocarbon), moderate activity in the 1% ether fraction, and virtually no activity of


fractions containing more ether (Fig. 5.2A).


The hexane and 1% ether fractions were, on


average, 76.7 and 19.9%, respectively, as active as an ootheca. Oothecal hydrocarbons

were then collected and separated on 10% AgNO3-impregnated silica gel by elution with


and 10% ether in hexane, and each fraction bioassayed as before. All the biological


activity was in the 10% ether fraction (Fig. 5.2B); this fraction was 62.6% as active as an


ootheca. Of 120 parasitoids bioassayed for each of the 0 and


ether fractions, 0 and 1


parasitoids, respectively, responded positively. For the 10% ether fraction, however, 86

out of 180 parasitoids responded positively.

GC analysis of total hydrocarbon from oothecae and frass were qualitatively (Fig.





72

hydrocarbons from oothecae and frass, respectively. The predominant hydrocarbon from

both sources was 6,9-heptacosadiene (peak 3 in Fig. 5.3; after Baker et al. 1963, Beatty &

Gilby 1969, Jackson 1972); it comprised 60 and 46% of all hydrocarbon from oothecae

and frass, respectively.


Separation of oothecal hydrocarbons was accomplished by elution with 0,


10% ether in hexane (Fig. 5.4). Collectively, Figures


and 5.4


provide strong evidence


that the hydrocarbon 6,9-heptacosadiene is the compound responsible for the attraction of


A. hagenowii to P. americana oothecae.

of the peak area in Fig. 5.4D. Though tl


0.28 mV


This compound accounted for 100% (13.2 mV)


he alkadiene was present in the 2% fraction (ca.


Fig. 5.4C), a vast majority was present in the 10% ether fraction (97.9% of all


the 6,9-heptacosadiene collected).


The 0% fraction contained only saturated


hydrocarbons.

On a per mg fresh weight basis, frass contained ca. 37.1 times the amount of 6,9-

heptacosadiene as oothecae (cuticular only). An ootheca contained ca. 2.09+0.30 ug 6,9-

heptacosadiene, or 0.024+0.003 ug per mg fresh weight (based on an oothecal weight of

87.9+1.29 mg, n=9). Frass, on the other hand, contained 0.881+0.10 ug of alkadiene per

mg fresh weight.

To confirm the biological activity of 6,9-heptacosadiene, the collection and

separation of hydrocarbon from adult female cockroaches was made for bioassay. GC

analysis of the total hydrocarbon fraction from female cockroaches showed it to be

qualitatively (Fig. 5.3C) and quantitatively (Table 5.1) similar to hydrocarbons in oothecae






73

collected. GC analysis of the purified 5 female equivalent sample by three temperature


programs estimated its purity at 96.8, 98+


, and 100%. Bioassay (n=4 bioassay replicates)


indicated that ca. one OE of 6,9-heptacosadiene was 92.2% as active as an ootheca;

subsequent bioassay of 0.5 ug elicited a response 78.9% as active as an ootheca (n=l

bioassay replicate).

Volatile 6,9-heptacosadiene from P. americana oothecae was effectively trapped

on Super Q (Fig. 5.5B). Though the adsorbent was prepared for use by elution with 4-5

mis of hexane, some contaminants remained and were eluted with the hydrocarbons after

volatile collection was complete (Fig. 5.5B & C (unlabeled peaks)). Identification of peak

3 in Fig. 5.5B was determined by several methods. First, via coinjection with alkane

standards its Kovats retention index (KI) was 2666 (n=2 independent volatile collections);

the KI of 6,9-heptacosadiene from P. americana oothecae was 2673. In addition,

coinjection of known 6,9-heptacosadiene (Fig. 5.4D) and the volatile collection from

oothecae (Fig. 5.5B) resulted in elution of a single peak at the retention time of the


alkadiene.


The injection was performed isothermally (60C with a


min hold, then 5C per


min to 320C with a


minute hold) and with the same temperature program reported


earlier. It was also suspected that peaks 1 and 2 in Fig. 5.5B were pentacosane and 3-


methylpentacosane, respectively, as compared with the reference chromatogram.


