INFOILVJ:aTlON TO USERS

This manuscriptPas been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted Thus, some thesis and dissenation' copies are in typewriter face, while others may be from any type ofcomputer printer.

The quality of this reproductioD is depeDdent DpOD the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improperalignment can adversely affect reproduction.

In the unlikely. event that the author did not send UMI a complete mam1script and there are missing pages, these will be noted. Also, if unauthoriZed copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right inequal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back ofthe book.

Photographs included in the origiDal mam1script have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

UMI A Bell & Howell Information Company 300 North Zeeb Road. Ann Arbor. MI48106-1346 USA 3131761-4700 800:521-0600

THE STATUS OF THE BIOLOGICAL CONTROL OF

MEALYBUGS IN HAWAD

A DISSERTATION SUBMITTED TO THE GRADUATE DIVISION OF THE UNIVERSITY OF IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

IN

ENTOMOLOGY

DECEMBER 1995

By

Hector Gonzalez-Hernandez

Dissertation Committee:

Marshall W. Johnson, Chairperson Neil 1. Reimer Diane E. Ullman John W. Beardsley Kenneth G. Rohrbach OMI Number: 9615523

UMI Microform 9615523 Copyright 1996, by UMI Company. All rights reserved.

This microform edition is protected against unauthorized copying under Title 17, United States Code.

UMI 300 North Zeeb Road Ann Arbor, MI 48103 © Copyright 1995 by Hector Gonzalez-Hernandez

III DEDICATION

Dedicated to my wife Pili and my daughter Mariana whose love and

dedication have enriched every day ofmy live.

IV ACKNOWLEDGMENTS

Special thanks to my major professor, Dr. Marshall W. Johnson, whose encouragement,

teaching, valuable advisory work, and friendship made it possible to successfully finish this

enterprise. Also to my advisor Dr. Neil 1. Reimer, for his effort, and time spent teaching

me to work with , and for his friendship. I would like to thank my other committee

members, Dr. John W. Beardsley, Dr. Diane E. Ullman, and Dr. Kenneth G. Rohrbach for

all their suggestions and support to conduct this work. Thanks to Dr. Bruce E. Tabashnik

for his suggestions through my experiments. I am pleased to thank the Consejo Nacional

de Ciencia y Tecnologia (National Council ofScience and Technology) and Colegio de

Postgraduados ofthe Mexican government; Hawaii Governor's Agricultural Coordinating

Committee and the Hawaii Sugar Planters Association for providing financial support to

conduct this Ph. D. program. Many thanks to Calvin H. Oda, Vaeleti Tyrell, and

Francisco Bareng from Del Monte Fresh Produce (Hawaii) Inc., and Dr. Herbert Fleisch

from the Maui Land and Pineapple Co., for providing pineapple fields and technical

support for my field experiments. Thanks to Claide Ragasa and Pamela Fukada for their

technical support. I am grateful to all the Faculty, Staffand graduate students ofthe

Department ofEntomology for their valuable help during my program. Special thanks to

Mildred E. Uegawachi, Lynn H. Hata and Jamie L Wong, secretaries ofthe Department of

Entomology for their hard work that was an essential part ofmy project. My gratitude goes to my parents Vicente Gonzalez Vaquera and Antelia Hernandez de Gonzalez, and

my brothers, especially Alejandro Gonzalez Hernandez who always encouraged me to continue working for a higher education.

v ABSTRACT

The pink pineapple , (Cockerell), and the gray pineapple mealybug, D. neobrevipes Beardsley, (Homoptera: Pseudococcidae) are associated with the disorder mealybug wilt ofpineapple (MWP). High mealybug densities promote the occurrence ofMWP, an important disease limiting pineapple production. In the absence ofants, such as the big-headed , megacephala (Fabricius), mealybug densities are usually low. However, in the presence ofants, mealybug densities on pineapple may reach damaging levels, and ultimately result in the occurrence ofMWP.

Several theories have been proposed to explain the relationships among the and the ants. Studies were undertaken to better define the role ofestablished biological control agents in suppressing mealybug populations in Hawaii.

In surveys in abandoned pineapple fields on the Hawaiian islands ofOahu and

Maui from July 1992 to November 1993, mealybug densities ranged from a mean of22 to

157 mealybugs per plant. P. megacephala was the dominant ant species found. Five natural enemy species, previously introduced into Hawaii for mealybug control, were associated with D. brevipes and D. neobrevipes. Anagynls ananatis Gahan

(: ) was the most common parasitoid associated with D. brevipes.

The only predators that appeared to have any potential were Nephus bilucernarius

Mulsant (Coleoptera: ) and Lobodiplosispseudococci (Felt.) (Diptera:

Cecidomyiidae).

Field evaluation studies using the biological check method suggested that A. ananatis was predominantly responsible for the decline ofD. brevipes densities in the

VI absence ofants. A study combining the biological check method and the paired-cage exclusion technique indicated that in the absence ofP. megacephala, D. brevipes densities were heavily impacted by natural enemies and probably the lack ofsanitation (e.g., honeydew removal). Laboratory studies indicated that P. megacephala significantly decreased mealybug parasitization and by adult A. ananatis and N. bilucernarius by 73 and 48 percent, respectively.

Biological studies were conducted on A. ananatis and N. bilucemarius to define and quantify biological characteristics such as mealybug host and stage preference, developmental times, longevity and fecundity. Additionally, the functional response ofA. ananatis was determined when using D. brevipes as a host species.

VII TABLE OF CONTENTS

Page

Acknowledgments v

Abstract vi

List ofTables xiii

List ofFigures xv

I. INTRODUCTION 1

II. LITERATURE REVIEW 3

Worldwide Pest Status ofMealybugs 3

Mealybugs Infesting and Other Crops 4

Nature ofMealybug Damage to Pineapple 5

Direct Damage 5

Indirect Damage...... 6

Factors Impacting Mealybug Population Ecology 7

Host Plants in Hawaii 7

Honeydew Accumulation 8

Association with Various Ant Species 8

Role ofAnts in Survival ofPM 9

Natural Enemies 10

Current Understanding ofMealybug Wilt ofPineapple Disease 10

Causes ofMealybug Wilt ofPineapple 10

Distribution ofMealybug Wilt ofPineapple 10

VlIl Management ofDysmicoccils Species 14

Management Options and Constraints 15

Chemical Control 15

Physical and Cultural Control 16

Host Plant Resistance 17

Biological Control 17

Interactions ofAnts and Biological Control Agents 19

Control ofAnts to Conserve Natural Enemies 22

Conclusions and Recommendations 23

References Cited 25

III. SURVEY OF THE NATURAL ENEMIES OF THE PINEAPPLE

l\1EALYBUGS 36

Introduction 36

Materials and Methods 37

Results 41

Parasitoids 41

Predators 45

Discussion 49

References Cited 53

IV. IMPACT OF PHEIDOLEMEGACEPHALA (F.) (HYMENOPTERA:

FORMICIDAE) ON THE CONTROL OF DYSMICOCCUS BREVIPES

(COCKERELL) (HOMOPTERA: PSEUDOCOCCIDAE) .. 58

ix Introduction 58

Materials and Methods 60

Biological Check Method 60

Combined Biological Check and Exclusion Method 62

Impact ofP. megacephala on Mealybug Natural Enemies 64

Nephus billicernarius 64

Anagynls ananatis 68

Results 68

Biological Check Method 68

Combined Biological Check and Exclusion Method 75

Impact ofP. megacephala on Mealybug Natural Enemies 79

Nephus billicernarius 79

Anagyrus ananatis 79

Discussion 80

References Cited 85

V. HOST PREFERENCE, REPRODUCTION, AND FUNCTIONAL

RESPONSE OF ANANGYRUS ANANATIS GAHAN (HYMENOPTERA:

ENCYRTIDAE), A PARASITOID OF DYSMICOCCUS BREVIPES

(COCKERELL) (HOMOPTERA: PSEUDOCOCCIDAE) 89

Introduction 89

Materials and Methods 91

Host Species Preference 92

x Host Stage Preference 93

Reproduction, Longevity, and Sex Ratio 93

Functional Response ofAnagynls ananatis 94

Results 95

Host Species Preference 95

Host Stage Preference 95

Reproduction, Longevity, and Sex Ratio 98

Functional Response ofAnagynls anana/is . 98

Discussion 102

References Cited 106

VI. PREY PREFERENCE, AND FECUNDITY OF NEPHUS

BILUCERNARIUS (COLEOPTERA: COCCINELLIDAE), A

PREDATOR OF DYSMICOCCUS BREVIPES (COCKERELL) 110

Introduction 110

Materials and ~.1ethods III

Prey Species Preference 112

Prey Stage Preference 112

Fecundity ofN. bilucemarius 113

Results 114

Prey Species Preference 114

Prey Stage Preference 114

Fecundity ofN. bilucemarius 114

Xl Discussion 116

References Cited 120

VII. CONCLUSIONS 123

Xli LIST OF TABLES

Table Page

2.1. Predators introduced into Hawaii from 1894 to 1936 for controlling

pineapple mealybugs 11

2.2. Parasitoids introduced into Hawaii from 1894 to 1936 for controlling

pineapple mealybugs in the genus Dysmicoccus spp. 13

3. 1. Locations and dates (1992-1993) that Anagyros ananalis was found

associated with pineapple mealybugs and ants in pineapple on the islands of

Oahu and Maui, Hawaii 42

3.2. Locations and dates (1992-1993) that Euryrhopalus propinquus was found

associated with pineapple mealybugs and ants in pineapple on the islands

ofOahu and Maui, Hawaii 44

3.3. Locations and dates (1992-1993) that Lobodiplosispseudococci was found

associated with pineapple mealybugs and ants in pineapple on the islands

ofOahu and Maui, Hawaii 46

3.4. Locations and dates (1992-1993) that Nephus bilucernarius was found

associated with pineapple mealybugs and ants in pineapple on the islands

ofOahu and Maui, Hawaii 47

XIll 3.5. Locations and dates (1992-1993) that Slicololhis nificeps was found

associated with pineapple mealybugs and ants in pineapple on the islands

ofOahu and Maui, Hawaii . 48

4.1. Analysis ofvariance ofthe mean densities ofD. brevipes, A. ananalis

and predators (N. billicernarilis and L. pselldococci) in ant-infested

and ant-free pineapple plots, before ant elimination, and 13, 56, 87, 126,

153, and 188 days after initial ant bait application . 69

4.2. Ratio ofdensities ofthe pink pineapple mealybug to the densities ofthe

encyrtid Anagynls ananatis in ant-infested and ant-free pineapple

plots 76

4.3. Orthogonal contrast ofthe densities ofD. brevipes in big-headed ant-infested

and ant-free cages, where natural enemies were excluded and permitted.

Split plot design experiment 77

5.1. Results ofchoice and no-choice experiments to determine host species

preference ofA. ananatis for D. brevipes and D. neobrevipes 96

5.2. Choice and no-choice experiments to determine host stage preference

ofA. ananatis when reared on the pink pineapple mealybug 97

6.1. Choice and no-choice experiments ofthe predator stage preference ofN.

billicernarius when reared on the pink pineapple mealybug D. brevipes .... 115

XIV LIST OF FIGURES

3.1. Oahu sites surveyed for natural enemies ofpineapple mealybugs .. 39

3.2. Maui sites surveyed for natural enemies ofpineapple mealybugs . 40

4.1. Schematic ofbiological check experiment showing layout ofant- free

and ant-infested plots in a complete randomized block design (B: block).

Located at the Del Monte Pineapple Co., Kunia, Oahu, HI .. 61

4.2. Schematic arrangement ofexperimental pineapple plots in a split plot

design. Exclusion experiment. Del Monte Pineapple Co., Kunia,

Oahu, In 63

4.3. Screen cages used on a field experiment for the exclusion ofnatural

enemies ofD. brevipes in the presence or absence ofants. PPP: potted

pineapple plant; a: closed cage type with a door; b: open cage type 65

4.4. Arena cages used for observing the impact ofants on the activity ofthe

natural enemies ofD. brevipes. A: ant colony nest box; B: arena cage

for the introduction ofnatural enemies, ants and mealybugs; C: arena

cage control for the introduction ofnatural enemies and mealybugs;

D: lateral view ofan arena cage; P: pineapple leaf 66

4.5. Mean number ofhealthy pink pineapple mealybugs (all stages except

crawlers) per pineapple plant (aerial part) in ant-infested and ant-free

experimental pineapple plots 71

xv 4.6. Mean numbers ofthe encyrtid Anagynls anana/is reared from pineapple

mealybugs per pineapple plant (aerial part) in ant-infested and ant-free

experimental pineapple plots 72

4.7. Percent ofparasitization ofDysmicocclis brevipes by Anagynls anana/is

per pineapple plant (aerial part) in ant-infested and ant-free experimental

pineapple plots 73

4.8. Mean number ofthe predators Lobodiplosis pselldococci and Nephus

bilucemarius per pineapple plant (aerial part) in ant-infested and

ant-free experimental pineapple plots 74

4.9. Mean numbers ofthe pink pineapple mealybugs per pineapple plant

in ant-infested and ant-free pineapple plots, and in the presence and

absence ofnatural enemies (NE) 78

5.1. Reproduction ofAnagynls anana/is when reared on Dysmicocclls

brevipes: A) daily mean number ofparasitized mealybugs; B) female

and male parasitoid 99

5.2. Linear regression ofthe 1/(ln ofthe mean proportion ofDysmicoccus

brevipes hosts surviving parasitization) by Anagynls anana/is as a

function ofhost density 100

5.3. Functional response ofthe observed and expected numbers of

DysmicocClis brevipes hosts attacked by Anagyrus anana/is as host

density increased 101

XVI 6.1. Mean daily fecundity ofthe coccinellid predator Nephus billicernarilis

(n = 7) reared on Dysmicoccus brevipes 117

XVll I. INTRODUCTION

Pineapple mealybugs, Dysmicoccils spp. (Homoptera: Pseudococcidae), are

serious pests almost everywhere pineapple is grown (Bartlett 1978). Pineapple plants

heavily infested with these mealybugs may suffer growth depression and wilt (Carter

1967). On developed fruit, mealybugs cause direct damage and cannery problems

(Beardsley 1993). The pink pineapple mealybug (pPM), Dysmicoccus brevipes

(Cockerell), has been associated with the widely distributed disorder called mealybug wilt

ofpineapple (MWP) (Beardsley 1959, Carter 1962). Another mealybug species, the long

tailed mealybug, Pselldococcus longispilms (Targioni), is occasionally associated with

MWP (Carter 1966, 1967; Petty 1978). Several ant species are associated with these PM

including the big-headed ant, (Fabricius), the argentine ant,

Liniepithema humile (Mayr), and the fire ant, Solenopsis geminata (Fabricius)

(Illingworth 1931, Carter 1967, Beardsley et al. 1982, Reimer et al. 1990). Some ant

subfamilies have a trophobiosis association with mealybugs and other homopterans based

on their consumption ofhoneydew excreted by the homopterans (Holldobler and Wilson

1990). Honeydew is rich in sugars as well as some amino acids and waxes (Heidari and

Copland 1993). Ants may protect mealybugs and other Homoptera from natural enemies

(Illingworth 1931, Carter 1933, Flanders 1951, Nixon 1951, Bess 1958, Way 1963, Jahn and Beardsley 1994) and may increase mealybug survivorship by removing mealybug excretions (Flanders 1951, Bess 1958). This is thought to be the basis for the mutualistic association among these species. In Hawaii, the big-headed ant is the dominant species in pineapple fields (Reimer et al. 1990, Rohrbach et al. 1988) where they are strongly associated with PPM and GPM.

Effective control ofMWP has been obtained when ant densities were maintained at low levels (Beardsley et al. 1982, Rohrbach et al. 1988). For example, when P. megacephala were eliminated, PM populations rapidly decreased (Carter 1967, Rohrbach et al. 1988, Petty and Tustin 1993). Use ofinsecticides , such as DDT, cWordane, and mirex, to control ants associated with PM had a detrimental effect on non-target organisms, and, therefore, were banned by the Environmental Protection Agency (Reimer and Beardsley 1990). Chemical ant control is currently obtained by the application of hydramethylnon in a bait formulation (Amdro ®) under an emergency local needs registration.

To provide effective, low cost control ofPM in Hawaii, a classical biological control project was initiated in the 1930's (Chapman 1938). However, the role and effectiveness ofintroduced natural enemies in suppressing PM was not evaluated.

The main objectives ofthis work were to: 1) evaluate the role ofpreviously introduced natural enemies in suppressing D. brevipes and D. neobrevipes; 2) evaluate the impact ofP. megacephala on the activity ofeffective natural enemies ofPM; and 3) determine important biological parameters ofthe encyrtid Anagyros ananatis Gahan and the coccinellid Nephus bilucernarius Mulsant relative to mass rearing these species.

2 II. LITERATURE REVIEW

Worldwide Pest Status ofMealybugs

Mealybugs (Homoptera: Pseudococcidae) are among the most economically

important pests (Zimmerman 1948). According to Williams and Granara de Willink

(1992), many ofthe 2000 described mealybug species are pests ofimportant crops. The

genera Pseudococcus, Phenacoccus, and Planococcus include many ofthe most

significant species ofagricultural importance (Bartlett 1978). Agricultural crops attacked

by mealybugs include cassava, potatoes, tomatoes, peppers, citrus, soybeans, beans,

coffee, cocoa, sugarcane, mango, banana, and other tropical and subtropical fruits

(Williams and Granara de Willink 1992). Mealybugs also attack many ornamental crops

(Bartlett 1978). Mealybugs frequently increase to high numbers, killing the host plant by

consuming the sap or by injecting toxins, and transmitting viruses (Williams and Granara

de Willink 1992). Excreted honeydew that accumulates on plant foliage or fruit is a

suitable medium for the growth ofsooty mold that causes reduction ofplant

photosynthesis rates (Bartlett 1978). Several mealybug species ofeconomical importance

have a mutualistic association with ants. In this association, ants feed on the mealybug

excreted honeydew, while mealybug survival is enhanced by activities ofants including

protection from natural enemies (Nixon 1951, Way 1963).

Some mealybugs have been associated with the transmission ofviruses inducing plant diseases. For example, Planococcoides njalensis (Laing) is the most important vector ofswollen shoot ofcocoa in Africa (Strickland 1951). Planococcus citri (Risso)

3 and Ferrisia virgata (Cockerell) are also vectors ofthis disease (Lamb 1974). In

Trinidad, the Cocoa Trinidad virus is transmitted by Dysmicoccus alazon Williams,

Dysmicocclis brevipes (Cockerell) ,F virgata and Planaococclis minor (Maskell)

(Williams and Granara de Willink 1992).

The majority ofmealybugs ofeconomic importance are introduced species

(Zimmerman 1948). These mealybugs are difficult to control in part because they lack

efficient natural enemies that maintained mealybug populations at low levels in their areas

oforigin.

Mealybugs Infesting Pineapple

In Hawaii the most important introduced pseudococcids attacking pineapple are

the pink pineapple mealybug, Dysmicocclis brevipes (Cockerell), and the gray pineapple

mealybug, D. neobrevipes Beardsley. Earlier works recognized Pseudococclis brevipes

(Cockerell) as the major mealybug species in pineapple (Zimmerman 1948), but Ferris

(1948) later renamed it Dysmicocclis brevipes (Cockerell). D. brevipes was well known

in other pineapple growing areas ofthe world and had been considered to include two

distinct types, the gray and the pink forms. Later, Beardsley (1959) reviewed the taxonomic status ofthe Dysmicoccus group in Hawaii, and based on consistent morphological differences between the pink and the gray forms, he proposed the name of

D. neobrevipes for the gray form. Pineapple is their common host, and they are associated with mealybug wilt ofpineapple (Beardsley 1993). Other mealybug species are present in pineapple fields, but they are not considered important because they occur sporadically in

4 low numbers (Carter 1967). Ofall described mealybug species present in Hawaii, D. hrevipes and D. neohrevipes are the most difficult to manage using natural enemies because oftheir strong association with ants (Bartlett 1978). Even when S. sacchari is associated with ants, the negative impact the ants have on the mealybugs' natural enemies is minimal ifany (Fluker et al. 1968).