The KI


of 3-methylpentacosane from volatile collection was 2571, while the KI from ootheca-


derived material was


2575.


Pentacosane was identified in the volatile collection by


comparison of its retention time with an authentic pentacosane standard, and its coelution





74

8.377+0.45 ng (n=2 determinations) 6,9-heptacosadiene quantified by coinjection with 0.1

ug hexacosane.


Discussion


Hydrocarbons account for the majority of cuticular lipids in P. americana; Gilby

and Cox (1963) and Jackson (1972) estimated that 75%, and 85-95%, respectively, of

cuticular lipids in P. americana are hydrocarbons. Baker et al. (1963) reported a bimodal

distribution of hydrocarbons from P. americana cuticle and hemolymph, but only the first

series of 5 peaks were identified. Hydrocarbon composition from the 2 sources was


qualitatively identical and quantitatively similar. The

hydrocarbon is in the first series of peaks, containing


vast majority (95% to 98%) of


compounds of 25-29 carbons: n-


pentacosane (12%), 3-methylpentacosane (20%), 6,9-heptacosadiene (65%), n-

heptacosane, and n-nonacosane. Qualitative and quantitative hydrocarbon data from

oothecae, frass, and adult female P. americana in the current study parallel those of earlier

researchers (see Fig 5.3 and Table 5.1); the major hydrocarbon from all three sources was

6,9-heptacosadiene.

Because of the simplicity of its hydrocarbon composition, P. americana has been

used as a model for hydrocarbon synthesis in insects. Nelson (1969) first demonstrated

that hydrocarbons were synthesized in the epidermis (i.e. oenocytes) of P. americana; it is

likely that a lipophorin is responsible for the transport of hydrocarbons to their final

destination (Katase & Chino 1982). Jackson and Baker (1970) and Conrad and Jackson

(1971) reported that injection of labelled iinoleic acid (C 18:2, at carbons 6 and 9) into





75

suggesting that linoleic acid is a precursor to this alkadiene. Dwyer et al. (1981) later

documented that 6,9-heptacosadiene was in fact synthesized in P. americana by elongation

of linoleic acid with acetate units, followed by decarboxylation to yield the alkadiene.

This study is the first evidence for kairomonal activity of an alkadiene, and the first

biological activity of 6,9-heptacosadiene, though its existence has been known from P.

americana since Baker et al. (1963) identified it; its identity has since been confirmed

(Beatty & Gilby 1969, Jackson 1972). Other alkadienes, however, are biologically active;

7,11-heptacosadiene is an aphrodisiac found in the cuticular wax of female Drosophila

melanogaster (Antony & Jallon 1982, Antony et al. 1985), and 6,9-nonadecadiene is a


component of the


sex pheromone of several Geometrid moth species (Szocs et al. 1984,


McDonough et al. 1986). Bartelt et al. (1982) identified the


sex pheromone blend of a


sawfly, Pikonema alaskensis, as a series of 9,19-alkadienes (C28 to C39). Bartelt and

Jones (1983) later discovered that some of the 9,19-alkadienes slowly air oxidized to

aldehydes, and one prominent aldehyde, (Z)-10-nonadecenal, had primary pheromonal

activity. The smallest alkadiene (C28) had the greatest amount of loss (47%) after a 24 h


exposure to air. The kairomonal activity of 6,9-heptacosadiene may be an aldehyde


product of its oxidation, but in our bioassays the alkadiene was not exposed to air until its


bioassay, and then for a period of only


0 minutes. Additionally, the alkadiene was


collected as a volatile on Super Q (Fig. 5.5B), an indication of its lack of oxidation.

Identification of a kairomone in the frass and cuticle of a host was first

documented by Jones et al. (1971); 13-methylhentriacontane was identified in the frass,








croceipes.


When parasitoids contact this hydrocarbon, they intensely search the area by


rubbing the substratum with their antennae.


The location of the methyl group and the


carbon chain length were very important to the activity of this kairomone, which was

active at 50 ng. Other methylated hydrocarbons have been shown to have kairomonal

properties; 13,17-dimethylnonatriacontane from moth scales was identified as effecting the


host-seeking behavior in T.


nubilale, an egg parasitoid of Ostrinia nubilalis (Shu et al.


1990).

The discovery of parasitoid-behavior modifying compounds is important not only

from a basic biology standpoint, but has also been shown to have practical applications as

well. Lewis et al. (1972) enhanced significantly the parasitism of Lepidoptera eggs by T.

evanescens in Petri dishes, greenhouses, and cotton fields by pre-release application of a

hexane-soluble host-location kairomone, later identified as tricosane, from moth scales.

Eggs placed in locations treated with the extract suffered higher parasitism than control


eggs


(i.e. no kairomone). In a later study, Lewis et al. (1975) discovered that broadcast


application of tricosane resulted in increased parasitism of sentinel and wild Lepidoptera


eggs by wild and released Trichogramma spp. Unlike the inundative release of


pheromone for mating disruption, large-scale application of a host-seeking stimulant did

not serve as a confusant for the parasitoids; it stimulated host-searching, resulting in


increased parasitism.


The possibility of utilizing 6,9-heptacosadiene as a similar host-


seeking stimulant for P. americana oothecae should be considered. The growing interest in

biological and biorational (IGRs and pheromones) control of urban pests, fueled by a






77

provide ample opportunity to investigate their chemical ecologies, and hopefully lead to

the development of alternative control strategies to aid in their management.























30 60 120 240


Airflow (ml/minute/flowmeter)


B All


Responses


Ootheca


None


Response


Ten minute Y-tube bioassay response (meanSE) of 10,


hagenowii to (A) a


l-d-old unfed


<10-d-old P. americana ootheca at each of four airflow rates, and


- -


I








A Total


Lipids


0 1 5 10 30


0 2


Fraction (% ether in hexane)


c~m~amrm



























Fig. 5.3. Gas chromatographic analysis of P. americana total hydrocarbon from (A)
1 to 7-d-old oothecae, (B) 1 to 5-d-old frass, and (C) adult females. Hydrocarbon
identification after Baker et al. (1963). Peak 1: n-pentacosane; Peak 2: 3-
methylpentacosane; Peak 3: 6,9-heptacosadiene; Peak 4: n-heptacosane; Peak 5: n-
nonacosane. See Table 5.1 for corresponding peak area percentages. For each
chromatogram, peak 3 is presented off scale to highlight smaller peaks. In
chromatogram B, peak S is 0.1 ug n-hexacosane standard.













Ootheca


Frass


Adult female


3


I I I
20


I I24
24


28


I I I
32


I I6 I
36


40


44


i 1






82






-l In

4- +1 +1 +1
0 00NO

('U
S~ C Om cc
7 cU M O0


~~I


ncn



Pd+1 +1 +1)
o enrl Coo
c-fld II
C It
It O 0
o o

m +1 1 +1
6Inn

O O t- ON
P;'I 6Y~C
E `66\
Cl +1+ 1e
U -c C
oY QIIc od
r~ a rO C
-o -
U C-)


c, C
) IO 0
00 +1 +1 +1I 0-

a -

4-a It, in ON


C) I

'U 6,UP


C) 0) 0 Li


'U~ II
H -


























Fig. 5.4. Gas chromatographic analysis of(A) total hydrocarbon from 1 to 7-d-old
P. americana oothecae (as a reference chromatogram), and ootheca hydrocarbons
which were separated on 10% AgNO3-impregnated silica gel by elution with (B) 0,
(C) 2, and (D) 10% ether in hexane. 0.2 ootheca equivalents were injected for each
of the 0, 2, and 10% fractions.


















24~


22~


Ootheca


0% ether


C


9.2
88--
Q~ _

9.0


8.8


2% ether


10% ether


16 20 24 28 52 56


C


J~hj



























Gas chromatographic analysis of(A) P. americana ootheca total


hydrocarbon (as a reference chromatogram), (B) volatile hydrocarbons from <10-
d-old oothecae, and (C) contaminants from Super Q volatile adsorbent. Peak l=n-


pentacosane, 2=3-methylpentacosane, 3=6,9-heptacosadiene,


4=n-heptacosane,


5=n-nonacosane.








100

80


60


40


20

0


1o0.