Nature ofMealybug Damage to Pineapple

Direct Damage. The type ofdamage to pineapple depends on the mealybug species infesting a plant. Feeding by D. neobrevipes causes green spotting on the pineapple leaves. This is a localized toxic effect that develops around the point offeeding

(Carter 1933, Serrano 1934, Petty 1978). Green spotting on leaves is characterized by the presence ofan outer green circle with a darker green center (Carter 1967). Feeding by D. brevipes causes only chlorotic areas in response to the depletion ofthe plant cell

(Zimmerman 1948). According to Carter (1962), growth depression is a systemic effect of mealybug feeding, and is best expressed in small seedlings, although heavily infested ratoon plants may show growth depression. Feeding ofboth Dysmicoccus species on pineapple leaves produces a toxic effect called mealybug stripe. This damage is caused by their saliva, which is toxic to the chloroplast and in time results in striped areas (green or blackened areas) (Carter 1967). Mealybugs that feed on or inside marketable fruit reduce product quality based on their presence (cosmetic damage). Mealybug infested fruit may result in the implementation ofquarantine restrictions on fresh fruit shipments (Beardsley

1993).

5 Indirect Damage. Feeding activity ofD. brevipes and D. neobrevipes induces the expression ofmealybug wilt ofpineapple (MWP). During the 1920's, many areas of pineapple in Hawaii were destroyed by this disease (Illingworth 1931). By 1930, MWP was considered the most wide spread and destructive pineapple disease in almost every pineapple producing country (Illingworth 1931). The rapid decline ofthe pineapple industry in Florida was attributed to wilt (Illingworth 1931).

Carter (1933) first divided MWP into two categories: "quick wilt" and "slow wilt."

He described quick wilt as a result ofa sudden infestation ofmealybugs, which fed for a short period. Recovery from this wilt type is common and results in production ofsmall fruit. In contrast, slow wilt is caused by a large mealybug population feeding over a long period. There is recovery from this stage ifrapid control measures are taken. Carter

(1967) later concluded that slow wilt was less common due to effective ant control.

The most important symptom ofMWP is drying and wilting ofthe leaves that begins at the leaftips, which turn a reddish-yellow color (Singh and Sastry 1974).

According to Carter (1 945a, 1962, 1967), above ground symptoms ofMWP include four stages: 1) preliminary reddish coloration ofthe leaves; 2) definite color change from red to pink, with an inward inflection ofthe leafmargins; 3) light reflection from leaf surfaces becomes pronounced; and 4) tips ofaffected leaves become desiccated. After the last stage, a recovery stage may occur, with fresh, apparently normal leaves growing out from the center ofthe plant. This stage was called "terminal wilt, normal center" (Carter

1945b, 1962, 1967). MWP includes collapse ofthe roots resulting in their death, and the initial appearance ofleafwilt symptoms (Carter 1948).

6 Illingworth (1931) observed that wilt disease was associated with the feeding

activity ofD. brevipes, and symptoms were seen only after D. brevipes fed upon wilting

plants. Carter (1933) conducted several experiments that supported Illingworth's idea

about the relationship between mealybugs feeding on pineapple plants and occurrence of

wilt. Additionally, Carter (1933, 1941, 1945a, 1951) first proposed that MWP was

caused by the injection ofa toxin during the mealybugs' feeding activity. Later, Carter

(1962, 1963) rejected his original hypothesis and suggested that in MWP, mealybugs

transmitted a latent factor (thought to be a latent virus), which was transmissible and

retained in a vegetative plant. Although some closterovirus particles have been isolated

(Gunashine and German 1989) and serologically detected in wilted pineapple plants

(Ullman et aI. 1989), the etiological significance ofMWP is not yet known (Ullman et aI.

1993). More recently, Wakman et al. (1995) found clostero-like virus particles from wilted and asymtomatic pineapple plants. According to them, these virus particles are similar in appearance to those found in Hawaii. They also found bacilliform virus particles in the same MWP affected plants which may be related to the badnavirus group. These badnavirus have also been found in MWP affected plants in Hawaii (Hu, Sether and

Ullman, unpublished data).

Factors Impacting Mealybug Population Ecology

Host Plants in Hawaii. D. Ileobrevipes has been recorded on 29 plant species in

Hawaii (Beardsley 1959, 1965; Nakahara 1981). It commonly occurs on pineapple, sisal

(Beardsley et al. 1982), and banana, as well as on other host species, but has not been

7 detected in grasses (Beardsley 1959). In contrast, D. brevipes commonly occurs on

pineapple and perennial grasses (including sugarcane), and has been recorded on more

than 50 plant species (Zimmerman 1948, Nakahara 1981, Rohrbach et al. 1988), including

groundnut (Williams 1985, Singh et al. 1988). In Hawaii, these two species a have few

plant hosts in common, such as Coccolobo lIvifera (L.), Colocasia esculenta (L.) and

Musa spp. (Nakahara 1981).

Honeydew Accumulation. Accumulation ofhoneydew may promote sooty mold

(Capnodillm sp.) growth and coccoidea crawlers may become stuck in excess honeydew

(Carter 1933, Flanders 1951, Way 1951, Bess 1958, Bartlett 1961, Petty 1978, Rohrbach

at aI. 1988). Other types offungi may grow in the honeydew and later may envelop and

destroy honeydew-producing (Way 1955).

Association with Various Ant Species. Phillips (1934) found 28 ant species in

pineapple fields in Hawaii, but he considered only P. megacephala and the fire ant,

Solenopsis gemf:;ata (Fabricius), as closely associated with PM. Later, Carter (1967)

included the crazy ants, Paratrechina bOllrbonica Forel, P. longicomis Latreille, and P.

vaga (Forel) (mis-identified as P. sharpii Forel, not established in Hawaii), the sugar ant,

Plagiolepis alluaudi Forel, and the Argentine ant, Linepithema humile (Mayr). Ofthese

ant species, P. megacephala and S. geminata are most closely associated with PM (Carter

1967). According to later studies ofReimer et al. (1990), only P. megacephala, S.

geminata, and L. hllmile are considered pestiferous (as defined by their PM association) in

pineapple fields. The first named species is the dominant ant in Hawaiian pineapple fields, the majority ofwhich are situated below 600 m elevation (Reimer et aI. 1990). The

8 impact ofP. megacephala is not restricted to agricultural plants alone, it also affects drip irrigation systems (Chang and Ota 1990), the urban environment, and native Hawaiian fauna (Reimer et al. 1990, Gillespie and Reimer 1992, Reimer 1993).

Role ofAnts in Survival ofPM. The sanitation hypothesis theorizes that ants remove and feed upon Homoptera produced-honeydew resulting in increased survival of the Homoptera (Flanders 1951, Bess 1958). However, laboratory studies conducted by

Jahn (1992) indicated that P. megacephala did not reduce D. neobrevipes mortality by removing honeydew. However, his study was short term and mealybugs densities in both his control and experimental treatments decreased from initiation ofthe study. In 1931,

Illingworth proposed that P. megacephala had a beneficial influence on the survival and distribution ofthe pineapple mealybugs. Carter (1932) also suggested that P. megacephala was the main factor in mealybug movement into new pineapple fields. Data ofJahn (1992) refuted this theory, and suggested thatD. neobrevipes was mainly dispersed by wind. Whatever the beneficial effect (protection and/or sanitation) ofP. megacephala upon D. neobrevipes, Carter (1933) concluded that a lower incidence of

MWP resulted from the absence ofP. megacephala and S. geminata in the pineapple fields.

Carter (1960) first established the PM/ant wilt relationship when he observed reduction ofa population, and then low incidence ofwilted plants in an ant-free field.

Unfortunately, this experiment did not satisfactorily prove the relationship because he did not compare mealybug populations ofan ant-free field with an ant-infested field. More convincing evidence supporting the relationship was obtained by Su (1979) and Beardsley

9 et a1. (1982). They observed that in plots with excellent ant control, mealybug wilted plants were not found. While in plots with ants, the number ofwilted plants increased during the second crop.

Natural Enemies. When mealybug natural enemies were introduced into Hawaii from Australia, South East Asia, and Central and South America, only a few became established (Carter 1945a, Zimmerman 1948, Bartlett 1978) (Tables 2.1 and 2.2).

Following introduction ofthese natural enemies, no evaluation studies were conducted, although mealybug decline was commonly observed in the absence ofants and pesticide residues.

Current Understanding ofMealybug Wilt ofPineapple Disease

Causes ofMealybug Wilt ofPineapple. MWP is the most important disease of pineapple where it is commercially grown (Carter 1963). Three mealybug species

(Homoptera: Pseudococcidae) have been associated with MWP with the ranking of importance (greatest to least) being: D. neobrevipes, D. brevipes (Beardsley 1993), and the long-tailed mealybug, PselidococCliS longispilllls (Targioni) (Carter 1966, 1967, Petty

1978, Rohrbach et ai. 1988). With respect to P. longispillus, no evidence has been found to associate this species with pineapple wilt in the field (Carter 1966, 1967; Petty 1978).

However, Carter (1966) was able to induce wilt symptoms on test pineapple plants using

P. longispinlls (= Pseudococclis adOllidllnl (L.», in laboratory experiments.

Distribution ofMealybug Wilt ofPineapple. MWP occurs in practically all the tropical areas ofthe world where pineapple is commercially grown (Carter 1962).

10 Table 2.1. Predators introduced into Hawaii from 1894 to 1936 for controlling mealybugs.

Year Species Family Location Reference introduced found 1893 Hyperaspis comparatus Casey. Coccinellidae California Chapman 1938 " Hyperaspis depressa Casey " " " " Hyperaspis postica Leconte " " " " Scymnussp. " " " 1894 Hyperaspis undulata Say " " " 1894 Cryptolaemus mountrouzieri " Australia Swezey 1938 Mulsant' 1894 Rhizobium ventralis Erichson " Australia Chapman 1938 1895 Hyperaspis japonica " California " 1895 Nephus bipul1ctatlls Kugel " Japan " ,. 1894-95 Sticholotis nificeps Weise Japan and Giffard 1906 China 1896 Hyperaspis lateralis Mulsant " California " 1904 Orcus lajertei Mulsant " Australia " 1906 Hyperaspis limbatus Casey " California " 1906 N. bipunctatus " South China " 1907 Hyperaspis 8-notata Casey " Mexico " 1914 N. bipllnctatus " Philippine " 1922 Nephlls sp. " Mexico 1922 Diomus sp. " " " 1922 Hyperaspis sylvestrii Weise " " " 1922 Diomus sp. 2 " " " 1922 Scyml1us (Pul/us) uncinatus " " Funasaki at al. Sicard' 1988 1924 Hyperasips albicoWs Gorh " Panama Chapman 1938 " Cleothera bromelicoa Sicard " " " Diomus margipal/ens " " " 1924 Seymnus pietus Gorh " Panama Chapman 1938 " S. (pullus) uncinatus ' " " " " Pseudiastata nebulosa Drosophilidae Mexico " Coquillet 1929 Sympherobills barberi Hemerobiidae Mexico " (Banks) , 1930 Lobodiplosis pseudoeocci CecidomyiidaeMexico Swezey 1938 Felt 1 1930 Nephlls bilucemarius Coccinellidae Mexico Funasaki et al. Mulsant' 1988 , Successfully established in Hawaiian pineapple fields.

11 Table 2.1. (Continued). Predators introduced into Hawaii from 1894 to 1936 for controlling mealybugs.

Year Species Family Location Reference introduced found 1931-37 P. nebulosa Coquillet Drosophilidae British Swezey 1938 Guyana, Guatemala, Panama & Trinidad 1933 Diomusfutahoshii Ohta Coccinellidae Formosa Chapman 1938 1935 Dicrodiplosis guatemalensis Guatemala Swezey 1939 Felt 1935-36 Hyperaspis nigrum Mulsant Coccinellidae Brazil " 1936 Hyperaspis quinqllenotata " " Chapman 1939 Mulsant 1938 Scymnus apicij/avlIs Motsch " Malay Swezey 1939

12 Table 2.2. Parasitoids introduced into Hawaii to control pineapple mealybugs in the genus Dysmicocclis spp.

Year Species Family Location Reference introduced found 1932 Zaplatycenlsful/awayi Encyrtidae Panama Swezey 1938 Timberlake 1935 Synaspidia pretiosa " Guatemala " Timberlake 1935 Aenasills cariococcliS " Colombia Chapman 1938 Compere. 1935 Aenasius colombiensis " Colombia Swezey 1939 Compere 1935-36 A. colombiensis " Colombia & Chapman 1938 Venezuela 1935-36 Anagyros ananatis Gahan 1,2 Encyrtidae Brazil Carter 1937 1935 Hambeltonia pseudococcina " Brazil " Compere 1935 H. pselldococcina 1 " Colombia & " Venezuela 1936 ElIryrhopallis propinqlllls " British Funasaki et al. Kerrich 1 Guyana 1988 1938 Anagyros aurantifrons Encyrtidae Africa Chapman 1938 Compere 1954 Pselldaphyclls allgeliclis " California Weber 1955 (Howard)

1 Successfully established in Hawaii pineapple fields. 2 Formerly mis-identified as Allagyros coccidivoros Dozier.

13 However, MWP is not equally important wherever pineapple is grown, because there are

several species ofPM involved and these may differ in their ability to cause wilt

(Beardsley 1993). This disease has been recorded in Hawaii (Illingworth 1931, Carter

1932); Jamaica and Central America (Carter 1934, 1949); Mexico (Trejo personal

communication); Puerto Rico (plank and Smith 1940; Carter 1949); Brazil (Carter 1949);

Florida (Carter 1933); Australia, Fiji, Malaya and Philippine Islands (Carter 1942); India

(Singh and Sastry 1974); Cote D'Ivoire (Guerout 1972); East and South Mrica (Carter

1942; Petty 1978, Petty and Tustin 1993).

Management ofDysmicoccus Species

D. brevipes and D. neobrevipes differ in their biology (type ofreproduction), hosts

and area ofthe plant that they attack (feeding behavior), and their associated internal

symbionts (Ito 1938, Beardsley 1959, Carter 1935a and 1967). Ito (1938) conducted a

life-cycle experiment using the gray strain and the pink strain ofP. brevipes. He observed

that the gray strain produced males, while the pink strain did not (parthenogenetic).

Although, this finding was true for Hawaii, it is not always so in other areas, such as

tropical America, where there is a bisexual form morphologically identical to Hawaiian D.

brevipes females (Beardsley 1993). According to Ito (1938), the gray strain produced a

mean of347 progeny, with a mean female life span of95 days. The pink strain produced a mean of234 progeny, and had a mean female life span of90 days. With respect to the internal symbionts, the gray strain has symbionts that are not present in the pink strain

(Carter 1935a, 1967). Another difference between these two species is that D.

14 neobrevipes feeding causes a symptom known as green spotting on the pineapple leaves.

On the other hand, D. brevipes produces only yellow chlorotic spots (Carter 1935a,

1967).

D. neobrevipes prefers to feed on the aerial part ofits host plants. In pineapple, it is found on lower leaves, slips, fruit, and crowns (Beardsley 1959, Carter 1962, Py et ai.

1984, Rohrbach et al. 1988). On pineapple fruit, D. neobrevipes can enter into the blossom cups (Petty 1978, Jahn 1992). Petty (1978), and Rai and Sinha (1980) observed

D. brevipes entering the blossom cups ofpineapple fruit in South Africa and Guyana, respectively.

Management Options and Constraints. Several control strategies have been suggested for direct control ofmealybugs and wilt. These include field insecticide applications (Carter 1967), chemical protection ofplanting material (Macion 1978, Rai and Sinha 1980), thermotherapy ofplanting material for mealybugs and wilt (plank and

Smith 1940, Singh and Sastry 1974, Ullman et al. 1991, Ullman et al. 1993, Hu ei al.

1995) and biological control ofmealybugs (Rohrbach et al. 1988, Jahn 1992). Biological control ofants tending mealybugs has had little attention, perhaps because few natural enemies ofants are known (Holldobler and Wilson 1990), and those studied have not been highly efficient regulators ofants (Jouvenez 1986).

Chemical Control. Oil sprays were first used to control PM, but were abandoned at the end ofthe World War II when DDT became available for ant control (Carter 1967).

After the development ofDDT, organic phosphate pesticides, such as parathion, malathion

15 and diazinon became available for mealybug control. According to Carter (1967),

diazinon was more effective at lower rates than parathion and malathion.

Treatments were made on planting material and (or) as field applications for direct

control ofPM. For example, malathion, diazinon (Macion 1978) and ometoate (Rai and

Sinha 1980) were effective when used as pre-plant dips for planting material to eliminate

any mealybugs infesting the planting material. In post-plant foliar application, malathion

(Macion 1978) and phoxim (Rai and Sinha 1980) were effective for control ofresidual

mealybug populations on field plants.

Physical and Cultural Controls. Physical and cultural controls have received

little attention in Hawaii. Some types could be implemented in integrated management

programs for PM. The "ant fence" was the first physical control technique successfully

used for ant control in pineapple in Hawaii. The ant fence consisted ofa 0.3 m wide board

sunk 0.15 m into the ground edgewise. This fence was regularly sprayed with a heavy oil residue (Carter 1967). Guard rows ofpineapple planted parallel to the border around the field were used commonly in Hawaii pineapple fields until 1950 (Rohrbach et al. 1988).

These guard rows reduced both ant and mealybug movement into new pineapple beds

(Carter 1967).

Plank and Smith (1940) satisfactorily eliminated D. brevipes infestations on pineapple slips (planting material) using thermotherapy with hot air, at 46° C for 6 hr.

Thermotherapy ofplants infected with mealybug wilt has been successfully applied to obtain apparent virus-free pineapple planting material (Singh and Sastry 1974, Ullman et al. 1991, Ullman et al. 1993, Hu et al. 1995). Ullman et al. (1991) proposed that the

16 application ofhot water (at 40° C for 60 min. or 50° C for 30 min.) on a large scale could increase pineapple fruit yields.

Illingworth (1931) suggested the maintenance ofclean culture along field borders and the burning ofold or dead pineapple plants provided cultural control ofPM. Plank and Smith (1940) recommended that instead offollowing a common fallow, pineapple plants should be removed and the land should be planted with a soil-improving crop (i.e., legumes) unsusceptible to mealybugs, before replanting with pineapple.

Host Plant Resistance. Breeding programs for the development ofmealybug wilt resistance have not been a successful alternative because some hybrids thought to be resistant have expressed only partial resistance, and some ofthese were susceptible to wilt

(Carter and Collins 1947, Carter and Ito 1956, Williams and Fleisch 1993). Collins and

Carter (1954) found three wilt-resistant hybrids ofCayenne, but none oftheir progenies survived the re-selection process. The most resistant hybrid, Cayenne-Queen hybrid 7898, expressed segregation ofthe resistant characters with susceptibility being dominant. This hybrid showed only mild leafsymptoms, but in heavy PM infestations the symptoms were severe (Carter and Collins 1947). At that time, wilt screening was eliminated from the breeding program because ofthe excellent ant control obtained with chlorinated hydrocarbon insecticides that greatly reduced MWP (Williams and Fleisch 1993).

Biological Control. From 1894 to 1936 many introductions ofpredators (Table

2.1) and parasitoids (Table 2.2) ofmealybugs were made into Hawaii (Chapman 1938,

Funasaki et al. 1988). By 1930,34 parasitoids and predators ofmealybugs were introduced, ofwhich eight were established but without appreciable effectiveness

17 (Chapman 1938). Introductions into Hawaii were more successful from 1930 to 1936

(Chapman 1938, Bartlett 1978). Species such as the predator Lobodiplosis pseudococci

(Felt) (Diptera: Cecidomyiidae), from Mexico (Chapman 1936, Carter 1944), the encyrtid parasitoids Anagynls ananatis Gahan (formerly mis-identified as A. coccidivonls Dozier),

Hambeltonia pselldococcina Compere (Compere 1936, Carter 1937), andEuryrhopailis propinqulls Kerrich (Bartlett 1978) from South and Central America, were established and controlled pineapple mealybugs in the absence ofants (Carter 1945a, Zimmerman 1948,

Bartlett 1978). Among those natural enemies, L. pselldococci was the most widely established ofany ofthe introduced species (Chapman, 1936; Carter, 1944). Another natural enemy, the coccinellid predator Nephus billicemarilis Mulsant, was introduced from Mexico in 1930 and successfully established (Swezey 1939, Funasaki et al. 1988).

Carter (1967) pointed out that natural enemies introduced in Hawaii may work effectively on remaining small mealybug colonies after ant elimination.

In 1935, Carter (1935a) found that Pseudiastrata nebulosa Coquillet (Diptera:

Drosophilidae) was the only effective predacious biological control agent in Guatemala and Jamaica. P. nebulosa was previously introduced into Hawaii from Mexico in 1924, but it did not establish (Carter 1935b, Chapman 1938). In Puerto Rico in 1936 and 1937,

Bartlett (1939) made releases ofA. coccidivonls imported from Brazil, and H. pseudococcina from Brazil and Venezuela via Hawaii, against D. brevipes. Only H. pseudococcina was able to establish in Puerto Rico. H. pselldococcina also was introduced and established in southern Florida in 1944 from Puerto Rico to control D. brevipes present on small pineapple plantings, but no information exists regarding its

18 abundance or effectiveness (Clausen 1956). Carter (1949) found that the parasitoids

PseudaphyclIs sp. and Anogynls sp. probably reduced infestation ofD. brevipes on pineapple slips in Brazil. Hence, a potential still remains for more introductions ofnatural enemies from South America, the site oforigin ofPM. Furthermore, very little is known about the biology ofthese introduced natural enemies. Only visual observations and informal reports exist about the host feeding behaviors and host or prey preferences ofthe natural enemies.

Interactions ofAnts and Biological Control Agents. Several ant species, particularly members ofthe most advanced subfamilies, the Myrmicinae, the

Dolichoderinae, and the Formicinae, have a trophobiotic relationship with certain

Homoptera and play an important role in their survival and development (Nixon 1951,

Holldobler and Wilson 1990). This mutualistic relationship between ants and homopterans, particularly in and scale insects, has been well reviewed by Nixon (1951) and Way

(1963). Ants remove and feed on the honeydew excreted by the Homoptera (Holldobler and Wilson 1990). The ants protect Homopterous colonies from natural enemies, and this behavior has been proposed or demonstrated in some Coccidae and Pseudococcidae species (Illingworth 1931, Carter 1933, Flanders 1951, Nixon 1951, Bess 1958, Bartlett

1961, Way 1963, Fluker 1969, Petty 1978, Nechols and Seibert 1985, Rohrbach et al.

1988, Hanks and Sadof 1990, Jahn 1992, Jahn and Beardsley 1993, Reimer et al. 1993).

However, other hypotheses have been proposed to explain the homopterous-ant relationship. The sanitation hypothesis proposes that ants remove honeydew that is deleterious to mealybug survival. Another hypothesis is that ants aid in mealybug

19 dispersal to new plants (Illingworth 1931, Carter 1932, Phillips 1934, Plank and Smith

1940, Way 1963, Petty 1978).

Effectiveness ofnatural enemies in controlling scale insects after ant removal has been evaluated for some Coccidae (Bartlett 1961, Reimer et aI. 1993) and Pseudococcidae

(Fluker et aI. 1968, Nechols and Seibert 1985, Jahn 1992, Jahn and Beardsley 1993) species. Bartlett (1961) found that in the absence ofL. hZ/mile, the coccids Saissetia o/eae (Bern.) and Coccus hesperidum (L.) were effectively controlled by Metaphycus he/vo/us Compere and M /uteo/us Timberlake, respectively. Reimer et al. (1993) observed that P. megacepha/a had a beneficial effect on (Green) survival by interfering with or preying on natural enemies, mainly on the coccinellid larvae Azya

/uteipes Mulsant, Curillus coentlells Mulsant, Crypto/aemus mOl/ntrouzieri Mulsant and

Rhizobius ventra/is (Erichson). They observed that in the absence ofants, natural enemies, primarily larval coccinellids, eliminated C. viridis from coffee trees. Reductions ofparasitism from 27.4 to 98.4 % occurred in the presence ofants. With regard to the pink sugarcane mealybug, Saccharicocclis sacchari (Cockerell), Fluker et aI. (1968) found that the presence ofP. megacepha/a had a beneficial effect on populations ofthis mealybug, but it was not detennined ifthis effect was because ofa suppressive effect of the ants on the host-specific encyrtid parasitoid Al1agynls saccharico/a (Timberlake), or some other factor. Nechols and Seibert (1985) observed reduction ofpercentage parasitism ofNipaecocclis vastator (Maskell) by AllagyntS illdicus Shafee when the ant

Technomyrmex a/bipes F. Smith was present. Jahn (1992) found evidence that when pineapple fields were free ofP. megacepha/a, there was an increase in populations of

20 generalist natural enemies. The only natural enemies he found in experimental pineapple

fields on the island ofMaui, Hawaii, were the coccinellid predator Curinus coerilleus

Mulsant and an undetermined spider species. Parasitism was not detected. He suggested

the necessity for identifying the most efficient natural enemies ofD. neobrevipes to

implement them in augmentation programs in order to decrease insecticide use.

In most experiments, ants negatively impacted the effectiveness ofnatural enemies

ofscale insects by interfering or killing them. However, some natural enemies were not

sensitive to the presence ofants. For example, the introduction into California ofthe

parasitoid Tetracnemus pretioslis Timberlake, reduced the abundance ofthe mealybugs

Pseudococcus gahni (Green) and P. aurailanatlls (Maskell), in spite ofthe ants being attracted to their honeydew (Flanders 1958). In most ofthe above examples ofants disrupting their associated Homoptera, the alleviation ofant interference with natural enemies, thereby conserving natural enemies, is recommended via reduction ofant densities to non-significance levels. Steyn (1954) indicated that the use ofsticky bands around trunks ofcitrus trees may keep out P. megacephala from the canopy, in order to obtain commercial control ofCalifornia red scale, Aonidiel/a aurantii (Maskell) and the soft scale, Coccus hesperidllm L., by means oftheir natural enemy activity (e.g. Aphytis chrysomphali Mercet). Although, A. allrantii does not produce honeydew, it is indirectly protected by the presence ofants tending honeydew-producing soft scales. A similar suggestion was made by Reimer et al. (1993) relative to green scale, Coccus viridis

(Green), on coffee trees. They suggested the use ofpolyester fiber/Tanglefoot bands to eliminate P. megacephala from the tree canopies. Bartlett (1961) stated that before an

21 ant-intolerant natural enemy is released against scale insects it would be desirable to first

eliminate ants. In Africa, successful biological control against the cassava mealybug,

Phenacoccus manihoti Matile-Ferrero, was obtained by the introduction ofthe encyrtid

Epidinocarsis lopezi (De Santis) (Cudjoe et al. 1993). Unfortunately it was observed that

ants in the genera Campollotus, Crematogaster and Pheidole had a negative effect on the

biological control ofthe cassava mealybug, and it was proposed to control these ants.

Control ofAnts to Conserve Natural Enemies. Before mirex and heptacWor

were banned in the United States, ants in pineapples were effectively controlled and MWP

was not a problem in Hawaii (Beardsley et aI. 1982, Reimer and Beardsley 1990, Reimer

et aI. 1991, Rohrbach and Apt 1986). Afterwards, new insecticide bait products were

tested and recommended for P. megacephala in pineapple fields. Su (1979), and Su et al.

(1980) found the compound hydramethylnon (Amdro®) tested in baits, as a good

candidate to replace mirex for pineapple ant control. Similarly, Reimer and Beardsley

(1989, 1989a, 1990) obtained more effective control ofP. megacephala in pineapple fields using commercial Amdro® (hydramethylnon) bait, compared to Logic®

(fenoxycarb) treatment. Today Amdro® provides good ant control in pineapple fields when it is used in combination with tillage at the time ofthe field preparation, and in applications along field edges and borders when needed (Reimer and Gonzalez-Hernandez

1993). Fenoxycarb and pyriproxyfen produce insect growth regulator activity (IGR) on P. megacephafa and S. illvicta. These materials have potential use as insecticide baits because they cause a cessation ofegg production, high mortality against irnmatures, and

22 shift in caste differentiation from workers to males (Glancey et al. 1990, Reimer et al.

1991).

Although products to control ants are available presently and have been annually approved for use, pineapple growers are concerned that those products may be suddenly removed from the market place. Presently, research to find alternatives to chemical baits is underway.

Conclusions and Recommendations

With the advent ofsynthetic pesticides, emphasis on developing biological and cultural controls for Dysmicocclis spp. in pineapple declined. Efforts continued to elucidate the relationship between mealybug densities and the occurrence ofMWP.

However, several avenues exist to further exploit and improve biological control.

Only anecdotal observations exist with respect to the presence and efficacy of established natural enemies ofDysmicocclis spp. Information is needed on the distribution and mealybug associations ofnatural enemies in pineapple plantings. Additionally, interactions between common ant species and natural enemies need to be elucidated.

Basic information on natural enemy biology will provide the groundwork for potential development ofaugmentation practices.

Given that the last introductions ofnatural enemies were made prior to the recognition ofD. brevipes and D. neobrevipes as distinct species, it is possible that species specific natural enemies may be sought for both ofthese species, especially given the

23 significant differences in their host plant ranges. Additionally, few natural enemies were introduced from South America the theorized site oforigin ofthese species.

24 Reference Cited

Bartlett, B. R. 1961. The influence ofants upon parasites, predators, and scale insects.

Ann. Ent. Soc. Amer. 54: 543-551.

Bartlett, B. R. 1978. Pseudococcidae, pp. 137-170. In C. P. Clausen (ed.), Introduced

parasites and predators ofarthropod pest and weeds: a world review. U. S. Dep.

Agric. Handb. No. 480.

Bartlett, K. 1939. Introduction and colonization oftwo parasites ofthe pineapple

mealybug in Puerto Rico. 1. Agric. Univ. ofPuerto Rico 23: 67-72.

Beardsley, 1. W. 1959. On the ofpineapple mealybugs in Hawaii with the

description ofa previously unnamed species (Homoptera: Pseudococcidae). Proc.

Hawaii. Entomol. Soc. 17: 29-37.

Beardsley, 1. W. 1965. Notes on the pineapple mealybug complex, with descriptions of

two new species (Homoptera: Pseudococcidae). Proc. Hawaii. Entomol. Soc. 19:

55-68.

Beardsley, 1. W. 1976. A synopsis ofthe Encyrtidae ofthe Hawaiian Islands with keys to

genera and species (Hymenoptera: Chalcidoidea). Proc. Hawaii. Entomol. Soc.

22: 181-228.

Beardsley, 1. W. 1993. The pineapple mealybug complex; taxonomy, distribution and host

relationships. First International Pineapple Symposium, Honolulu, Hawaii. Acta

Horticulturae 334: 383-386.

Beardsley, 1. W. Jr., T. H. Su., F. L. Mac Ewen, and D. Gerling. 1982. Field investigation

on the interrelationship ofthe big headed ant, the gray pineapple mealybug, and the

25 pineapple mealybug wilt disease in Hawaii. Proc. Hawaii. Entomol. Soc. 24: 51­

57.

Bess, H. A. 1958. The green scale, Coccus viridis (Green) (Homoptera: Coccidae), and

ants. Proc. Hawaii. Entomol. Soc. 16: 349-355.

Carter, W. 1932. Studies ofpopulations ofPselldococcus brevipes (Ckl.) occurring on

pineapple plants. Ecology 3: 296-304.

Carter, W. 1933. The pineapple mealybug, Pselidococclis brevipes (Ckl.) and wilt of

pineapple. Phytopathology 23: 207-242.

Carter, W. 1934. Mealybug wilt and green spot in Jamaica and Central America.

Phytopathology 24: 424-426.

Carter, W. 1935a. The symbionts ofPselldococcus brevipes (Ckl.). Ann. Entomol. Soc.

America 28: 60-64.

Carter, W. 1935b. Studies on the biological control ofPseudococcus brevipes (Ckl.) in

Jamaica and Central America. 1. Econ. Entomol. 28: 1037-1041.

Carter, W. 1937. Importation and laboratory breeding oftwo chalcid parasites of

Pseudococclis brevipes (Ckl.). 1. Econ. Entomol. 30: 370-372.

Carter, W. 1941. Insects and the spread ofplant diseases. Smithsonian Report for 1940,

pp. 329-342.

Carter, W. 1942. The geographical distribution ofmealybug wilt with notes ofsome

other insect pests ofpineapple. 1. Econ. Entomol. 35: 10-15.

Carter, W. 1944. Biological control ofPseudococclls brevipes (Ckl.). Proc. Hawaii.

Entomol. Soc. 12. 15.

26 Carter, W. 1945a. Encyrtid parasites ofPseudococcus brevipes (Cockerell). Proc.

Hawaii. Entomol. Soc. 12: 489.

Carter, W. 1945b. Some etiological aspects ofmealybug wilt. Phytopathology 35: 305.

Carter, W. 1948. The effect ofmealybug feeding on pineapple plants grown in finely

atomized nutrient solution. Phytopathology 38: 645-657.

Carter, W. 1949. Insect notes from South America with special reference to

Pselldococclls brevipes and mealybug wilt. 1. Econ. Entomol. 42: 761-766.

Carter, W. 1951. The feeding sequence ofPseudococcus brevipes (Ckl.) in relation to

mealybug wilt ofpineapple in Hawaii. Phytopathology 41: 769-780.

Carter, W. 1960. A study ofmealybug populations (Dysmicoccus brevipes (Ckl.)) in an

ant-free field. 1. Eeon. Entomol. 53: 296-299.

Carter, W. 1962. Insects in Relation to Plant Disease. 1. Willey & Sons, N. Y., pp. 238­

265.

Carter, W. 1963. Mealybug wilt ofpineapple; a reappraisal. Ann. New York Acad. Sci.

105: 741-764.

Carter, W. 1966. Experiment on the relation ofPselldococcus adonidum to mealybug wilt

and mealybug stripe ofpineapple. Pineapple Research Institute ofHawaii,

Research Report PRI121: 1-8.

Carter, W. 1967. Insects and Related Pests ofPineapple in Hawaii. A manual for

fieldman. Pineapple Research Institute (Restricted publication) pp. 36-53.

Carter, W. and 1. L. Collins. 1947. Resistance to mealybug wilt ofpineapple with special

reference to a Cayenne-Queen Hybrid. Phytopathology 37: 332-338.

27 Carter, W. and K. Ito. 1956. Study ofthe value ofpineapple plants as sources ofthe

mealybug wilt factor. Phytopathology 46: 601-605.

Chang, V. C. S., and A. K. Ota. 1990. Ant control in Hawaiian drip irrigation system, pp:

708-715. Ill: Applied Myrmecology. A world perspective. R. K. Vander Meer, K.

Jaffe and A. Cedeno (ed.). West View Press, Boulder, Colo.

Chapman, R. N. 1938. Biological control (Project 26). Experimental Station ofPineapple

Producers Cooperative Association. Honolulu, Hi. Monthly Report, December,

pp.319-333.

Clausen, C. P. 1956. Biological control ofinsect pest in the Continental United States.

USDA, Tech. Bull. No. 1139, 151 p.

Collins, J. L., and W. Carter. 1954. Wilt resistance mutations in the Cayenne variety of

pineapple. Phytophatology 44: 662-666.

Compere, H. 1936. A new genus and species ofEncyrtidae parasitic in the pineapple

mealybug, Pseudococclis brevipes (CkI.). Proc. Hawaii. Entomol. Soc. 9: 171­

174.

Cudjoe, A. R., P. Neuenschwander, and M. J. W. Copland. 1993. Interference by ants in

biological control ofthe cassava mealybug, Phenacoccus manihoti (:

Pseudococcidae) in Ghana. Bull. Entomol. Research 83: 15-22.

Ferris, G. F. 1948. Atlas ofScale Insects ofNorth America. Series V. Pseudococcidae.

Stanford University Press, Stanford, CA. 278 p.

Flanders, S. E. 1951. The role ofthe ant in the biological control ofhomopterous insects.

Can. Ent. 83: 93-98.

28 Flanders, S. E. 1958. The role ofthe ant in the biological control ofscale insects in

California. Proc. Inter. Congr. Entomol. 10th (1956), Montreal 4: 579-584.

Fluker, S. 1969. Sympatric association among selected ant species and some effects on

sugarcane mealybugs in Hawaii. Ph.D. Dissertation, University ofHawaii,

Honolulu, Hi.

Fluker, S. S., E. W. Huddleston, and 1. W. Beardsley, 1968. Some effects ofthe big­

headed ant on populations ofthe pink sugarcane mealybug. 1. Econ. Entomol. 61:

474-477.

Funasaki, G. Y., P. Y. Lai, L. M. Nakahara, 1. W. Beardsley, and A. K. Ota. 1988. A

review ofbiological control introductions in Hawaii: 1890-1985. Proc. Hawaii.

EntomoI. Soc. 28: 105-160.

Glancey, B. M., N. Reimer, and W. A. Banks. 1990. Effects ofIGR fenoxycarb and

sumitomo S-31183 and the queens oftwo Myrmicinae ant species, pp. 604-614.

Ill: Applied Myrmecology. A world perspective. R. K. Vander Meer, K. Jaffe,

and A. Cedeno (eds.), West View Press. Buolder, Colo.

Gray, R. A. 1952. Composition ofhoneydew excreted by pineapple mealybugs. Science

115: 129-133.

Guerrout, R. 1972. Repercussion du wilt sur la production de l' en Cote D'Ivore.

Fruits 27: 179-184.

Giffard, W. M. 1907. Notes and exhibition. Proc. Hawaiian Entomol. Soc. 1: 173-174.

29 Gillespie R. G., and N. Reimer. 1993. The effect ofalien predatory ants (Hymenoptera:

Formicidae) on Hawaiian endemic spiders (Aranea: Tetragnathidae). Pacific

Science 47: 21-33.

Gunasinghe, U. B., and T. L. German. 1989. Purification and partial characterization ofa

virus from pineapple. Phytopathology 79: 1337-1341.

Hanks, L. M., and C. S. Sadof 1990. The effect ofants on nymphal survivorship of

Coccus viridis (Homoptera; Coccidae). Biotropica 22(2): 210-213.

Heidari, M, and M. 1. W. Copland. 1993. Honeydew: a food resource or arrestant for the

mealybug predator Cryptolaemus montrouzieri? Entomophaga 38(1): 63-68.

Holldobler, B., and E. O. Wilson. 1990. The ants. The Bleknap Press ofHarvard

University Press, Cambridge, MA.

Hu,1. S., D. M. Sether, H. M. Harrington, and D. E. Ullman. 1995. Two-step heat

treatment ofpineapple crowns increases thermotolerance. HortTech 5: 63-66.

Illingworth, 1. F. 1931. Preliminary report on evidence that mealybugs are an important

factor in pineapple wilt. J. Econ. Entomol. 24: 877-889.

Ito, K. 1938. Studies on the life history ofthe pineapple mealybug Pseudococcus brevipes

(Ckl.). 1. Econ. Entomol. 31: 291-298.

Jahn, G. 1992. The ecological significance ofthe big-headed ant in mealybug wilt disease

ofpineapple. Ph.D. Dissertation, University ofHawaii, Honolulu, Hi.