9_



-1
-M
8-
t
7-
-I


Ootheca




5


Volatiles


Super Q

Contaminants


II l m m~pm7miii11111liili tI 11111~IIIIII iili












REFERENCES CITED


Ali, M. & W


Islam. 1983. Morphology of Tetrastichus hagenowii Chalcidoidea


Tetrastichidae a parasite on the oothecae of cockroach. Bangladesh J. Zool. 11(2):
79-86.


Antony,


C., T


L. Davis, D. A. Carlson, J.


Pechine & J. M. Jallon. 198


Compared


behavioral responses of male Drosophila melanogaster (Canton S) to natural and


synthetic aphrodisiacs. J. Chem. Ecol.


11(12):


1617-1629.


Antony


C. & J. M. Jallon.


1982.


The chemical basis for


sex recognition in Drosophila


melanogaster. J. Ins.

Appel, A. G. & M. K. Rust.


Physic

1985.


873-880.


Outdoor activity and distribution of the smokybrown


cockroach, Periplaneta fuliginosa (Dictyoptera: Blattidae). Environ. Entomol. 14:
669-673.


Arthur, A. P. 1981
Jones & W


Host acceptance by parasitoids, pp. 97-120. In D. A. Nord
J. Lewis [eds.], Semiochemicals: Their role in pest control.


*


lund, R. L.
Wiley &


Sons, New York, New York.


Auger, J., C. LeComte, J. Paris & E.


Thibout. 1989. Identification of leek-moth and


diamondback-moth volatiles that stimulate parasitoid, Diadromus pulchellus. J.


Chem. Ecol.


15(4): 1391-1398.


Baehrecke, E. H.,


Vinson & H. J.


Williams.


1990. Foraging behavior of Campoletis


sonorensis


in response to Heliothis virescens


and cotton plants. Entomol. Exp.


Appl.


Baker, G. L., H. E.


Vroman & J. Padmore. 1963.


cockroach. Biochem.


Biophys. Res. Commun.


drocarbons of the American
13: 360-365.


Barlin, M. R., S. B.


Vinson & G. L. Piper. 1981. Ultrastructure of the antennal sensilla of


the cockroach-egg parasitoid, Tetrastichus hagenowli (Hymenoptera: Eulophidae).
J. Morph. 168: 97-108.


,1.





88

Bartelt, R. J. & R. L. Jones. 1983. (Z)-10O-nonadecenal: A pheromonally active air
oxidation product of (Z,Z)-9,19-dienes in yellowheaded spruce sawfly. J. Chem.
Ecol. 9(9): 1333-1341.


Bartelt, R. J., R. L. Jones & H. M. Kulman. 1982.


yellowheaded spruce sawfly


Hydrocarbon components of the


sex pheromone: A series of(Z,Z)-9,19-dienes. J.


Chem. Ecol. 8(1): 95-114.

Beatty, I. M. & A. R. Gilby. 1969. The major hydrocarbon of a cockroach cuticular wax.
Naturwissenshaften 25: 373.


Bouchard,


& C. Cloutier. 1984. Honeydew as a source of host-searching kairomones


for the aphid parasitoid Aphidius nigripes


(Hymenoptera: Aphidiidae).


Can. J.


Zool.


62: 1513-1520.


Bouchard,


.&C.


Cloutier. 1985. Role of olfaction


Aphidius nigripes (Hymenoptera: Aphidiidae).


n host finding by aphid parasitoid
J. Chem. Ecol. 11(6): 801-808.


Brenner, R. J. 1988. Focality and mobility of some peridomestic cockroaches in Florida
(Dictyoptera: Blattaria). Ann. Entomol. Soc. Am. 81(4): 581-592.

Brenner, R. J. & R. R. Pierce. 1991. Seasonality of peridomestic cockroaches (Blattoidea:
Blattidae): mobility, winter reduction, and effect of traps and baits. J. Econ.


Entomol. 84(6):


1735-1745.


Brill, J. H. & W


Bertsch. 1985


An investigation of sampling methods for the analysis of


cuticular hydrocarbons. J. Entomol. Sci. 20(4):


435-443.


Burks, B. D. 1943. The North American parasitic wasps of the genus Tetrastichus- a
contribution to biological control of insect pests. Proceedings of the United States
National Museum 93(3170): 505-608.