Jahn, G., and 1. W. Beardsley. 1993. Big-headed ants, Pheidole megacephala:

interference with biological control ofgray pineapple mealybugs. In: Exotic Ants:

30 Biology, Impact, and Control ofIntroduced Species. D. F. Williams (ed.), pp:

199-205. Westview Studies in Insect Biology. Boulder, Colorado.

Jouvenaz, D. P. 1986. Diseases offire ants: Problems and opportunities. pp. 327-338.

111: Fire ants and LeafCutting Ants: A Synthesis ofthe Current Knowledge. C. S.

Lofgren and R. K. VanderMeer, eds. Westview Press, Boulder, CO.

Lamb, K. P. 1974. Economic Entomology in the Tropics. Academic Press. London. pp.

37-45.

Madon, E. A. 1978. Ecological, pathological and control aspects ofthe mealybug-ant­

wilt complex in pineapple. Ph.D. Dissertation, University ofHawaii, Honolulu, Hi.

Nakahara, S. 1981. List ofthe Hawaiian Coccoidea (Homoptera: Sternorhyncha). Proc.

Hawaiian Entomol. Soc. 23: 387-419.

Nechols,1. R. and T. F. Seibert. 1985. Biological control ofthe spherical mealybug,

Nipaecocclis vasta/or (Homoptera: Pseudococcidae): assessment by ant exclusion.

Environ. Entomol. 14: 45-47.

Nixon, G. E. 1. 1951. The association ofants with aphids and coccids. Commonwealth

Institute ofEntomology, London 36 pp.

Petty, G. L. 1978. The pineapple mealybug. Pineapple H.15. Farming in South Africa, 8

p.

Petty G. 1., and H. Tustin. 1993. Ant (Pheidole megacephala F.) - mealybug

(Dysmicoccus brevipes Ckll.) relationships in pineapples in South Africa. First

International Pineapple Symposium, Honolulu, Hawaii, Acta Horticulturae 334:

387-395.

31 Phillips, 1. S. 1934. The biology, distribution and control ofants in Hawaiian pineapple

fields. Ph.D. Dissertation, University ofHawaii, Honolulu, Hi.

Plank, H. K. and M. R. Smith. 1940. A survey ofthe pineapple mealybug in Puerto Rico

and preliminary studies ofits control. 1. Agric. Univ. ofPuerto Rico 24: 49-75.

Py, C., 1. 1. Lachoehuilhe, and C. Teisson. 1984. L'ananas sa Culture, ses Produits. G.P.

Maisonneure & Larose Agence de Coperation Cuiturelle et Technique. Paris, 562

p.

Rai, B. K. and K. Sinha. 1980. Pineapple: chemical control ofmealybugs and associated

ants in Guyana. 1. Econ. Entomol. 73: 41-45.

Reimer, N. 1. 1993. Distribution and impact ofalien ants in vulnerable Hawaiian

ecosystems, pp. 11-22. 111: Exotic Ants. Biology, impact and control ofintroduced

species. D. F. Williams (ed.). Westview Press, Boulder, Colo.

Reimer, N. J., and 1. W. Beardsley, Jr. 1989. Pineapple, Allanas comosus (L.) Merr.

Bigheaded ant: Pheidole megacephala (F.). Control test 1986. Insecticide and

Acaricide Tests 14: 69-70.

Reimer, N. J., and 1. W. Beardsley, Jr. 1989a. Pineapple, Allanas comoslis (L.) Merr.

Bigheaded ant: Pheidole megacephala (F.). Control test 1987. Insecticide and

Acaricide Tests 14: 69.

Reimer, N. J., and 1. W. Beardsley. 1990. Effectiveness ofhydramethylnon and

fenoxycarb for control ofbigheaded ant (Hymenoptera: Forrnicidae), an ant

associated with mealybug wilt ofpineapple in Hawaii. 1. Econ. Entomol. 83: 74­

80.

32 Reimer, N., 1. W. Beardsley, and G. Jahn. 1990. Pest ant in the Hawaiian Islands, pp. 40­

50. 111: Applied Myrmecology. A world perspective. R. K. Vander Meer, K. Jaffe

and A. Cedeno (ed.). West View Press, Boulder, Colo.

Reimer, N. 1., M. Cope and G. Yasuda. 1993. Interference ofPheidole megacephala

(Hymenoptera: Formicidae) with biological control ofCoccus viridis (Homoptera:

Coccidae) in coffee. Environ. Entomol. 22(2): 483-488.

Reimer, N. 1., B. M. Glancey, and 1. W. Beardsley. 1991. Development ofPheidole

megacephala (Hymenoptera: Formicidae) colonies following ingestion of

fenoxycarb and pyriproxyfen. 1. Econ. Entomol. 84: 55-60.

Reimer, N. 1., and H. Gonzalez-Hernandez. 1993. Evidence for colony founding by

haplometrosis in the big-headed ant, Pheidole megacephala. First International

Pineapple Symposium, Honolulu, Hawaii, Acta Horticulturae 334: 397-405.

Rohrbach, K. G., 1. W. Beardsley, T. L. German, N. 1. Reimer, and W. G. Sanford. 1988.

Mealybug wilt, mealybugs, and ants on pineapple. Plant Disease 72: 558-565.

Rohrbach, K. G. and W. L. Apt. 1986. Nematode and disease problems ofpineapple.

Plant Disease 70: 81-87.

Serrano, F. B. 1934. Pineapple mealy-bug wilt in the Philippines. Philipp. 1. Sci. 55:

363-377.

Singh, S. 1. and S. M. Sastry. 1974. Wilt ofpineapple, a new virus disease in India.

Indian Phytopathology 27: 298-303.

Singh, T. V. K., T. R. Goud, and K. M. Azam. 1988. Attack ofmealybug, Dysmicoccus

brevipes on groundnut. Indian 1. Entomol. 48: 358.

33 Steyn, J. J. 1954. The effect ofthe cosmopolitan brown house ant (Pheidole

megacephala F.) on citrus red scale (Aonidiella aumntii Mask.) at Letaba. 1.

Entomol. Soc. S. Africa 17: 252-264.

Strickland, A. H. 1951. The entomology ofswollen shoot cacao. 1.- The insect species

involved, with notes on their biology. Bull. Entomol. Research 41: 725-748.

Su, T. H. 1979. The biology and control ofthe big-headed ant and its associated

mealybugs in Hawaiian pineapple fields. Ph.D. Dissertation, University ofHawaii,

Honolulu, Hi.

Su, T. H., 1. W. Beardsley, and F. L. McEwen. 1980. AC-217,300, a promising new

insecticide for use in baits for control ofthe bigheaded ant in pineapple. 1. Econ.

Entomol. 73: 755-756.

Swezey, O. H. 1939. Recent records ofthe introduction ofbeneficial insects into the

Hawaiian islands. Proc Hawaii. Entomol. Soc. 10: 349-352.

Ullman, D. E., T. L. German, U. B. Gunasinghe, and R. H. Ebesu. 1989. Serology ofa

closterovirus particle associated with mealybug wilt ofpineapple. Phytopathology

79: 1341-1345.

Ullman, D. E., T. L. German, C. E. McIntosh, and D. D. F. Williams. 1991. Effect ofheat

treatment on a closterovirus-like particle associated with mealybug wilt of

pineapple. Plant Disease 75: 859-861.

Ullman, D. E., 1. S. Hu, D. Sether, A. Gonzalves, D. F. Williams and H. Fleisch. 1993.

Heat treatment ofpineapple: subsequent growth and occurrence ofmealybug wilt

34 ofpineapple. First International Pineapple Symposium, Honolulu, Hawaii, Acta

Horticulturae 334: 407-410.

Wakman, W., D. S. TeakIe, 1. E. Thomas, and R. G. Dietzgen. 1995. Presence ofa

c1ostero-like virus and a bacilliform virus in pineapple plants in Australia. Aust. 1.

Agric. Res. 46: 947-958.

Way, M. 1. 1963. Mutualism between ants and Honeydew-producing Homoptera. Ann.

Rev. Entomol. 8: 307-344.

Weber, P. W. 1955. Recent liberations ofbeneficial insects in Hawaii- IV. Proc. Hawaii.

Entomol. Soc. 15: 635-638.

Williams, D. D. F., and H. Fleisch. 1993. Historical review ofpineapple breeding in

Hawaii. First International Pineapple Symposium, Honolulu, Hawaii, Acta

Horticulturae 334: 67-76.

Williams, D. J. 1985. Australian Mealybugs. British Museum (Natural History), pp. 116­

118.

Williams, D. 1., and M. C. Granara de Willink. 1992. Mealybugs ofCentral and South

America. c.A.B. International Institute ofEntomology. London. 635 p.

Zimmerman, E. C. 1948. Insects ofHawaii. Vol. 5. Homoptera: Sternorhyncha.

University ofHawaii Press. Honolulu, Hi. pp. 189-201.

35 Ill. SURVEY OF THE NATURAL ENEMIES OF THE PINEAPPLE

MEALYBUGS

Introduction

The pink pineapple mealybug, Dysmicoccus brevipes (Cockerell), is a serious pest

everywhere pineapple is grown and is associated with the disease mealybug wilt of

pineapple (MWP) (Beardsley 1959, 1993, Carter 1962, Bartlett 1978). MWP is the most

destructive disease ofpineapple worldwide (Carter 1962). High mealybug populations in

Hawaii are associated with the presence ofants, primarily the big-headed ant, Pheidole

megacephala (F.) (Carter 1960, 1967). Control ofMWP has been obtained when ants are

controlled (Beardsley et al. 1982, Rohrbach et al. 1988). Why mealybug populations

suddenly decrease in the absence ofants is not well understood. Three hypotheses have

been proposed to explain this phenomenon. The most accepted is that ants protect

Homoptera from their associated natural enemies (Illingworth 1931, Carter 1933, Flanders

1951, Nixon 1951, Bess 1958, Barttlet 1961, Way 1963, Fluker 1969, Petty 1978,

Nechols and Seibert 1985, Jahn 1992, Cudjoe et al. 1993, Reimer et al. 1993). A second

hypothesis is that ants remove and feed upon accumulated honeydew (sanitation

hypothesis) which benefits survival ofHomoptera (Flanders 1951, Bess 1958). Lastly,

ants may aid in mealybug dispersal to new plants (Illingworth 1931, Carter 1932, Phillips

1934, Plank and Smith 1948, Petty 1978). Although natural enemies are believed to play

a role in mealybug suppression in the absence ofants, few studies in Hawaii have identified the major predators and parasitoids attacking mealybugs in pineapple plantings.

36 Attempts at biological control ofmealybugs in Hawaii were started in 1894 with the introduction ofseveral natural enemies ofthe mealybugs. They were introduced into

Hawaii from Australia, Japan, South China, California, Central and South America from

1894 to 1936 (Chapman 1938, Funasaki et al. 1988). From 1930 to 1936 several entomophagus species were introduced to control Dysmicoccus spp. (Chapman 1938,

Barttlet 1978) from Central and South America (Carter 1967). It was believed that the natural enemies that successfully established in Hawaii impacted pineapple mealybugs in the absence ofants (Carter 1945, Zimmerman 1948, Barttlet 1978). According to Carter

(1967), only a small proportion ofthe natural enemies released established due the parallel use ofinsecticides. Introductions were continued in 1946 without good results, and continued up to 1958 with the introduction and release ofthe parasitoid Pseudaphycus angelicus (Howard) (from the Commonwealth Institute ofBiological Control laboratory at Fontana, California) on Oahu, Lanai, and Maui in 1954 (Weber 1955), and the parasitoid Pselldaphyclls dysmicocci Bennet from Trinidad (Barttlet 1978). Following these multiple introductions, no studies were made to determine natural enemy establishment or to evaluate the effectiveness ofalready established natural enemies in controlling pineapple mealybugs. The objectives ofthis study were to identify the natural enemies attacking Dysmicoccus spp. in pineapple plantings as well as the Dysmicoccus and ant species present.

Materials and Methods

Dysmicoccus spp., their natural enemies and associated ants were surveyed in

37 abandoned pineapple fields on the Hawaiian islands ofOahu and Maui. Oahu surveys

were conducted in July 1992, January, June and November 1993 in two pineapple fields

(No.5, 310 m; No. 69,270 m) (latitude: 21°30' N and longitude: 158° 10' W) located at

Del Monte Fresh Produce (Hawaii) Inc., Kunia, Oahu (Fig. 3.1). Maui surveys were made in July 1992, February, June and October 1993 in three pineapple fields belonging to the

Maui Land and Pineapple Co. Two fields were located at the Honolua Plantation (No.9,

90 m; No. 43,240 m) (latitude: 20° 50' N and longitude: 156° 40' W) and one field at the

Haliimaile Station (No. 70, 390 m) (latitude: 20° 50' N and longitude: 153° 20' W) (Fig.

3.2). From each pineapple field, 20 pineapple plants were randomly selected on each survey date. Each plant was separated into its roots and aerial parts. Plant components were placed in individual plastic bags and taken to the laboratory for examination. Each sample was kept in closed 1 I polyethylene containers, and held under laboratory conditions (28 ± 20 C, 60-70% RH, and L:D of 14-10) for possible emergence ofnatural enemy adults. All natural enemies (predators, primary and secondary parasitoids), and ants found on plants were held for identification. When no D. neobrevipes males

(cocoons or adults) were found on a plant, five mealybug adults were slide mounted to confirm the presence ofonly D. brevipes. Slide mounts were prepared from each plant not infested with male cocoons, according to a modified method described by Williams and Granara de Willink (1992). Emerged natural enemies were mounted and identified.

Pineapple mealybugs were identified by John W. Beardsley and Neil 1. Reimer,

Department ofEntomology, University ofHawaii at Manoa. Encyrtidae and Coccinellidae

38 Field 68 ---;~~~~=.:

Field 5

O'ahu

Fig. 3.1. Oahu sites surveyed for natural enemies of pineapple mealybugs.

39 Field 43 Honolua Station Field 70 Haliimaile Station

Maui

\,===:f S"'"

Fig 3.2. Maui sites surveyed for natural enemies ofpineapple mealybugs.

40 were identified by J. W. Beardsley. Cecidomyiidae weres identified by Elmo D. Hardy,

Department ofEntomology, University ofHawaii at Manoa.

Results

A total offive natural enemies, previously introduced into Hawaii to control

pineapple mealybugs, were found associated with D. brevipes and D. neobrevipes. All

natural enemies were found attacking pineapple mealybugs tended by ants. Ofthese

natural enemies, two were primary parasitoids, Anagyms ananatis Gahan (Hymenoptera:

Encyrtidae), and Euryrhopa/lis propinqlllls Kerrich (Hymenoptera: Encyrtidae). The

remaining species were predators, Lobodip/osis pselldococci (Felt.) (Diptera:

Cecidmomyiidae), Nephlls billicernarilis Mulsant and Sticholotis nificeps Weise

(Coleoptera: CoccineIIidae). Additional entomophagous species found included the

hyperparasitoid Prochiloneums sp. (Hymenoptera: Encyrtidae) associated only with the

encyrtid parasitoid E. propinqllus. and the parasitoid Homa/Olylus sp. (Hymenoptera:

Encyrtidae) found attacking larvae ofboth coccineIIid predators, N. billicernarius and S.

nificeps. The latter predator was found only sporadically, once on Oahu and on two

sampling dates on Maui.

Parasitoids. A. ananatis was found at both surveys sites on Oahu, and on all

sampling dates except 31 January 1992. A. ananatis was reared only from D. brevipes

infesting pineapple leaves and fruits, although D. brevipes was also present on pineapple roots (Table 3.1). On Oahu, D. brevipes was associated with the big headed-ant and the black ant, Ochetel/us glaber (Mayr), the former being the dominant species.

41 Table 3.1. Locations and dates (1992-1993) that Anagyrlls ananatis was found associated with pineapple mealybugs and ants in pineapple on the islands ofOahu and Maui, Hawaii.

MeanNo.A. % Field Elevation Mean No. PM ananatis reared Parasiti- Associated Island Site No. (m) Date PM"·b per Plant ± SD ±SD zation AntsC Oahu Kunia 5 310 31 July 92 Pink 145.0 ± 111.5 1.8 ± 2.9 3.3 0. g., P. m. 30 June 93 Pink 62.3 ±46.6 2.0 ±2.8 2.4 0. g., P. m. 17 Nov 93 Pink 24.8 ± 13.8 0.1 ± 0.3 0.3 0. g., P. m.

68 270 30 June 93 Pink, Gray 75.6 ± 92.4 0.3 ± 0.4 0.4 P.m. 17 Nov 93 Pink, Gray 23.0 ± 26.1 0.1 ±0.2 0.3 P.m.

Maui Haliimaile 70 390 4 Feb 93 Pink 28.2 ± 41.6 0.6 ± 2.5 0.6 M·f,P.m ~ 8 June 93 Pink 2.8 ± 4.3 9.9 H. spp., P. m. N 22.9 ± 29.0 5 Oct 93 Pink 72.1 ± 52.3 2.7 ±4.3 2.3 P.m.

Honolua 9 90 21 July 92 Pink 100.0 ± 36.4 3.2 ± 6.05 2.5 P.m. 4 Feb 93 Pink 157.2 ± 175.6 1.2 ± 1.6 1.0 P.m. 9 June 93 Pink 137.1 ± 136.4 5.1 8.7 3.0 P.a.,P.m.

43 240 21 July 92 Pink, Gray 112.0 ± 70.5 3.8 ±4.0 2.5 P.m. 4 Feb 93 Pink, Gray 118.7 ± 162.7 0.1 ± 0.3 0.5 P.m. 9 June 93 Pink, Gray 79.9 ± 54.0 1.2 ± 1.8 2.7 P.m. 5 Oct 93 Pink, Gray 57.0 ± 64.2 0.5 ±0.9 1.0 P.m. " PM == Pineapple mealybugs; Pink == Dysmicocclis brevipes; Gray == Dysmicocclis ueobrevipes. bNote that the gray pineapple mealybug did not serve as a host ofA. auanatis. C Formicidae species: H. spp. == Hypoponera sp.; M f == MOllomorium floricola (Jerdon); 0. g. == Ochetelllls glaber (Mayr); P. a. == Plagiolepis alluaudi Forel; P. m. == Pheidole megacephala (F.). On Maui, A. ananatis was found at all survey sites, and on all sampling dates. It attacked

only D. brevipes infesting pineapple leaves, fruits and blossom cups (Table 3.1). Only D.

brevipes was found infesting pineapple roots, leaves, fruits, blossom cups and crowns at

Field 9, Honolua. Pineapple plants were infested with both mealybug species in Field 43,

Honolua. D. brevipes was found infesting pineapple roots, leaves, fruits, blossom cups

and crowns while D. neobrevipes only infested pineapple leaves, fruits and crowns. D.

brevipes was the only mealybug detected on pineapple roots, leaves, fruits, blossom cups

and crowns in Field 70, Haliimaile. On Maui, mealybugs were usually associated with P.

megacephala in all fields, and in some cases with Monomorillmjloricola Jerdon, and

Hypoponera sp. on field 70, and with Plagiolepis alluaudi Forel, the little yellow ant, in

the Field 9 (Table 3.1). Percent parasitization ofD. brevipes by A. ananatis was low in

the presence ofants, ranging from 0.3 to 9.9 percent.

The encyrtid ElIryrhopalus propinquus Kerrich was detected on both islands

(Table 3.2). In Kunia, Oahu, E. propinquus was found in the Field 68 where both mealybug species were present. In Field 5, only D. brevipes was detected on all the sampling dates, and E. propinqlllls was reared from it. At Kunia, the dominant ant species was P. megacephala, although 0. glaber was restricted to small patches (Table 3.2). On

Maui, E. propinqulls parasitized D. brevipes and both mealybugs in Fields 9 and 43, respectively, at Honolua.

P. megacephala was found in all Maui fields on all sampling dates, while M f1oricola, P. alluaudi, and Hypoponera sp. each were found once but on different dates

(Table 3.2).

43 Table 3.2. Locations and dates (1992-1993) that ElIryhropallis propillqlllls was found associated with pineapple mealybugs and ants in pineapple on the islands ofOahu and Maui, Hawaii.

MeanNo.E. % Field Elevation Mean No. PM propillqlllls Parasiti- Associated Hyper-parasitoid C Island Site No. (m) Date PMa per Plant ± SD reared ± SD zation Antsb Oahu Kunia 5 310 31 July 92 Pink 145.0 ± 111.5 0.05 ±0.2 0.3 D.g., P. m. Prochilollellrlls sp. 30 June 93 Pink 62.3 ± 46.6 0.05 ± 0.2 0.2 o.g., P. m. ProchilollclIrlls sp.

68 270 30 June 93 Pink, Gray 75.6 ± 92.4 0.10±0.4 0.2 P.m. ProchiJollellrlls sp. 17 Nov 93 Pink, Gray 23.0 ± 26.1 0.50 ± 0.2 0.1 P.m.

Maui Honolua 9 90 21 July 92 Pink 100.0 ± 36.4 1.20 ± 1.6 0.2 P.m. ProchiJollclIrlis sp.

~ ~ 43 240 21 July 92 Pink, Gray 112.0 ± 70.5 2.20 ±3.0 1.7 P.m. Prochilollellrlls sp. 9 June 93 Pink, Gray 79.9 ± 54.0 0.60 ± 1.3 0.7 P.m.

a PM = Pineapple mealybugs; Pink = Dysmicoccus brevipes; Gray Dysmicocclis lIeobrevipes. b Formicidae' species: O. g. = Ochetelllls glaber (Mayr); P.m. =Pheidole megacephala (F.).

C Hymenoptera: Encyrtid. Predators. The cecidomyiid predator, Lobodip/osis pselldococci (Felt.) was detected on both islands and in all fields sampled (Table 3.3). The highest number of individuals (5.75 1. pselldococci larvae and adults! plant) was found in Honolua, Field 43, where a mixed Dysmicoccus population infested pineapple plants. In Honolua, Field 9, it was found only once preying on D. brevipes feeding on fruit, and in Haliimaile Field 70 only one individual was collected from a pineapple fruit infested with D. brevipes.

The coccinellid Nephus bilucemarius Mulsant was found on Oahu and Maui

(Table 3.4). In Kunia, Oahu, it was found only in Field 5 attacking D. brevipes infesting pineapple fruit. In Honolua, Maui, Field 9, N. bilucemarius preyed on D. brevipes infesting pineapple fruit and blossom cups. In Field 43, Honolua, this predator fed upon both mealybugs present on pineapple fruit and crowns. In Haliimaile, Field 70, it was found attacking D. brevipes infesting pineapple fruit (Table 3.4). Mean densities ofN. bilucemarills ranged from 0.1 to 1.4 individuals per plant.

The coccinellid Stich%tis ntjiceps Weise was collected on both islands. This particular predator was found in the lowest numbers (range from 0.05 to 0.2 individuals per plant) and on the fewest pineapple plants. In Kunia, Oahu it was found at both survey sites attacking both mealybugs infesting pineapple fruit and crowns (Table 3.5). In

Honolua, Field 43, a similar behavior.was observed. In Haliimaile, Field 70, it was found attacking D. brevipes infesting roots as well as fruit. This was the only natural enemy found attacking D. brevipes infesting pineapple roots (Table 3.5).

45 Table 3.3. Locations and dates (1992-1993) that Lobodiplosispselldococci was found associated with pineapple mealybugs and ants in pineapple on the islands ofOahu and Maui, Hawaii.

Mean No. L. Field Elevation Mean No. PM pselldococci per Associated Island Site No. (m) Date PMa per Plant ± SD plant ± SD Antsb Oahu Kunia 5 307.8 31 July 92 Pink 145.0 ± 111.5 5.75 ± 9.6 O.g., P. m. 30 June 93 Pink 62.3 ± 46.6 0.10 ± 0.4 O.g.,P.m. 17 Nov 93 Pink 24.8 ± 13.8 0.10 ± 0.4 o.g., P. m.

68 270 30 June 93 Pink, Gray 75.6 ± 92.4 1.75 ± 5.2 P.m. 17 Nov 93 Pink, Gray 23.0 ± 26.1 0.35 ± 1.0 P.m.

Maui Haliimaile 70 390 4 Feb 93 Pink M.j,P.m. ~ 28.2 ± 41.6 0.05 ± 0.2 0\ Honolua 9 90 21 July 92 Pink 100.0 ± 36.4 2.00 ± 2.8 P.m.

43 240 21 July 92 Pink, Gray 112.0 ± 70.5 3.00 ± 3.0 P.m. 4 Feb 93 Pink, Gray 118.7 ± 162.7 0.15±0.5 P.m. 9 June 93 Pink, Gray 79.9 ± 54.0 0.30 ± 1.3 P.m. 5 Oct 93 Pink, Gray 57.0 ± 64.2 4.00 ± 10.3 P.m.

a PM =Pineapple mealybugs; Pink =Dysmicocclls brevipes; Gray Dysmicoccils Ileobrevipes. b Formicidae species: M. f =MOllomorium floricola (Jerdon); O. g. =Ochetelllls glaber (Mayr); P.m. =Pheidole megacephala (F.). Table 3.4. Locations and dates (1992-1993) that Nephlls bilucernarills was found associated with pineapple mealybugs and ants in pineapple on the islands ofOahu and Maui, Hawaii.

Mean No. Mean No. N. % Parasitization by Field Elevation PM per billlcernarills Associated Homoityills sp.c Island Site No. (m) Date PMa Plant ± SD found per plant Antsb Oahu Kunia 5 310m 31 July 92 Pink 145.0 ± 111.5 1.8±2.4 0. g., P. m. 14.0 30 June 93 Pink 62.3 ±46.6 0.1 ± 0.3 0. g., P. m 0.0

Maui Haliimaile 70 390m 5 Oct 93 Pink 72.1 ± 52.3 0.1 ± 0.3 P.m. 0.0

Honolua 9 90m 21 July 92 Pink 100.0 ± 36.4 0.2 ± 0.4 P.m. 0.0 4 Feb 93 Pink 157.2 ± 175.6 0.4 ± 0.5 P.m. 9 June 93 Pink 137.1 ± 136.4 1.4±1.2 P.a.,P.m. 3.6 .j:>. -.l 43 240m 21 July 92 Pink, Gray 112.0 ± 70.5 0.2 ± 0.4 P.m. 50.0 9 June 93 Pink, Gray 79.9 ± 54.0 0.2 ± 0.7 P.m. 0.0

a PM = Pineapple mealybugs; Pink =Dysmicocclis brevipes; Gray DysmicocclIs neobrevipes. b Formicidae species: O. g. = Ochetelllls glaber (Mayr); P. a. =Plagiolepis allualldi Forel; P.m. =Pheidole megacephala (F.).

C Hymenoptera: Encyrtidae. Table 3.5. Locations and dates (1992-1993) that Sticolothis nificeps was found associated with pineapple mealybugs and ants in pineapple on the islands ofOahu and Maui, Hawaii.

% Parasitization Mean No. S. by Hom%tY/lis Field Elevation Mean No. PM rllficeps per Associated sp. c Island Site No. (m) Date PM3 per Plant ± SD plant ± SD Antsb Oahu Kunia 5 310 m 31 July 92 Pink 145.0 ± 111.5 0.10 ± 0.4 O.g.,P.m 50.0 30 June 93 Pink 62.3 ±46.6 0.05 ± 0.2 O.g.,P.m 0.0 17 Nov 92 Pink 24.8 ± 13.8 0.05 ± 0.2 O.g.,P.m 0.0

68 270m 30 June 93 Pink, Gray 75.6 ± 92.4 0.20 ± 0.7 P.m. 0.0

Maui Haliimaile 70 390m 8 June 93 Pink 22.9 ± 29.0 0.10±0.4 H. spp., P. m. 0.0 ~ 00 5 Oct 93 Pink 72.1 ± 52.3 0.05 ± 0.2 P.m. 0.0

43 240m 21 July 92 Pink, Gray 112.0 ± 70.5 0.1O±0.4 P.m. 50.0

3 PM = Pineapple mealybugs; Pink = Dysmicocclis brevipes; Gray Dysmicocclis neobrevipes. b Formicidae species: H. spp. =Hypoponera sp.; 0. g. =Ochetelllls g/aber (Mayr); P.I11. = Pheido/e megacepha/a (F.).

C Hymenoptera: Encyrtidae. Discussion

Survey results suggest that approximately 20 percent ofthe almost 20 natural

enemies imported and released against DysmicocclIsspp. from 1924 to 1958 established and persist today in Hawaii's pineapple fields (Chapman 1938, Weber 1855, Bartlett

1978). According to Chapman (1938), all natural enemy introductions prior to 1930 for mealybug control in general were unsuccessful relative to control ofpineapple mealybugs.

Those introductions were the product ofexplorations conducted from Mexico to Central

America and some from California and Australia. Successful introductions of entomophagous species associated with pineapple mealybugs were achieved in the 1930's when exploration was extended from Mexico to northern South America and Brazil

(Chapman 1936, Compere 1936, Carter 1937, Swezey 1939, Carter 1944, Bartlett 1978,

Funasaki et al. 1988). Ofthe few natural enemies that established in the 1930's, in particular the promising encyrtid Hambeltonia pselldococcina Compere imported from

Venezuela for D. brevipes control, none were found during these surveys. Carter (1945) last reported this parasitoid in Hawaii on D. brevipes in one Maui locality.

After successful establishment ofnatural enemies in Hawaii, nothing was done to conserve them in pineapple plantings because ofthe parallel development ofeffective pesticides at the time (Carter 1967). Results ofthis study suggest that additional attempts should be made to discover and introduce natural enemies ofD. brevipes and D. neobrevipes from

South America.

Most natural enemies found in this survey, except the coccinellid S. ruficeps, were associated with mealybugs feeding on aerial plant parts ofpineapple. This suggests that

49 these natural enemies were originally collected in other countries attacking mealybugs

present on the aerial plants part. No previous reports exist on natural enemies attacking

Dysmicoccus mealybugs feeding on the pineapple roots. Densities ofnatural enemies

found were low compared to the Dysmicoccus densities. This probably was a result ofthe

interference in natural enemy activity caused by ants such as P. megacephala.

Ofthe natural enemies found in this study, A. ananatis was the most common. It

was found in all pineapple fields on both islands, on almost all the sampling dates, and in

the highest numbers. However it attacked only D. brevipes. Survey data indicate that D.

brevipes was the dominant mealybug in the pineapple fields surveyed, and therefore it may

be one reason why A. ananatis was found in all fields. This finding is contrary to that of

Beardsley (1993) who stated that D. lleobrevipes was the most important mealybug

infesting fruits in pineapple.

A. ananatis is a primary endoparasitoid, purposely introduced to Hawaii in 1935

and 1936 from Brazil by D. T. Fullaway (Carter 1937, Chapman 1938). Carter (1949)

reported that Anagyros spp. was the most common parasitoid attacking D. brevipes in a pineapple growing area ofSao Paulo, Brazil. A. allanatis has been reared only from D. brevipes in Hawaii (Zimmerman 1948, Beardsley 1976), while in South America it has been reported to parasitize D. brevipes as well as the cosmopolitan mealybugs:

Dysmicoccus boninsis (Kuwana), Ferrisia virgata (Cockerell), Pianococclis citri Risso, and Antonina graminis (Maskell) (Noyse and Hayat 1994). In this survey, parasitization was observed on D. brevipes tended by ants. Therefore, good potential exists for use of

A. ananatis in augmentative programs against D. brevipes in Hawaii.

50 In contrast, E. propillquus was present in almost all localities, except Haliimaile,

Maui, but in low numbers compared to A. allallatis. Nevertheless, this parasitoid may be a

good candidate for use in biological control programs because it attacks both D. brevipes

and D. Ileobrevipes (Kerrich 1967, Beardsley 1976). It was purposely introduced to

Hawaii in 1936 from British Guyana and established by 1948 (Beardsley 1976). Although,

capable ofparasitizing both mealybug species, it appeared to prefer D. neobrevipes as a

host when mixed mealybug populations were present in pineapple plantings. Ofadditional

concern with this wasp is its susceptibility to the encyrtid hyperparasitoid ProchiJoneurus

sp.

The only predators which appeared to have any potential in Dysmicoccus

biological control are the coccinellid N. bilucernarius and the cecidomyiid 1. pseudococci.

Both predators attack D. brevipes and D. neobrevipes, and both were present in almost all

surveyed areas. N. biJucemarius was introduced into Hawaii from Mexico in 1930 and was reported established in 1939 (Swezey 1939, Leeper 1976). N. bilucemarius is a more prey specific predator. In Hawaii it has been previously reported to attack D. brevipes and N. vastator (Maskell) (Funasaki et. al. 1988). Therefore, it is a good candidate for natural enemy augmentative programs for Dysmicoccus control. Another attribute ofthis predator is that its immature and adult stages feed on all the mealybug stages. 1. pseudococci was introduced from Mexico in 1930, and reported established in 1934

(Chapman 1938, Swezey 1989, Hardy 1960). According to Chapman (1938) and Carter

(1944) this cecidomyiid, among other introduced natural enemies, was one ofthe best in pineapple fields where it gave sufficient control ofmealybugs feeding on fruit. Attacking

51 mealybugs on fruit makes this predator a good candidate for augmentative programs. In

Hawaii L. pselldococci has been found preying upon D. brevipes; the lebbeck mealybug,

Nipaecoccus vaslalor (Maskell); the citrus mealybug, Planococclls citri (Risso); and the pink sugarcane mealybug, Saccharicoccus sacchari (Cockerell) (Hardy 1960).

Considering the current distribution ofthe natural enemies found in this study and their feeding behaviors, A. ananalis, E. propillqlllls, L. pseudococci, andN. billicernarius, should be studied to determine their potential in augmentative biological programs against

Dysmicocclls spp. in pineapple.

52 References Cited

Bartlett, B. R. 1961. The influence ofants upon parasites, predators, and scale insects.

Ann. Entomol. Soc. Amer. 54: 543-551.

Bartlett, B. R. 1978. Pseudococcidae, pp. 137-170. In C. P. Clausen (ed.), Introduced

parasites and predators ofarthropod pest and weeds: a world review. U. S. D. A.

Handbook. No. 480.

Beardsley, 1. W. 1959. On the taxonomy ofpineapple mealybug in Hawaii with the

description ofa previously unnamed species (Homoptera: Pseudococcidae). Proc.

Hawaii. Entomol. Soc. 17: 29-37.

Beardsley, 1. W. 1976. A synopsis ofthe Encyrtidae ofthe Hawaiian Islands with keys to

genera and species (Hymenoptera: Chalcidoidea). Proc. Hawaiian Entomol. Soc.

22: 181-228.

Beardsley, 1. W. 1993. The pineapple mealybug complex; taxonomy, distribution and host

relationships. First International Pineapple Symposium, Honolulu, Hawaii. Acta

Horticulturae 334: 383-386.

Beardsley, J. W., Jr., T. H. Su., F. 1. Mac Ewen, and D. Gerling. 1982. Field

investigation on the interrelationship ofthe big headed ant, the gray pineapple

mealybug, and the pineapple mealybug wilt disease in Hawaii. Prec. Hawaii.

Entomol. Soc. 24: 51-57.

Bess, H. A. 1958. The green scale, Coccus viridis (Green) (Homoptera: Coccidae), and

ants. Prec. Hawaii. Entomol. Soc. 16: 349-355.

Carter, W. 1932. Studies ofpopulations ofPseudococclis brevipes (Ckl.) occurring on

53 pineapple plants. Ecol. 3: 296-304.

Carter, W. 1933. The pineapple mealybug, Pselidococclis brevipes (CkI.) and wilt of

pineapple. Phytopath. 23: 207-242.

Carter, W. 1937. Importation and laboratory breeding oftwo chalcid parasites of

Pseudococcus brevipes (Ckl.). 1. Econ. Entomol. 30: 370-372.

Carter, W. 1944. Biological control ofPseudococcus brevipes (Ckl.). Proc. Hawaii.

Entomol. Soc. 12: 15.

Carter, W. 1945. Encyrtid parasites ofPseudococcus brevipes (Cockerell). Proc. Hawaii.

Entomol. Soc. 12: 489.

Carter, W. 1960. A study ofmealybug populations (Dysmicoccus brevipes (CkI.» in an

ant-free field. 1. Econ. Entomol. 53: 296-299.

Carter, W. 1962. Insects in Relation to Plant Disease. 1. Willey & Sons, N. Y., pp. 238­

265.

Carter, W. 1967. Insects and Related Pests ofPineapple in Hawaii. A manual for

fieldmen. Pineapple Research Institute (Restricted publication) pp. 36-53.

Chapman, R. N. 1938. Biological control (Project 26). Experimental Station ofPineapple

Producers Cooperative Association. Honolulu, Hi. Monthly Report, December,

pp. 319-333.

Cudjoe, A. R., P. Neuenschwander, and M. 1. W. Copland. 1993. Interference by ants in

biological control ofthe cassava mealybug Phenacoccus manihoti (Hemiptera:

Pseudococcidae) in Ghana. Bull. Entomol. Res. 83: 15-22.

Flanders, S. E. 1951. The role ofthe ant in the biological control ofhornopterous insects.

54 Can. Ent. 83: 93-98.

Fluker, S. 1969. Sympatric association among selected ant species and some effects on

sugarcane mealybugs in Hawaii. Ph.D. Dissertation, University ofHawaii,

Honolulu, Hi.

Funasaki, G. Y., P. Y. Lai, L. M. Nakahara, 1. W. Beardsley, and A. K. Ota. 1988. A

review ofbiological control introductions in Hawaii: 1890-1985. Proc. Hawaii.

Entomol. Soc. 28: 105-160.

Hardy, D. E. 1960. Insect ofHawaii. X. Diptera: Nematocera-Brachycera. University of

Hawaii Press.

Illingworth, 1. F. 1931. Preliminary report on evidence that mealybug are an important

factor in pineapple wilt. 1. Econ. Entomol. 24: 877-889.

Jahn, G. 1992. The ecological significance ofthe big-headed ant in mealybug wilt disease

ofpineapple. Ph.D. Dissertation, University ofHawaii, Honolulu, Hi.

Kerrlch. G. 1. 1967. On the classification ofthe Anagyrine Encyrtidae, with a revision of

some ofthe genera (Hymenoptera: Chalcidoidea). Bull. ofthe British Museum

(Natural History) (Entomology) 20: 143-250.

Leeper, J. R. 1976. A review ofthe Hawaiian Coeeinellidae. Proe. Hawaii. Entomol.

Soc. 22: 279-306.

Neehols,1. R. and T. F. Seibert. 1985. Biological control ofthe spherical mealybug,

Nipaecocclis vastator (Homoptera: Pseudoeoccidae): assessment by ant exclusion.

Environ. Entomol. 14: 45-47.

Nixon, G. E. 1. 1951. The association ofants with aphids and coceids. Commonwealth

55 Instit. Entomol., London 36 pp.

Noyes J. S., and M. Hayat. 1994. Oriental Mealybug Parasitoids ofthe Anagyrini

(Hymenoptera: Encyrtidae). Natural History Museum. CAB International,

London. 554 pp.

Petty, G. L. 1978. The pineapple mealybug. Pineapple H.15. Farming in South Africa. 8

p.

Phillips, J. S. 1934. The biology, distribution and control ofants in Hawaiian pineapple

fields. Ph.D. Dissertation, University ofHawaii, Honolulu, Hi.

Plank, H. K. and M. R. Smith. 1940. A survey ofthe pineapple mealybug in Puerto Rico

and preliminary studies ofits control. J. Agric. Univ. ofPuerto Rico 24: 49-75.