Cameron, E.


1955


On the parasites and predators of the cockroach. I.


Tetrastichus


hagenowii (Ratz.). Bull. Entomol. Res.


137-147.


Carde, R.


T. & H. Lee. 1989. Effect of experience on the responses of the parasitoid


Brachymeria intermedia (Hymenoptera: Chalcididae) to its host, Lymantria dispar
(Lepidoptera: Lymantriidae), and to kairomone. Ann. Entomol. Soc. Am. 82(5):
653-657.


Carlson, D. A. & M.


W. Service.


1980. Identification of mosquitoes of Anopheles


i .1. A 0 T i fl ....- -------__--_--------- C' .. n






89

Carroll, K. K. 1961. Separation of lipid classes by chromatography on florisil. J. Lipid
Research. 2(2): 135-141.


Cave, R. D., M. J. Gaylor & J. T. Bradley. 1987


. Host handling and recognition by


Telenomus reynoldsi (Hymenoptera: Scelionidae), an egg parasitoid of Geocoris
spp. (Heteroptera: Lygaeidae). Ann. Entomol. Soc. Am. 80: 217-223.


Ceianu, C. 1986. Two parasites ofBlatta orientalis L. oothecae:


Tetrastichus hagenowii


(Ratzeburg) (Hymenoptera: Eulophidae) and Prosevania fuscipes (Illiger)
(Hymenoptera: Evaniidae). Arch. Roum. Pathol. Exp. Microbiol., Bucharest.
45(2): 161-167.


Chiri, A. A. & E. F.


Legner.


Response of three Chelonus (Hymenoptera:


Braconidae) species to kairomones in scales of six Lepidoptera. Can. Ent. 118:
329-333.


Clement, S. L.,


L. Rubinik & D. A. McCartney.


1986. Larviposition response of


Bonnetia comta (Dipt.:


Tachinidae) to a kairomone of Agrotis ipsilon (Lep.:


Noctuidae). Entomophaga 31(1): 277-284.


Conover,


. J., M. E. Johnson & M. M. Johnson.


981. A comparative study of tests for


homogeneity of variances, with applications to the outer continental shelf bidding
data. Technometrics 23(4): 351-361.


Conrad, C.


L. Jackson. 1971.


Hydrocarbon biosynthesis in Periplaneta americana.


J. Ins. Physiol.


1907


-1916.


Cornwell, P. B. 1968.


Dmoch, J.


The cockroach, vol. 1. Hutchinson, London.


Lewis, P. B. Martin & D. A. Nordlund. 1985. Role of host-produced


stimuli and learning in host selection behavior of Cotesia (=Apanteles)


marginiventris (Cresson). J. Chem. Ecol.


1(4): 453-463.


Drost,


Y. C., W. J. Lewis, P. O. Zanen & M. A. Keller. 1986. Beneficial arthropod
behavior mediated by airborne semiochemicals. I. Flight behavior and influence of


preflight handling ofMicroplitis croceipes (Cresson). J. Chem. Ecol. 12(6): 1247-
1262.

Dwyer, L. A., M. de Renobales & G. L. Blomquist. 1981. Biosynthesis of(Z,Z)-6,9-


heptacosadiene in the American cockroach. Lipids 16(


1): 810-814.


& L.






90

Edmunds, L. R. 1955. Biological notes on Tetrastichus hagenowii (Ratzeburg), a
chalcidoid parasite of cockroach eggs (Hymenoptera: Eulophidae; Orthoptera:
Blattidae). Ann. Entomol. Soc. Am. 48: 210-213.


Eller, F


. J., J. H.


Tumlinson & W. J. Lewis. 1988. Beneficial arthropod behavior mediated


by airborne semiochemicals:


Source ofvolatiles mediating the host-location flight


behavior ofMicroplitis croceipes (Cresson) (Hymenoptera: Braconidae), a
parasitoid ofHeliothis zea (Boddie) (Lepidoptera: Noctuidae). Environ. Entomol.
17(4): 745-753.


Elzen, G.


W., H. J.


Williams & S. B.


Vinson. 1983.