Reimer, N. J., and 1. W. Beardsley, Jr. 1989. Pineapple, Ananas comosus (L.) Merr.

Bigheaded ant: Pheidole megacephala (F.). Control test 1986. Insecticide and

Acaricide Tests 14: 69-70.

Reimer, N. 1., M. Cope and G. Yasuda. 1993. Interference ofPheidole megacephala

(Hymenoptera: Formicidae) with biological control ofCoccus viridis (Homoptera:

Coccidae) in coffee. Environ. Entomol. 22(2): 483-488.

Rohrbach, K. G., J. W. Beardsley, T. L. German, N. J. Reimer, and W. G. Sanford. 1988.

Mealybug wilt, mealybugs, and ants on pineapple. Plant Dis. 72: 558-565.

Swezey, O. H. 1939. Recent records ofthe introduction ofbeneficial insects into the

Hawaiian islands. Proc Hawaii. Entomol. Soc. 10: 349-352.

Way, M. J. 1963. Mutualism between ants and Honeydew-producing Homoptera. Ann.

Rev. Entomol. 8: 307-344.

56 Weber, P. W. 1955. Recent liberation ofbeneficial insects in Hawaii-IV. Proc. Hawaii.

Entomol. Soc. 15: 635-638.

Williams, D. 1., and M. C. Granara de Willink. 1992. Mealybugs ofCentral and South

America. c.A.B. International Institute ofEntomology. London. 635 p.

Zimmerman, E. C. 1948. Insects ofHawaii. Vol. 5. Homoptera: Sternorhyncha.

University ofHawaii Press. Honolulu, Hi. pp. 189-201.

57 IV. IMPACT OF PHEIDOLE MEGACEPHALA (F.) (HYMENOPTERA:

FORMICIDAE) ON THE BIOLOGICAL CONTROL OF DYSMICOCCUS

BREVIPES (COCKERELL) (HOMOPTERA: PSEUDOCOCCIDAE)

Introduction

Ant species, especially members ofthe advanced formicid subfamilies Myrmicinae,

Dolichoderinae and Formicinae, play an important role in the survival and development of many honeydew-producing homopterans (Holldobler and Wilson 1990). The ants depend on the honeydew excreted by some aphids, mealybugs, and scale insects (Nixon 1951).

Honeydew is an important food source for ants because it is rich in sugars, amino acids, and waxes (Gray 1952, Heidari and Copland 1993). In exchange for this nutrient source,

Homoptera receive several benefits. Protection from natural enemies is considered the main benefit that Homoptera obtain from this association (Buckley 1987). The guardian role ofants has been proposed or demonstrated in some Coccidae and Pseudococcidae species (Illingworth 1931, Carter 1933, Flanders 1951, Nixon 1951, Stein 1954, Bess

1958, Flanders 1958, Bartlett 1961, Way 1963, Fluker 1969, Petty 1978, Nechols and

Seibert 1985, Rohrbach et aI. 1988, Hanks and Sadof 1990, Holldobler and Wilson 1990,

Jahn 1992, Gross 1993, Reimer et al. 1993, Jahn and Beardsley 1994). Sanitation is another benefit Homoptera derive from tending ants. Accumulation ofhoneydew promotes sooty mold growth and crawlers may become stuck in excess honeydew

(Flanders 1951, Nixon 1951, Bess 1958, Bartlett 1961, Carter 1967, Petty 1978,

Rohrbach et. al. 1988). Ants remove and feed upon honeydew, thereby increasing

58 Homoptera survival (Flanders 1951, Nixon 1951, Bess 1958). Movement and redistribution ofHomoptera by ants is another benefit, and has been observed or proposed for some anti Homoptera associations (Illingworth 1931, Carter 1932, Phillips 1934,

Plank and Smith 1940, Nixon 1951, Way 1963, Petty 1978).

In Hawaiian pineapple fields, Reimer et al. (1990) considered the big-headed ant,

Pheido/e megacepha/a (Fabricius), the fire ant, So/enopsis geminata (Fabricius), and the

Argentine ant, Linepithema humi/e (Mayr) (= lridomyrmex humi/is), as the most pestiferous ant species, because oftheir association with the pink pineapple mealybug,

DysmicocclIs brevipes (Cockerell) and the gray pineapple mealybug, D. neobrevipes

Beardsley. P. megacepha/a is the dominant ant species in Hawaii's pineapple fields, the majority ofwhich are below 600 m elevation (Reimer et al. 1990).

In a survey conducted in pineapple plantings on the Hawaiian islands ofOahu and

Maui, five natural enemies were found associated with pineapple mealybugs. These were the encyrtids Anagynls ananatis Gahan, and Euryrhopa/us propinqllus Kerrich; the cecidomyiid predator, Lobodip/osis pseudococci (Felt.); and the coccinellids Nephus bi/ucemarius Mulsant and Stich%tis nificeps Weise (See Chapter III). Ofthese, A. ananatis was the most common found. However, all ofthese species were found in low densities « 10 individualsl plant) and always in plantings infested with ants that may have reduced their efficacy as natural enemies. No evaluation ofnatural enemy effectiveness on pineapple mealybugs has been conducted using standard evaluation methods (De Bach and

Huffaker 1971).

59 The objectives ofthis study were to evaluate the biological control offield

populations ojD. brevipes and the impact ofP. megacephala on this mealybug and its

natural enemies.

Materials and Methods

Biological Check Method. Field studies were conducted in an abandoned

pineapple planting at Kunia, Oahu, Hawaii, to determine the efficacy ofnatural enemies of the pink pineapple mealybug (PPM) in the absence and presence ofants. The biological check method (De Bach and Huffaker 1971) was used in which ants were allowed to interfere with established biological controls ofPPM. The field site setup consisted ofa complete randomized block design. The two treatments were ants present and ants absent. Treatments were replicated three times (Fig. 4.1). The study was initiated on 24

January 1993 and completed on 25 August 1993. To eliminate ants from the ant-free plots, a bait formulation ofhydrametyinon (Amdro ®, American Cynamid Co., Wayne,

NJ) at the rate of2 kg baitl ha (0.77% hydrametylnon by weight) was applied on 24

January, 13 May and 21 June 1993. Field monitoring was conducted on 24 January 1993 before bait application and 13, 56, 89, 126, 153, and 188 days after initial ant bait application. For each treatment, five plants per plot were randomly selected. Sampled plants were completely removed, and aerial parts were placed in individual plastic bags for later laboratory dissection. In the laboratory, samples were kept in 5 liter polyethylene containers and covered with a lid containing an organdy covered hole of 13.3 cm diam.

Numbers ofthe pink pineapple mealybugs, parasitized mealybugs, and immature and adult

60 61 m I" ·1

B 1 Present Absent I ~ I I I

B 2 I Absent 10 I Present I

B 3 I Absent Present 10 I I ~N

Fig. 4.1. Schematic ofbiological check experiment showing layout ofant-free and ant infested plots in a complete randomized block design (B: block). Located at the, Del Monte Pineapple Co., Kunia, Oahu, HI.

61 predators were recorded only from the aerial parts ofthe plant. Data were analyzed by

ANOVA (Wilkinson 1990).

Combined Biological Check and Exclusion Method. Field studies were conducted in an abandoned pineapple planting at Kunia, Oahu, Hawaii, to examine the interactions ofP. megacephala with D. brevipes and the mealybug's natural enemies. A combination ofevaluation techniques was used in this study. These were the exclusion technique and interference technique (De Bach and Huffaker 1971). Open and closed cages over plants (paired cage technique) were used to exclude natural enemies and ants were used to interfere with the mealybugs natural enemies (biological check method).

Experimental design consisted ofa split plot with ants presence/absence as the variable of the main plot. Natural enemy presence/absence was subplotted within these main plots by means ofcaged plants (Fig. 4.2), Therefore, the experiment contained a total offour treatments replicated three times. Pineapple plants (Smooth Cayenne) used in this study were grown from crowns planted in Finson's Sunshine Mix soil in polyethylene black mulch bags of(20.3 x 14 x 35.6 cm) and held in screen cages of(1.1 x 1.1 x 2.1 m) located on the rooftop ofGilmore Hall, University ofHawaii at Manoa. One month later, plants were moved indoors where they were individually infested with 10 gravid D. brevipes females obtained from a one year old laboratory colony reared on butternut squash. Two weeks later, the infested potted plants were placed inside screened experimental cages randomly distributed throughout the pineapple field serving as the experimental site. Plants were hand watered once a week. In the ant-free blocks, Amdro

(0.77% hydrametylnon by weight) was applied at the rate of2 kg bait/ha on 14 March

62 61 m I" ~I c • An1s c o c An$ I • 0 fiE] • c 0 Absent Pmsent • • I I: •• I c a c An1s • c c· An1s c •c • Absent • I· a Pmsent I • I D • I a • .. An1s c An1s • c Absent • I· Pmsent 10" • "0I D [] [] "I --N [] - plants in open cages, free access to natural enemies • plants in closed cages, exclusion of natural enemies

Fig. 4.2. Schematic arrangement ofexperimental pineapple plots in a split plot design. Exclusion experiment. Del Monte Pineapple Co., Kunia, Oahu, m.

63 1994 and 18 April 1994 to eliminate ants. To exclude natural enemies from plants, 0.61 x

2 0.61 x 0.61 m screened cages covered with fine mesh of 1,369 cells per cm were used

(closed cages). A door was placed on the top ofeach cage to allow for inspection of

caged plants (Fig. 4.3.a). The bottom edges ofthese cages were buried in the ground.

Screen cages (0.61 x 0.61 x 0.61 m) with three holes of 15 cm diameter on each oftwo

opposite vertical sides were used (open cages) to allow access ofnatural enemies (Fig.

4.3. b). D. brevipes numbers and presence or absence ofants in caged plants were

monitored weekly through the sample period, from 25 February 1993 to August 1994.

Caged plants were removed at the end ofthe sampling period, dissected in the laboratory,

and examined for healthy and parasitized D. brevipes, parasitoids, and predators on the

aerial portions and root systems. Data on numbers ofmealybugs infesting plants was

analyzed by ANaVA for a split plot design and orthogonal contrast (Wilkinson 1990).

Impact ofP. megacepha!a on Mealybug Natural Enemies. Nephus

bi!ucernarius. Laboratory studies were conducted to determine what life stages ofthe

coccinellid N. bilucemarius were affected (, immature, adult) by the presence ofP. megacephala. For this experiment, two arena cages were used: one connected to a laboratory ant colony (treated cage) and the other without a connection (control cage)

(Fig. 4.4). Arena cages consisted ofa transparent plastic shoe box (15 cm wide x 9 cm high x 29 cm long). To provide ants access to the arena, a plastic tube (5 mm diam.) was connected to another similar shoe box containing a P. megacephala colony. The ant colony was obtained from a preexisting laboratory colony, from which several

64 ...... ~~~~§~ f Z~ v . ::: ...... ·.

a b

Fig. 4.3. Screen cages used in a field experiment for the exclusion ofnatural enemies ofD. brevipes in the presence or absence ofants. PPP: potted pineapple plant; a: closed cage type with a door; b: open cage type.

65 A

B

=- Ants o Mealybugs

c

...... 0

D ?. ,{~Q :i 9.

Fig. 4.4. Arena cages used for observing the impact ofants on the activity ofthe natural enemies ofD. brevipes. A: ant colony nest box; B: arena cage for the introduction of natural enemies, ants and mealybugs; C: arena cage control for the introduction ofnatural enemies and mealybugs; D: lateral view ofan arena cage; P: pineapple leaf.

66 mated queens, 4 cm ofbrood, workers and soldiers were transferred. The ant colony was

maintained in the nest box containing food (modified Banks diet) (Banks et al. 1981),

water, and a nest chamber (a 15 x 199 mm petri dish with a 5 mm layer ofPlaster ofParis

in the bottom) (modified Bishops cells) (Bishop et al. 1980). The top edges ofeach shoe

box were coated with Fluon AD-l® (Norton Products, Inc., Woodsocket, Rl) to prevent

escape ofthe ants. All laboratory experiments were conducted for 24 h at 24 ± l oC and

photophase of 14 hr.

To observe the impact ofants on N. bi/ucernarius adults, two fresWy cut pineapple

leaves (approximately 3 cm inside at the base, 15-20 cm in length) were infested each with

100 PPM first instars. One infested leafwas placed into each arena cage. One hour later, an adult coccinellid was introduced into each experimental arena cage for one day. At the end ofthe experiment, the condition (live or dead) ofthe coccinellid and final number of mealybugs were recorded for each arena. To observe the impact ofants on N. bilucemarius larvae, a similar procedure was followed with the exception that pineapple leaves were infested with 10 D. brevipes adults. Condition (live or dead) ofcoccinellid third instar larvae and number ofconsumed mealybugs were recorded at the end ofthe experiment. Impact ofants on N. bilucernarius was also tested. Similar arena cages were used for this experiment. Ten coccinellid eggs were placed in a 4.0 cm diam x 0.6 cm petri dish cover and exposed for 24 hr to ants in the arena cage. A day later, the number ofeggs remaining in the petri dish cover was recorded. All experiments for all coccinellid stages were replicated 10 times. An ANOVA analysis was conducted to

67 compare differences in prey consumed by adults and larvae ofthe coccinellid in the

presence or absence ofants (Wilkinson 1990).

Anagyrus ananatis. Laboratory studies were conducted to determine the impact ofP. megacephala on searching A. ananatis adult females. The laboratory setup consisted ofthe same types ofarena cages and ant nest colony used for observing the impact ofP. megacephala on N. billicernarilis. The only differences were the stages and numbers ofmealybugs used. In this case, pineapple leaves were individually infested with

80 mealybugs (third instar or adult females). Once the pineapple leafwas infested with mealybugs, P. megacephala was given free access to the arena cage. A day later, a pair of

A. ananatis adults was introduced to each cage for 24 hr. Afterwards, mealybugs were removed, placed in separate plastic vials covered with fine mesh, and held until mummies were formed. This experiment was replicated eight times and conducted for 24 h at 24 ±

1°C and photophase of 14 hr. An ANOVA test was conducted to compare differences in numbers ofA. ananatis parasitized hosts, in the presence and absence ofants (Wilkinson

1990).

Results

Biological Check Method. There were no significant differences among D. hrevipes densities in experimental pineapple plots before ant elimination (F= 1.64; df= 1;

P = 0.21) (Table 4.1). Highly significant differences in D. hrevipes densities between treatments were observed 56 days after initial ant bait application (F= 16.76; df= 1; P <

68 Table 4.1. Summary ofthe analyses ofvariance ofthe mean densities ofD. brevipes, A. ananatis and predators (N. billicernarilis and L. pselldococci) in ant-infested and ant-free pineapple plots, before ant elimination, and 13, 56, 89, 126, 153, and 188 days after initial ant bait application.

D. brevipes A. anana/is Predators

Time (days) F P F P F P

0 1.64 0.210 0 0 0.00 0.00

13 0.19 0.660 0 0.000 1.00 0.33

56 16.76 < 0.001 ** 9.99 0.004 ** 4.24 0.05 *

0\ \0 89 27.36 < 0.001 ** 6.74 0.015 * 0.79 0.38

126 12.38 0.002 ** 2.24 0.140 0.00 0.00

153 29.98 < 0.001 ** 0 1.000 1.71 0.20

188 22.44 < 0.000 ** 0.03 0.870 4.07 0.05 * 0.001), with higher mealybugs densities in the ant-infested treatment continuing up to the

last day ofsampling (188 days) (Fig. 4.5). Ants disappeared completely (visual

observations) from the treated plots somewhere between 13 and 56 days after treatment.

During this experiment, P. megacephala was the dominant ant species tending D. brevipes

on pineapple plants. Other ant species, glaber (Mayr), Hypoponera

punctactissima (Roger), H. zwaluwenbllrgi (Wheeler) and Stnlmigenys rogeri Emery,

were found on a few occasions, in low densities, and in scattered areas.

The presence ofnatural enemies in the ant-free plots was recorded 56 days after

bait treatment (experiment initiation), as compared to the ant-infested treatment where

natural enemies were first observed 89 days after experiment initiation (Fig. 4.6, 4.7 and

4.8). Four natural enemy species were associated with D. brevipes: the encyrtid

parasitoids, Anagyrlls ananatis Gahan and ElIryrhopallis propinqllus Kerrich, the

cecidomyiid predator, Lobodiplosis pselldococci (Felt) and the coccinellid predator,

Nephlls bilucemarius Mulsant. A. ananatis was the most common natural enemy in the ant-free plots, followed by N. bilucemarills. As D. brevipes densities decreased at 56 days after bait treatment, A. anGnatis densities increased in the absence ofants (Fig. 4.6).

Significantly higher A. ananatis densities were recorded in the ant-free plots 56 days (F =

9.99; df= 1; P = 0.004) and 89 (F= 6.75; df= 1; P = 0.015) after initial bait treatment

(Table 4.1). The presence ofA. ananatis was detected in ant-infested plots 89 days after experiment initiation, and continued to the end ofthe study. Significantly higher percentages ofmealybug parasitization by A. ananGtis were recorded in the ant-free plots

70 90 ~ ~ -0- Art-irte;ted 80 Pmdro Pmdro Pmdro -11- Art-free applied applied applied 70 -; 15. 60

} 50 >-

l 40 '0 0 z 30 ~ 20

10 .... 0 .--. .- 0 50 100 150 200 Days after initial ant t:ait appIicatic:n

Fig. 4.5. Mean number ofhealthy pink pineapple mealybugs (all stages except crawlers) per pineapple plant (aerial part) in ant-infested and ant-free experimental pineapple plots.

71 5 t t t -0- Art-irfested Pm:Iro Plrdro /lmdro -tI- Art-free appiaj appie:l appiaj -; 4 11. l f 3 T 1... .~ ! 2 "::( 0 - 1 ~ ~

a 50 100 150 200 Days after initial ant tait appication

Fig. 4.6. Mean numbers ofthe encyrtid Anagyms ananatis reared from pineapple mealybugs per pineapple plant (aerial part) in ant-infested and ant-free experimental pineapple plots.

72 60 .Amdro cql/ied Imdro appied 1-0- Pllt-infested I -11- Pllt-free 50

(I) 0) ::J ~40 cu ~ i 30 I!! a.cu 020 c: -~ Q) a.. 10

0 0 50 100 150 200 250 l:Bys after in~ial ant bait application

Fig. 4.7. Percent ofparasitization by Al1agynls ananatis on DysmicocclIs brevipes per pineapple plant (aerial part), in ant-infested and ant-free experimental pineapple plots.

73 1.4 + + + Prrdro appied Pm:lro appia::l -0- Mt-infested --- Mt-free I 1.2 ti ~ 1.0 ~ ~ 0.8 Ii '0 0.6 z0

~ 0.4

0.2

0.0 a 50 100 150 200 Days after initial ant bait appicatioo

Fig. 4.8. Mean number ofthe predators Lobodiplosispselldococci and Nephus billicemarilis per pineapple plant (aerial part) in ant-infested and ant-free experimental pineapple plots.

74 56 (F= 21.7; df= 1; P < 0.001),89 (F= 33.5; df= 1; P < 0.001), 126 (F= 5.03; df= 1;

P = 0.035), 153 (F= 10.8; df= 1; P = 0.003), and 188 days after experiment inititation (F

= 18.4; df= 1; P < 0.001) (Fig 4.7). Significantly higher predator densities (L.

pselldococci and N. billicemarills) were recorded in ant-free plots only at 56 days after

bait treatment (F= 4.24; df= 1; P = 0.05) (Table 4.1) (Fig. 4.8). In ant-infested fields,

the ratio ofmealybugs to A. anana/is individuals reared was significantly higher than ratios

in ant-free fields at 89, and 188 days after treatment (Table 4.2), indicating a greater

parasitoid presence in ant-free plots.

Combined Biological Check and Exclusion Method. There were no significant

differences in D. brevipes densities among treatments before application ofthe ant bait.