Response by the parasitoid


Camooletis sonorensis


(Hymenoptera: Ichneumonidae) to chemicals (synomones)


in plants: Implications for host habitat location. Environ. Entomol. 12:1873-1877


Elzen, G.


f-. J


Williams & S. B.


Vinson.


1984. Isolation and identification of cotton


synomones mediating searching behavior of parasitoid Campoletis sonorensis. J.


Chem. Ecol.


0(8): 1251-1264.


Elzen, G.


H. J.


Williams & S. B.


Vinson. 1986.


Wind tunnel flight responses by


hymenopterous parasitoid Campoletis sonorensis to cotton cultivars and lines.


Entomol. Exp. Appl.


285-289.


Fleet, R. R. & G.


. Frankie.


1975. Behavioral and ecological characteristics of a


eulophid egg parasite of two species of domiciliary cockroaches. Environ.
Entomol. 4(2): 282-284.


Fleet, R. R., G. L. Piper & G.


Frankie. 1978.


Studies on the population ecology of the


smokybrown cockroach, Periplaneta fuliginosa, in a Texas outdoor urban


environment. Environ. Entomol.


807-814.


Fukushima, J.,


Kuwahara & T


Suzuki.


1989. Isolation and identification of a


kairomone responsible for the stinging behavior of Bracon hebetor Say
(Hymenoptera: Braconidae) from frass of the Almond moth Cadra cautella Walker.


Agric. Biol. Chem. 53(


1): 3057-3059.


Gilby, A. R. & M. E. Cox.


1963.


The cuticular lipids of the cockroach, Periplaneta


americana (L.). J. Ins. Physiol. 9: 671-681.


Greany, P. D.,


B. Vinson & W


J. Lewis.


1984. Insect parasitoids: Finding new


opportunities for biological control. Bioscience 34(11): 690-696.








Gueldner, R. C., D. A. Nordlund,


J. Lewis, J. E. Thean & D. M. Wilson. 1984.


Kairomones and their use for management of entomophagous insects. XV.
Identification of several acids in scales of Heliothis zea moths and comments on
their possible role as kairomones for Trichogramma pretiosum. J. Chem. Ecol.
10(2): 245-251.

Hagenbuch, B. E., R. S. Patterson & P. G. Koehler. 1989. Biological control of the
American cockroach (Orthoptera: Blattidae) with inundative releases of


Tetrastichus hagenowii (Hymenoptera: Eulophidae). J. Econ.


Entomol. 82(1): 90-


Hagenbuch, B. E., R. S. Patterson, P. G. Koehler & R. J.
of the cockroach oothecal parasitoid, Tetrastichus


Brenner. 1988. Mass production
hagenowii (Hymenoptera:


Eulophidae), and its host, the American cockroach (Orthoptera: Blattidae). J.
Econ. Entomol. 81(2): 531-535.


Hagvar


, E. B. & T. Hofsvang. 1989. Effect of honeydew and hosts on plant colonization
by the aphid parasitoid Ephedrus cerasicola. Entomophaga 34(4): 495-501.


Harlan, H. J. & R. D. Kramer. 1981. Limited host specificity in Tetrastichus hagenowii
(Ratzeburg). J. Georgia Entomol. Soc. 16(1): 67-70.


Heitmans, W. R. B., P. Haccou & J. J. M. van Alphen. 1992. Egg supply, clutch size and
survival probability in Aprostocetus hagenowii (Ratz) (Hymenoptera: Eulophidae),
a gregarious parasitoid of cockroach oothecae. Proc. Exper. & Appl. Entomol. 3:
62-69.


Henson, R. D., S. B.


Vinson & C. S. Barfield.


1977.


Ovipositional behavior of Bracon


mellitor Say, a parasitoid of the boll weevil (Anthonomus grandis Boheman). III.
Isolation and identification of natural releasers of ovipositor probing. J. Chem.
Ecol. 3(2): 151-158.


Herard, F., M. A. Keller,


J. Lewis & J. H.


Tumlinson.


behavior mediated by airborne semiochemicals. III.


988a. Beneficial arthropod
Influence of age and


experience on flight chamber responses ofMicroplitis demolitor Wilkinson. J.


Chem. Ecol.


14(7):


1583-1596.


Herard, F., M. A. Keller,


W. J. Lewis & J. H.


Tumlinson.


behavior mediated by airborne semiochemicals. IV


oriented flight chamber responses
Ecol. 14(7): 1597-1606.


1988b. Beneficial arthropod
. Influence of host diet on host-


ofMicroolitis demolitor Wilkinson. J. Chem.








Hochberg, Y. & A. C. Tamhane. 1987
New York, New York.


Multiple Comparison Procedures.


Wiley & Sons,


Howard, R.


W. Flinn. 1990. Larval trails of Cryptolestes ferrugineus (Coleoptera:


Cucujidae) as kairomonal host-finding cues for the parasitoid Cephalonomia
waterstoni (Hymenoptera: Bethylidae). Ann. Entomol. Soc. Am. 83(2): 239-245.

Jackson, L. 1972. Cuticular lipids of insects-IV. Hydrocarbons of the cockroaches
Periplaneta japonica and Periplaneta americana compared to other cockroach


hydrocarbons. Comp. Biochem. Physiol.


Jackson, L. L. & G. L. Baker.


41B: 331-336.


1970. Cuticular lipids of insects. Lipids 5(2): 239-246.


Jie, L. & N. Wenqing. 1984. Bionomics of Tetrastichus hagenowii parasitizing in the
oothecae of Periplaneta americana. Acta Entomol. Sin. 27(4): 406-409.


Jones, R. L.,


J. Lewis, M. Beroza, B. A.


Bierl & A. N. Sparks. 1973. Host-seeking


stimulants (kairomones) for the


parasite,


Trichogramma evanescens. Environ.


Entomol. 2(4): 593-596.


Jones, R. L.,


. J. Lewis, M.


C. Bowman, M. Beroza & B. A. Bierl.


1971. Host-seeking


stimulant for parasite of corn earworm:


Science


Isolation, identification, and synthesis.


73: 842-843.


Kanayama, A., E


Yoshida & T. Honma. 1975


The parasitism by Tetrastichus


hagenowii


(Ratzeburg) on oothecae of the smoky brown cockroach, Periplaneta fuliginosa


(Serv.) collected in Shizuoka city. Jpn. J. Sanit. Zool.


Katase, H. & H. Chino. 1982


27(2): 157-162.


Transport of hydrocarbons by the lipophorin of insect


hemolymph. Biochim. Biophys. Acta 710: 341-348.


Keller, M. A.


990. Responses of the parasitoid Cotesia rubecula to its host Pieris rapae in


a flight tunnel. Entomol.


Kumarasinghe, N. C.


Exp. Appl.


& J. P. Edirisinghe. 1987


:243-249.

. Oothecal parasites of Periplaneta


americana: Parasitization and development in relation to host age. Insect Sci.
Applic. 8(2): 225-228.

Kumarasinghe, N. C. & J. P. Edirisinghe. 1991. A preliminary survey of domiciliary


cockroaches and their oothecal parasites in Sri Lanka.


Vidyodaya J. Sci. 3(2): 35-


& P.






93

LeBeck, L. M. 1991. A review of the Hymenopterous natural enemies of cockroaches
with emphasis on biological control. Entomophaga 36(3): 335-352.


Leonard, D. E., B. A. Bierl & M. Beroza. 1975.
behavior of the parasitoids Brachymeria


Gypsy moth kairomones influencing
intermedia and Apanteles melanoscelus.


Environ. Entomol.


4(6): 929-930.


Leonard, D. E., Z.


Wu & D. N. Ferro. 1987


Responses of parasite Edovum puttleri to


kairomone from eggs of Colorado potato beetle, Leptinotarsa decemlineata. J.
Chem. Ecol. 13(2): 335-344.


Levene, H. 1960. Robust tests for equality of variances, pp. 278-292. In I. Olkin [ed.],
Contributions to Probability and Statistics. Stanford University Press, Palo Alto.


Lewis,


L. Jones. 1971. Substance that stimulates host-seeking by Microplitis


croceipes (Hymenoptera: Braconidae),
Entomol. Soc. Am. 64(2): 471-473.


a parasite of Heliothis species.


Ann.