Twenty five days after experiment initiation, significantly more mealybugs were recorded

in ant-infested plots than in ant-free plots (F= 25.86; df= 1,44; P < 0.001) (Table 4.3).

Mealybugs remained significantly higher in ant-infested plots until termination ofthe experiment (Fig. 4.9). In the absence ofants, higher mealybug densities were observed in closed cages than in open cages 52 days after the start ofthe experiment (F = 5.95; df= 1,

44; P < 0.02). Mealybug densities on plants open to natural enemies started to decline 30 days after experiment initiation, and after 70 days were less than 2.0 mealybugs per pineapple plant. In the absence ofnatural enemies, mealybug densities were significantly higher in ant-infested plots than in ant-free plots 25 days after the start ofthe experiment

(F= 8.57; df= 1,44; P < 0.01) (Table 4.3), and on the remaining sample dates (Fig. 4.9).

In the presence ofants higher mealybug densities were observed in open cages than in closed cages, 70, 76, 84, 90 and 99 days after start ofexperiment. At the end ofthe

75 Table 4.2. Ratio ofdensities ofthe pink pineapple mealybug to the densities ofthe encyrtid Anagyrus ananatis in ant infested and ant-free pineapple plots.

Ant Status

Time (days) Ant-infested Ant-free P

0 0.00 0.00

13 321.00 0.00

56 0.00 1.13

89 41.80 1.21 1.86 0.04

126 13.95 3.84 -1.33 0.10

153 45.08 2.18 2.17 0.02

188 32.85 1.86 2.46 0.01

76 Table 4.3. Orthogonal contrast ofthe densities ofD. brevipes in big-headed ant-infested and ant-free cages, where natural enemies were excluded and permitted. Split plot design experiment.

Ants absent Ants present Natural enemies absent NE present & NE absent NE present & NE absent Ants absent & Ants present Time (days) F P F P F P 0 NS NS NS 12 NS NS NS 17 NS NS NS 25 NS NS 8.57 < 0.01 31 NS NS 32.86 < 0.001

~ 39 NS NS 55.79 < 0.001 -....l 46 NS NS 40.43 < 0.001 52 5.95 <0.02 NS 41.02 < 0.001 64 NS NS 23.27 < 0.001 70 NS 5.07 <0.05 18.41 < 0.001 76 NS 6.89 <0.05 6.06 <0.05 84 NS 9.44 < 0.01 10.20 < 0.01 90 NS 12.75 < 0.001 14.15 < 0.001 99 NS 5.60 <0.05 10.43 < 0.01 108 NS 12.17 < 0.001 12.17 < 0.001 115 25.47 < 0.001 10.49 < 0.01 10.49 < 0.01 140

Amdro applied Amdro applied 120 t 1 -cro -O-Ant-free + NE -.-Ant-free & NE-free Q. 100 -o-Ant-infested + NE & ~ -e Ant-infested NE free -a. ....a. 80 ...0 .cQl E 60 ::l C -.J C 00 ro Ql 40 ~

20 --.~ -~I~ ~-~ • :II~ ~-,[;L-J:;L_ _ ~_-. .~._._ o . :CJ-...,r:1_~.--n_r::t-r:'I-l7 0 10 20 30 40 50 60 70 80 90 100 110 120 Time (days)

Fig. 4.9. Mean numbers ofpink pineapple mealybugs per pineapple plant in ant-infested and ant-free pineapple plots, in the presence and absence ofnatural enemies (NE). experiment, A. anana/is was the only natural enemy found attacking D. brevipes on 4 out ofthe 12 plants in open cages. Although mealybugs were recorded on the pineapple roots at the end ofthe experiment, mean mealybug densities were significantly higher on pineapple roots in the ant-infested plots with open cages (37.6 mealybugs) than in ant­ infested plants with closed cages (20.3 mealybugs), ant-free plants with open cages (5.58 mealybugs), and ant-free closed cages (6.8 mealybugs) (Tukey HSD: F= 8.31; df= 3; P =

< 0.001). No natural enemies were detected attacking D. brevipes on the roots.

Impact ofP. megacephala on Mealybug Natural Enemies. N. bilucernarius.

In the absence ofants, N. billicemarilis adults consumed an average of53.6 D. brevipes crawlers in 24 hr, while in the presence ofants, only 14.6 crawlers were consumed (F =

53.6; df= 1; P < 0.001). Thus, mealybug predation was decreased 73% by the presence ofants. Ants killed three (out of 10) coccinellid adults. There were significant differences in the number ofmealybugs killed by the coccinellid larvae in the presence and absence of ants (F = 12.9; df= 1; P = 0.002). In the presence ofants, coccinellid larval predation was reduced 33.3 % (a mean of0.5 mealybugs killed) compared to predation in the ant­ free arena (a mean of 1.5 mealybugs killed). Ants killed 8 (out of 10) coccinellid larvae.

Ants consumed 79% ofthe 100 coccinellid eggs placed in the arena cages.

Anagyrus ananatis. P. megacephala interfered with the searching behavior ofA. anana/is adults. Significantly more mealybugs were parasitized by A. anana/is when ants were absent (15.9 mealybugs per female wasp) than when present (8.3 mealybugs per female) (F = 6.59; df= 1; P = 0.019). Percent parasitization by A. anana/is in the presence ofants was 48% lower than in the absence ofants.

79 Discussion

D. brevipes had the highest densities in the field when ants were present, and

natural enemies could not access them as readily as compared to an ant-fee environment.

The four natural enemies recorded causing D. brevipes mortality were less commonly

found when ants were present. These natural enemies, and particularly A. allallatis,

apparently were responsible for the decline ofmealybugs densities in the ant-free plots 56

days after initial ant bait application. Illingworth (1931) made greenhouse observations on

the presence oftwo unidentified coccinellid predators in ant-free pineapple plots.

Although he did not test the protection hypothesis, he suggested that ants protected

mealybugs from these natural enemies. Jahn (1992) working in pineapple on the island of

Maui suggested that P. megacephala suppressed populations ofthe coccinellid Clirinlis coen,lelis Mulsant preying on D. lleobrevipes. He observed that in ant-free plots there was an increase in coccinellid densities that appeared to be associated with in a decline of mealybug populations. Similar findings have been observed in other crop systems where

P. megacephala is the dominant ant species. Reimer et al. (1993) concluded that P. megacephala was responsible for high densities ofCoccus viridis (Green) infesting Coffea arabica L. High scale densities resulted when ants interfered with natural enemy activity.

Cudjoe et al. (1993) conducted an interference experiment in Ghana and found that P. megacephala, Camponotlls sp. and Crematogater sp. reduced parasitization rates by the encyrtid Epidillocarsis lopezi (De Santis), a parasitoid ofthe cassava mealybug,

Phenacocclis manihoti Matile-Ferrero.

80 Results shown here are consistent with the idea that Homoptera benefit from ant

association when ants promote sanitation within mealybugs infestations, and! or protect

mealybugs from natural enemies or by other interactions (such as construction ofmud

shelters on the Homoptera colonies).

In the combined biological check! exclusion method study, there were lower mean

mealybug densities in open cages, from 46 to 108 in the ant-free section. However, a

significant difference in densities between closed and open cages in ant-free plots only

occurred at 52 days after experiment initiation. This suggests that natural enemies may

cause mortality in the mealybug population. Mealybugs densities increased on plants

within closed cages in ant-free plots f~om 31 to 52 days after the experiment was started.

Additionally, in the closed cages, there were more mealybugs in the presence ofants than

in the absence ofants. The decline in mealybugs in ant-free closed cages (after day 52)

suggests that ants may have some influence on the mealybug population other than

protecting mealybugs from their natural enemies. The decline could be related to the lack

ofhoneydew removal. In contrast, Jahn (1992) concluded, from a laboratory experiment,

that sanitation by P. megacephala had no effect on D. neobrevipes survival. However, his

study was conducted using a different mealybug species and ran for 56 days compared

with the 120 days during which this system was studied. In a field experiment, Jahn

(1992) found that the only beneficial effect ofthe ants onD. neobrevipes survival was

protection from natural enemies. Nevertheless, he did not exclude natural enemies from

experimental pineapple plots to test other hypotheses (other than sanitation). Similar results were obtained by Reimer et al. (1993) but in a different crop system and with

81 another species. They found that on coffee trees, P. megacephala enhanced

survival ofgreen scale, Coccus viridis (Green), by interfering with the scale's natural

enemies.

In the presence ofants and absence ofnatural enemies, mealybugs were expected

to have the highest densities than in the presence ofants and presence ofnatural enemies;

despite this, the contrary was observed. One explanation could be related to the micro­

environment inside the closed screen cages. When temperature was recorded inside cages

it was approximately 4° C higher in the closed screen cages than in the open screen cages.

The higher temperature may have been detrimental when mealybug densities were high.

Also the closed environment ofthe closed cages, may have kept humidity high. P. megacephala may avoid high humidity, resulting in less honeydew removal, and reduced mealybug numbers. The presence ofnatural enemies may have changed the ants' behavior, causing a reduction in the time devoted to removal ofold and dead mealybugs and in an increase in the time spent guarding mealybugs and pursuing intruding natural enemies.

Comparison ofthe results obtained in biological check and the combined biological check and exclusion methods suggests that the former technique in not adequate to evaluate natural enemy efficacy ifincreases in mealybug densities are partially associated with sanitation provided by the ants.

The effect ofthe ants on D. brevipes infesting the roots ofpineapple has not been well studied in Hawaii. Although significant differences in mealybug densities attacking the roots were observed between ant-infested and ant-free pineapple areas at the end of

82 the experiment, this is not sufficient evidence that ants are essential for mealybug survival

on pineapple roots. Jahn (1992) observed that the presence ofD. brevipes infesting roots

ofexperimental potted pineapple plants was independent ofthe presence ofP.

megacephala. However, Petty and Tustin (1993) found thatD. brevipes pineapple root

infestation decreased to zero 12 weeks after ant elimination.

Results oflaboratory experiments indicated that P. megacephala had a negative

impact on mealybug mortality induced by A. ananatis and N. bilucernarius adults. Ants

interfered mostly with natural enemy searching behavior. However, ants actually killed

immature stages ofN. bilucernarills, regardless ofthe fact that the larvae morphologically

mimic their prey. The aggressiveness ofants toward natural enemies oftended­

Homoptera depends on several factors: the ant species involved, Homoptera density, and

the natural enemy searching and handling behavior. The tree ant, Oecophylla smaragdina

(p.), may attack any insects that enter its arboreal habitat, while Lasius and Formica are less aggressive towards enemies oftended aphids (Nixon 1951). Ants are more aggressive towards intruders when coccids are at low densities (Way 1963). In contrast, the attitude ofants toward other competitors is much less antagonistic when high coccid densities exist and large amounts ofhoneydew are available (Nixon 1951). Flanders (1951) stated that ants cause a high degree ofinterference in the behavior ofparasitoids that exhibit host feeding. Bartlett (1961) found that with two ant-tolerant parasitoids, Metaphycus stanleyi

(Comp.) and Coccophaglls scutellaris Dalm., that require long oviposition times, the presence ofL. hllmile caused reductions in the rate ofparasitism of65 and 27%, respectively. This may be compared to the ant-sensitive parasitoid, Metaphiclls luteolus

83 Timblake. Although it oviposits quickly, the presence ofL. humile reduced its

parasitization rate by 98%. A. ananatis also displays rapid oviposition (less than 10 sec.)

and is easily disturbed by the presence ofP. megacephala. IfA. ananatis and N.

bilucemarills have evolved to avoid ant interference, they have not successfully modified their oviposition or predation behaviors, respectively, to totally circumvent ant aggression.

However, this does not mean that they are not efficient natural enemies in the absence of ants. It has been observed that natural enemies adapted to the presence ofants (ant­ tolerant species) have little effect on their host densities compared to ant-sensitive natural enemies (Flanders 1958, Bartlett 1961, Way 1963).

A. ananatis, the parasitoid ofD. brevipes, as well as the predator N. bilucemarills that preys on D. brevipes and D. neobrevipes (see Chapter III) could potentially be used in mass release programs to control mealybugs in ant-infested pineapple fields via inundative releases. This suggestion is based on several factors. A. anana/is is the predominant parasitoid ofD. brevipes. Ants interfere with A. anana/is and N. bilucemarius adult searching behavior but do not actually kill them. Therefore, periodic augmentation offield densities ofthese natural enemies, would increase the ratio of natural enemies to ants and increase the probability ofbiological control without the necessity ofeliminating ants.

84 References Cited

Banks, W. A, C. S. Lofgren, D. P. Jouvenaz, C. E. Stringer, P. M. Bishop, D. F.

Williams, D. P. Wojcik, and B. M. Glancey. 1981. Techniques for collecting,

rearing, and handling imported fire ants. U. S. D. A. Science and Education

Administration. Advances in Agricultural Technology. A 106.24:S-21, 9 pp.

Bartlett, B. R. 1961. The influence ofants upon parasites, predators, and scale insects.

Ann. Ent. Soc. Amer. 54: 543-551.

Bess, H. A 1958. The green scale, Coccus viridis (Green) (Homoptera: Coccidae), and

ants. Proc. Hawaii. Entomol. Soc. 16: 349-355.

Bishop, P. M., W. A Banks, D. F. Williams, C. E. Stringer, 1. A. Mittchell, and C. S.

Lofgren. 1980. Simple nest for culturing imported fire ants. J. Georgia Entomol.

Soc. 15: 300-304.

Buckley, R. C. 1987. Interactions involving plants, Homoptera, and ants. Ann. Rev.

Ecol. Syst. 18: 111-135.

Carter, W. 1932. Studies ofpopulations ofPseudococcus brevipes (Ckl.) occurring on

pineapple plants. Ecology 3: 296-304.

Carter, W. 1933. The pineapple mealybug, PselldococclIS brevipes (Ckl.) and wilt of

pineapple. Phytopathology 23: 207-242.

Carter, W. 1967. Insects and Related Pests ofPineapple in Hawaii. A manual for

fieldman. Pineapple Research Institute (Restricted publication) pp. 36-53.

85 Cudjoe, A. R., P. Neuenschwander, and M. 1. W. Copland. 1993. Interference by ants in

biological control ofthe cassava mealybug Phellacocclis manihoti (Hemiptera:

Pseudococcidae) in Ghana.. Bull. Entomol. Research 83: 15-22.

DeBach, P., and C. F. Huffaker. 1971. Experimental techniques for evaluation ofthe

effectiveness ofnatural enemies. pp. 113-140. Ill: Biological Control. C.F.

Huffaker (ed.), Plenum Rosetta Ed. New York, NY.

Flanders, S. E. 1951. The role ofthe ant in the biological control ofhomopterous insects.

Can. Ent. 83: 93-98.

Flanders, S. E. 1958. The role ofthe ants in the biological control ofscale insects in

California. Proc. Inter. Congo Entomol. 10th. Montreal, 4: 579-584.

Fluker, S. 1969. Sympatric association among selected ant species and some effects on

sugarcane mealybugs in Hawaii. Ph.D. Dissertation, University ofHawaii,

Honolulu, Hi.

Gray, R. A. 1952. Composition ofhoneydew excreted by pineapple mealybugs. Science

115: 129-133.

Gross, P. 1993. Insect behavioral and morphological defenses against parasitoids. Ann.

Rev. Entomol. 38: 251-273.

Hanks, L. M., and C. S. Sadof 1990. The effect ofants on nymphal survivorship of

Coccus viridis (Homoptera; Coccidae). Biotropica 22(2): 210-213.

Heidari, M, and M. 1. W. Copland. 1993. Honeydew: a food resource or arrestant for the

mealybug predator Cryptolaemlls mOlltrollzieri? Entomophaga 38(1): 63-68.

86 Holldobler, B. and E. O. Wilson. 1990. The ants. The Bleknap Press ofHarvard

University Press, Cambridge, MA.

Illingworth, J. F. 1931. Preliminary report on evidence that mealybug are an important

factor in pineapple wilt. 1. Econ. Entomol. 24: 877-889.

Jahn, G. 1992. The ecological significance ofthe big-headed ant in mealybug wilt disease

ofpineapple. Ph.D. Dissertation, University ofHawaii, Honolulu, HI.

Jahn, G., and 1. W. Beardsley. 1994. Big-headed ants, Pheidole megacephala:

interference with biological control ofgray pineapple mealybugs. pp: 199-205.

Ill: Exotic ants: Biology, Impact, and control ofintroduced species. D. F.

Williams (ed.),. Westview Studies in Insect Biology. Boulder, Colorado.

Nechols,1. R. and T. F. Seibert. 1985. Biological control ofthe spherical mealybug,

Nipaecoccus vas/a/or (Homoptera: Pseudococcidae): assessment by ant exclusion.

Environ. Entomol. 14: 45-47.

Nixon, G. E. 1. 1951. The association ofants with aphids and coccids. Commonwealth

Institute ofEntomology London 36 pp.

Petty, G. L. 1978. The pineapple mealybug. Pineapple H.15. Farming in South Africa 8

p.

Petty G. 1., and H. Tustin. 1993. Ant (Pheidole megacephala F.) - mealybug

(Dysmicoccus brevipes Ckll.) relationships in pineapples in South Africa. First

International Pineapple Symposium, Honolulu, Hawaii, Acta Horticulture 334:

387-395.

87 Phillips, 1. S. 1934. The biology, distribution and control ofants in Hawaiian pineapple

fields. Ph.D. Dissertation, University ofHawaii, Honolulu, Hi.

Plank, H. K. and M. R. Smith. 1940. A survey ofthe pineapple mealybug in Puerto Rico

and preliminary studies ofits control. 1. Agric. Univ. ofPuerto Rico 24: 49-75.

Reimer, N. 1. 1993. Distribution and impact ofalien ants in vulnerable Hawaiian

ecosystems. pp. 11-22. In: Exotic Ants. Biology, impact and control of

introduced species. D. F. Williams (ed.). Westview Press, Boulder, Colo.

Reimer, N., 1. W. Beardsley, and G. Jahn. 1990. Pest ant in the Hawaiian Islands. pp.

40-50. Ill: Applied Myrmecology. A world perspective. R. K. Vander Meer, K.

Jaffe and A. Cedeno (ed.). West View Press, Boulder, Colo.

Reimer, N. 1., M. Cope and G. Yasuda. 1993. Interference ofPheidole megacephala

(Hymenoptera: Formicidae) with biological control ofCoccus viridis (Homoptera:

Coccidae) in coffee. Environ. Entomol. 22(2): 483-488.

Rohrbach, K. G., 1. W. Beardsley, T. L. German, N. 1. Reimer, and W. G. Sanford. 1988.

Mealybug wilt, mealybugs, and ants on pineapple. Plant Disease 72: 558-565.

Stein, J. J. 1954. The effect ofthe cosmopolitan brown house ant (Pheidole megacephala

F.) on citrus red scale (Aonidella aurantii Mask.) at Letaba. 1. Entomol. Soc. S.

Africa 17: 252-264.

Way, M. J. 1963. Mutualism between ants and Honeydew-producing Homoptera. Ann.

Rev. Entornol. 8: 307-344.

Wilkinson, L. 1990. Systat: the system for statistic. Evanston, IL.