Lewis,


W. J., R. L. Jones, D. A. Nordlund & A.
use for management of entomophagous


N. Sparks. 1975. Kairomones and their
insects: I. Evaluation for increasing rates


ofparasitization by Trichogramma spp. in the field. J. Chem. Ecol. 1(3): 343-347


Lewis,


. J., R. L. Jones & A. N. Sparks.


1972. A host seeking stimulant for the egg


parasite Trichogramma evanescens: Its source and a demonstration of its


laboratory and field activity. Ann.


Entomol. Soc. Am. 65(5): 1087-1089.


Lewis,


J.&D


A. Nordlund. 1984.


Semiochemicals influencing fall armyworm


parasitoid behavior: Implications
343-349.


for behavioral manipulation. Fla. Entomol. 67(3):


Lewis,


. J.&J.


Tumlinson. 1988. Host detection by chemically mediated associative


learning


n a parasitic wasp. Nature 331:


7-259.


Lewis,


W.J.


, J. H. Tumlinson & S. Krasnoff. 1991. Chemically mediated associative


learning: An important function in the foraging behavior ofMicroplitis croceipes
(Cresson). J. Chem. Ecol. 17(7): 1309-1325.


Li, L.


. 1990. Mass production of Trichogramma spp. and Anastatus japonicus Ashmead
with artificial diets in China. FFTC-NARC International Seminar on 'The use of
Parasitoids and Predators to Control Agricultural Pests', Tukuho Science City,
Ibaraki-Ken, 305, Japan, October 2-7, 1989.


J. & R.








Ma, R.


P. D. Swedenborg & R. L. Jones. 1992. Host-seeking behavior of Eriborus


terebrans (Hymenoptera: Ichneumonidae) toward the european corn borer and the
role of chemical stimuli. Ann. Entomol. Soc. Am. 85(1): 72-79.

Mack, G. A. & J. H. Skillings. 1980. A Friedman-type rank test for main effects in a two-
factor ANOVA. J. of the Amer. Stat'l. Assn. 75: 947-951.


McAuslane, H. J., S. B.


Vinson & H. J.


Williams.


1990. Effect of host diet on flight


behavior of the parasitoid Campoletis sonorensis


(Hymenoptera: Ichneumonidae).


J. Entomol. Sci. 25(4): 562-570.


McAuslane, H. J., S. B.


Vinson & H. J.


Williams.


1991a.


Stimuli influencing host


microhabitat
58: 267-277.


location in the parasitoid Campoletis


sonorensis. Entomol. Exp. Appl.


McAuslane, H. J., S. B.


Vinson & H. J.


Williams.


1991b.


Influence of adult experience on


host microhabitat location by the generalist parasitoid, Campoletis sonorensis
(Hymenoptera: Ichneumonidae). J. Ins. Beh. 4(1): 101-113.


McDonough, L. M., J


B. Bailey, M. P. Hoffmann, B. A. Leonhardt, D. F. Brown, C. L.


Smithhisler & K. Olsen. 1
Geometridae) component


986. Sabulod


es caberata Guenee (Lepidoptera:


s of its sex pheromone gland. J. Chem.


Ecol. 12(12):


2107-2116.


Morgan, P. B., B. J. Smittle & R. S. Patterson


1986. Use of irradiated pupae to mass


culture the microhymenopterous pupal parasitoid Spalangia endius Walker
(Hymenoptera: Pteromalidae). I. Musca domestic L. (Diptera: Muscidae). J.
Entomol. Sci. 21(3): 222-227.


Mudd, A. & S. A


Corbet.


982. Response of the Ichneumonid parasite Nemeritis


canescens to kairomones from the flour moth, Ephestia kuehniella. J. Chem. Ecol.
8(5): 843-850.


Mudd, A. J., H. H.


Walters & S. A. Corbet. 1984. Relative kairomonal activities of


acylcyclohexane-1,3-diones in eliciting oviposition behavior from the parasite


Nemeritis canescens (Grav.). J. Chem. Ecol.


): 1597-1601.


Muzaffar, N. & C. Inayatullah. 1986. Role of salivary glands of Pyralid borer larvae in
host selection by Apanteles flavipes (Cameron) (Hym: Braconidae). J. Agric. Res.
24(3): 203-205.