88 v. HOST PREFERENCE, PARASITIZATION CAPACITY, AND FUNCTIONAL

RESPONSE OF ANAGYRUS ANANATIS GAHAN (HYMENOPTERA:

ENCYRTIDAE), A PARASITOID OF DYSMICOCCUS BREVIPES

(COCKERELL) (HOMOPTERA: PSEUDOCOCCIDAE)

Introduction

Attempts to biologically control mealybugs in the genus Dysmicoccus in Hawaii

were started around 1894 with the introductions ofseveral natural enemies from Central

and South America. These attempts were unknowingly hampered because the systematics ofthe Dysmicocclls spp. in Hawaii were unclear (Beardsley 1959). The pineapple mealybug, Dysmicocclis brevipes (Cockerell), was only recognized although D. neobrevipes Beardsley was also present. D. neobrevipes was considered a distinct strain

(the gray form) ofD. brevipes. By 1930, 34 natural enemy species were introduced against D. brevipes, but only eight became established and without appreciable effectiveness (Chapman 1938). More successful biological control agents were introduced in the 1930's including the predator Lobodiplosis pseudococci (Felt) (Diptera:

Cecidomyiidae) from Mexico (Chapman 1938, Carter 1944); the encyrtid parasitoids

Anagynls ananatis Gahan (mis-identified as Anagyms coccidivoms Dosier), Hambletonia pseudococcina, and Euryrhopailis propinqulls Kerrich (Compere 1936, Carter 1937,

Bartlett 1978), and the predator Nephus bilucemarills Mulsant (Coleoptera:

CoccineIlidae) (Swezey 1939, Funasaki et al. 1988) from Central and South America.

89 These natural enemies established and partially controlled mealybugs colonies in the absence ofants (Carter 1945, 1967, Zimmerman 1948, Bartlett 1978).

Only visual observations were used to evaluate the effectiveness ofthese natural enemies. For example, Carter (1944) visually correlated low mealybug densities in the presence ofthe predator 1. pseudococci. He stated that this predator was the best established ofany introduced natural enemy. However, Carter (1945) later made field observations and suggested that A. ananatis and H. pselldococcina were effective in controlling pineapple mealybugs in one pineapple field in west MauL He observed high a degree ofparastism ofpineapple mealybugs. Little information exists about basic biological parameters (host preference, fecundity and longevity) or effectiveness ofthese natural enemies. A. ananatis and H. pselldococcina have been reported to attack D. brevipes only (Beardsley 1976). While E. propinqulIs was reported to parasitize both D. brevipes and D. neobrevipes (Beardsley 1976). 1. pseudococci has a broader prey range, and has been reported preying on the mealybugs D. brevipes, Nipaecoccus vastator

(Maskell), Planococclls cirri (Risso), and Saccharicoccus sacchari (Cockerell) (Funasaki et aI. 1988). The predator N. billicernarilis has been reported to attack the mealybugs D. brevipes and N. vastator. Clansy and Pollard (1947) described A. ananatis as a some what polyphagous wasp. Also, they observed that in field situations a generation ofthis parasitoid was completed in about 20 days. Jahn (1992) found that in pineapple plots free ofthe big headed ant, Pheidole megacepahala (F.), there was an increase in generalist predators (coccinellid Clirinlis coerlilelis Mulsant). He suggested the need for identification ofthe most efficient natural enemies ofthe pineapple mealybugs.

90 Gonzalez (Chapter III) surveying Dysmicocclis infested pineapple plantings in

Hawaii found the coccinellid predators N. billicemarius and S. ruficops Weise, the

cecidomyiid predator L. pseudococci, and the encyrtid parasitoids A. ananatis and E.

propinquus attacking the mealybugs. In another study, Gonzalez (Chapter IV) evaluated

the impact ofants on D. brevipes and its natural enemies and observed that A. ananatis

was the most common natural enemy, and appeared to impact D. brevipes field densities

in the absence ofants. In pineapple fields on Oahu and Maui, D. brevipes may be the

most important pineapple mealybug, where it occurs alone and in mixed populations with

D. neobrevipes. No hyperparasitoids were detected attacking A. ananatis, while on the

other encyrtid present, E. propinqulIs, a hyperparasitoid (Prochiloneurus sp.) was

detected. An important aspect ofthe genus Anagynls is that it is the in the tribe Anagyrini

that is successfully used in pest control worldwide. At least 15 introductions ofAnagyrini

have resulted in complete or partial control ofthe targeted pest species (Noyes and Hayat

1994). Thus, given the lack ofbiological information available and the potential ofA.

Qnanatis as a biological agent, studies were undertaken to determine A. ananatis host

preference, host stage preference, adult longevity, sex ratio, parasitization capacity and

functional response. This information will be used to appraise the role ofthis parasitoid in

mealybug control and to develop mass rearing protocols for this species.

Materials and Methods

Mealybugs and the parasitoid were initially collected from a pineapple planting in

Kunia, Oahu Is., Hawaii. D. brevipes and D. neobrevipes first instars were collected and

91 later separately reared on butternut squash under laboratory conditions of 26 ± 2°C and

14 hr photophase. Mummified mealybugs were isolated from those field collections and

held for parasitoid emergence. Emerged A. ananatis adults were reared in the laboratory

on D. brevipes infesting butternut squash, under the same laboratory conditions as used to

rear D. brevipes. These mealybugs and parasitoid colonies were reared for one year

before they were used in this experiment. In all experiments, exposure ofmealybugs to the

parasitoids was accomplished by infesting newly grown Smooth Cayenne pineapple leaves

approximately 3 cm at the base and 15-20 cm in length (obtained from a 2 month old greenhouse planting). Leaves were initially infested with mealybugs 24 hr before exposure to A. ananatis. All experiments were conducted under laboratory conditions (26 ± 2°C;

14 hr. photophase).

Host Species Preference. Laboratory studies were conducted to determine susceptibilities ofD. brevipes and D. neobrevipes to A. ananatis parasitization. Choice and no-choice experiments were conducted for this study. For the choice experiment, pineapple leaves were individually infested with five adult females ofeach D. brevipes and

D. neobrevipes. In the no-choice experiment, pineapple leaves were individually infested with either 10 adult females ofD. brevipes or D. neobrevipes. Infested pineapple leaves were individually kept inside a polyethylene container (5 liter capacity). One day after infesting leaves, an A. ananatis female and male (2-3 days old) were introduced into the container. After 24 hr, parasitoids were removed, and the mealybugs were transferred to a

37 ml polyethylene vial. Vials were capped with a lid with a hole (1.2 cm diam.) covered

2 with fine mesh (1,369 cells / cm ). Mealybugs were held in vials until they mummified or

92 parasitoids adults emerged from parasitized mealybugs. Numbers ofparasitized

mealybugs from each treatment were recorded. Treatments were replicated 20 times in

both sets ofexperiments. Data were analyzed with ANOVA and Tukey mean comparison test (Wilkinson 1990).

Host Stage Preference. Laboratory studies were conducted to detennine which developmental stages ofD. brevipes were susceptible to oviposition by A. ananatis.

Choice and no-choice experiments were conducted. In choice experiments, a pineapple leafwas individually infested with 3 individuals each ofD. brevipes first, second, third instars and adult females. In no-choice experiments, immature and adult mealybugs stages were exposed separately in groups of 10. Two day old adult A. ananatis females were used. After 24 hr exposure mealybugs and parasitoids were processed and analyzed as described above.

Reproduction, Longevity, and Sex Ratio. Laboratory studies were conducted to determine reproduction ofA. ananatis reared on D. brevipes. A pineapple leafwas infested with 30 (15 third instars and 15 adult females) to 50 adult females ofD. brevipes.

The pineapple leafwas kept inside a 5 liter polyethylene container. Twenty four hr later, a two day old parasitoid female and male were introduced to the container. After 24 hr, parasitoids were transferred to a new container, having the same mealybug density, until the female parasitoid died. Treatments were replicated 20 times. Oviposition period, number ofparasitized hosts or offspring produced by the parasitoid, and longevity of females and males were recorded and analyzed.

93 Functional Response ofAnagyrus anllnatis. Laboratory studies were conducted to determine A. ananaJis functional response when using D. brevipes as a host. Pineapple leaves were individually infested with 1, 5, 10,25,50 and 100 adult mealybug females (2-

5 days old). Each mealybug density was replicated 10 times. Each leafwas maintained in a 5 liter polyethylene container. Twenty-four hr later, one two-day old adult female parasitoid was introduced into the polyethylene container. D. brevipes were exposed to parasitoids for 24 hr. Exposed mealybugs were held on leaves until emergence ofadult parasitoids. Chi-square analysis was conducted to test independence between the observed and expected rates ofhost attack (Wilkinson 1990). Expected values were obtained using the random parasitoid equation proposed by Rogers (1972). This equation assumes an exponential decay in host density, and it is useful for long term closed experiments where no replacements are done (Houck and Strauss 1985):

- (T, * a' *P,) / (1 + a' * Th *Nt) N par = Nt [1 - e ]

Where N par = the number ofhosts parasitized; Nt = the number ofhosts present; Pt =the number ofparasites present; a ' = the instantaneous attack rate; Th = the handling time; and Tt = the total time available. The parameters a . and Th were obtained by linear regression analysis of 1/ In (proportion ofhosts surviving) as a function ofhost density

(Wilkinson 1990).

94 Results

Host Species Preference. In the choice experiment, A. ananatis significantly

preferred to attackD. brevipes over D. neobrevipes (F= 59.2; df= 1; P < 0.001) (Table

5.1), parasitizing 11.5 fold more individuals ofthe former versus the latter species.

From exposed D. neobrevipes three mummies were produced, but adult parasitoids

emerged from two mummies only. Similarly in no-choice tests, A. ananatis attacked only

D. brevipes (F= 40.9; df= 1; P < 0.001) (Table 5.1), and did not parasitize a single D.

neobrevipes.

Host Stage Preference. In the choice experiment, A. ananatis preferred to attack

D. brevipes adult females, and in only one case was a mealybug nymph (second instar)

parasitized (Table 5.2). Analysis indicated that D. brevipes adults were preferred over all other stages (F= 39.3; df= 3; P < 0.001). Sex ratio was determined in this experiment because in other encyrtids progeny sex ratio are related to the size and stage ofthe selected host (Nechols and Kikuchi 1985, de Jong and van Alphen 1989). Only 8 females

(sex ratio =8: 0 F ; M) were obtained from mummified female mealybugs and no males.

In the no-choice experiment, A. ananatis attacked D. brevipes second and third instars, and adult females, with 67 % ofthe attacks being made on adult females (Table 5.2). A. ananatis preferred to parasitize adult females the greatest followed by third instar D. brevipes (Tukey HSD) (Table 5.2). Sex ratio ofA. anana/is that emerged from parasitized third instars was 1.0 : 0.6 (F : M), and from adult females was 1.0 : 0.2 (F :

M).

95 Table 5.1. Results ofchoice and no-choice experiments to determine host species preference ofA. anQnalis for D. brevipes and D. neobrevipes.

Mean number ofparasitized mealybugs

Treatment D. brevipes D. neobrevipes

Choice 3 2.3 * 0.2

No-choice b 2.6 * o

3 Based on 5 mealybugs ofeach species per experimental container. b Based on 10 mealybugs ofone species per container. * Significantly different (P < 0.001).

96 Table 5.2. Choice and no-choice experiments to determine host stage preference ofA. ananatis when reared on the pink pineapple mealybug.

Mean number ofparasitized mealybugs:!

Nymph instars

Treatment I II III Adult

Choice O.OOb 0.05 b O.OOb 1.45 a

No-choice O.OOd 0.10 c 1.10 b 2.45 a a Means followed by the same letter in a row are not significantly different as determined by Tukey HSD mean comparison test.

97 Reproduction, Longevity, and Sex Ratio. Over its adult life, A. ananatis

parasitized a mean of27.7 mealybugs, with a range of 11-51 parasitized mealybugs. The

mean parasitization rate per day was 3.4 (± 1.6) mealybugs. The maximum mean number

ofmealybugs parasitized on a single day by A. ananatis was 6.6 ±4.3 and that was

obtained on the first day ofexposure. The oviposition period lasted for a mean of6.6

days (± 2.09) (Fig. 5.1 A).

A. ananatis females had a mean longevity of9.8 days (± 1.97), while the males

lived for a mean of 10.8 days (± 2.80) (Fig 5.1 B). Immature males completed their

development (egg to adult) in 21.2 days, while immature females required 23.3 days. The

sex ratio ofA. ananatis that emerged from both third instars and adult females was 1.0 :

0.8 (F: M).

Functional Response ofAnagyrus ananatis. The number ofhosts attacked

increased as mealybugs densities were increased from 1 to 25 individuals, and then

plateaued at densities above 25 individuals (Fig. 5.2). This observation was characteristic

ofthe type II model ofthe Holling disk equation (Murdoch 1973). Linear regression of

I/ln (mean proportion ofhosts surviving) as a function ofhost density provided an estimation ofthe instantaneous attack rate, a', as 0.413 (= - a'lTt ) and the handling time,

Th, as 2.6 hr (= - bITt) (Fig. 5.3). The estimated handling time was high (I67-fold greater) compared to the observed handling time ofapproximately 5.1 (± 1.9) sec. In estimating the parameters ofinstantaneous attack rate and handling time, total time was set at 24 hr as used in the study. Chi-square analysis ofthe observed and expected host attack rates

98 7

A) ~6 .[ (;j 5 ~ "i ~ 4 I! ~3 ~ c nI III ~ 2

o 2 4 6 8 10 12 14 16 Day of exposure

, - Females I o Males 20- - B) -

.

5

o 2 4 6 8 10 12 14 16 Days of exposure

Fig. 5.1. Reproduction ofA. anallatis when reared on D. brevipes: A) daily mean number ofparasitized mealybugs; B) female and male parasitoids surviving.

99 2

0 y =- 0.107 x - 0.101 r =0.993 P < 0.001 en c: -2 '5 .~ ~ -4 1ii ..c:0 c: :e0 -6 8. §. c: -8 ..-::::::

-10

-12 0 20 40 60 80 100 H::lst density

Fig. 5.2. Linear regression ofthe lI(ln ofthe mean proportion ofD. brevipes hosts surviving parasitization) by A. ananatis as a function ofhost density.

100 12

l!! 10 ;:, .c0 8 I ~ I -8 (1J I =fa 6 I en 'tiS .c0 [•--ExpededObseMd I '0 4 .... iI I E 2 a' = 0.413 ~ Th= 2.6 hr

0 0 20 40 60 80 100 Host density

Fig. 5.3. Functional response ofthe observed and expected number ofD. brevipes hosts attacked by A. Gnanatis as host density increased.

101 showed independence (Chi-square = 14.3; df= 5; P> 0.05), indicating that the predicted

values did not correspond with the observed values.

Discussion

A. ananatis preferred to attack adult females ofD. brevipes over D. neobrevipes

females. D. brevipes appeared to be host specific in its attacks and has been reared in

Hawaii from D. brevipes only (Beardsly 1976). Availability ofsuitable host stages for

parasitoid development determines how the parasitoid exploits an insect host population

(Nechols and Kikuchi 1985). When a choice was possible, A. ananatis preferred to

oviposit and successfully parasitize adult D. brevipes females over immature stages. This

suggests that A. ananatis preferred the larger hosts to parasitize. Similar observations

were obtained by Nechols and Kikuchi (1985) with Anagynls agraensis Saraswat (syn.

non. ofA. illdicus Shafee) parasitizing Nipaecoccus vastator (Maskell). They observed in

choice experiments that A. agaellsis preferred to oviposit in the largest mealybug hosts

(third instar and adult female) provided. Similar conclusions were made by de Jong and

van Alphen (1989) regarding Leptomastix dactylopii Howard parasitizing PlallocliccUs

citri (Risso), and by Cross and Moore (1992) studying Allagynls mangicola Noyes

parasitizing Rastrocclls illvadells Williams. In contrast, the following encyrtid species

preferred to parasitize the smaller host stages (first to second instars): Leptomastix (=

Paraleptomastix) abllormis (Girault), a parasitoid ofP. citri (Smith 1917); Clausellia josefi Rosen, a parasitoid ofPiallococclis vitis (Niedieslsky) (Berlinger 1973), and

Gyranusoidea tebygi Noyes, a parasitoid ofR. invadens (Willingks and Moore 1988).

102 Knowledge ofpreferred host species and host stages is essential to designing efficient

mass rearing protocols and developing augmentative programs against this mealybug using

A. ananatis. Effective timing ofA. ananatis releases should be based on the presence of

early infestations ofthird instar and adult females.

Host stage and host size have been directly related to progeny production in some

encyrtid species. This phenomenon has been observed in Anagyros mangicola Noyes

parasitizing R. invadens on which the female allocates males to smaller hosts (second

instar to adult female) and females to larger hosts (Cross and Moore 1992). In the no­

choice host stage preference experiment, A. ananatis showed the same behavior ofA.

mangicola by preferentially parasitizing adult females, although second and third instars

were parasitized also.

Results suggest that A. ananatis parasitizes most ofthe hosts it is biologically able

to parasitize during the first days after adult emergence. Eighty-eight percent ofthe

parasitized mealybugs were attacked during the first six days ofadult life. The encyrtid C. josefi produced 50% ofits offspring during the first four days ofits adult life (Berlinger

1973). A. ananatis parasitized a mean of3.5 mealybugs per day totaling 27.7 parasitized

mealybugs over its adult life. This value is consistent with those ofencyrtid species

successfully used against other economically importance mealybugs. L. abnormis was

able to produce a mean of33.1 progeny over its life when P. citri was used as a host

(Viggiani and Maresca 1977). When attacking Maconellicoccus hirslltlls (Green),

Anagyros dactylopii (Howard) produced a mean of39.3 offspring. Tingle and Copland

(1989) observed that A. pselldococcii, L. dactylopii and L. abnormis had mean progeny

103 production of 7.3, 6.2 and 6.4 offspring per female, respectively, per 5 hr ofexposure, when P. citri was used as a host. 1. dactylopii and 1. abnormis have been used in biological control programs against P. citri in many countries ofthe world (Noyes and

Hayat 1994).

The longevity ofA. anana/is adult females and males was shorter than other

Anagynls spp. reported in the literature. For example, females and males ofA. dactylopii had a longevity of22 and 14.5 days (25-280 C), respectively (Mani and Thontadarya

1988), and females and males ofA. mangicola lived for 20.2 and 16.7 days (270 C), respectively (Cross and Moore 1992). In contrast, both females and males of1. abnormis lived for 3.6 days (25-25° C) (Viggiani and Maresca 1977). Nevertheless, the short longevity ofA. anana/is adults appears adequate to successfully parasitize an effective number ofmealybugs. Because developmental time ofD. brevipes is also short in Hawaii and overlapping generations exist all year round, adult longevity may not be as important as in more temperate environments.

The curvilinear response obtained by the nonlinear regression equation ofRogers

(1972) apparently did not adequately describe the observed rates ofhost attack by A. anana/is on D. brevipes. There are several factors that may explain the mechanisms involved in this parasitoid's host searching behavior. For example, the estimated large handling time of2.6 hr used to fit the curve was high and unrealistic compared to that directly observed in the laboratory. A handling time between 5.1 (± 1.9) seconds was observed at host densities of10, 25, 50 and 100 mealybugs per unit. Thus, the large handling time estimated using Rogers (1972) technique made it difficult to fit the data to

104 the curve. Weidemann and Smith (1993) when studying the functional response ofthe

braconid parasitoid, Cotesiaflavipes (Cameron) on the sugarcane borer, Diatraea

saccharalis (F.) were unable to fit the observed rates ofhost attack to the random

parasitoid model. Instead they obtained a linear functional response. They observed that

handling time generated by a nonlinear least square fitting method was artificially high and

unreliable as found here with A. ananatis. Another limiting factor that influences the final

rate ofhost attack may be related to the capacity ofthe parasitoid to produce eggs.

Dissections ofmated A. ananatis females, prior to exposure to suitable mealybug hosts, revealed that the maximum number ofmature eggs found was 24. Observations indicate that A. ananatis stops ovipositing after a given number ofhosts are attacked, even when it is provided with a new group ofhealthy , unparasitized mealybugs. Therefore, egg depletion may be a factor limiting the rate ofhost attack for this parasitoid and not handling time. Collins et al. (1981) found that in Aphelilll1s thomsoni Graham, a parasitoid ofthe sycamore , Drepahnosiphllm platanoids (Schrank), the rate of attack decreased at high host density as a result ofegg depletion.

In a laboratory experiment looking at the impact ofants onthe host searching behavior ofA. ananatis (see Chapter IV), it was observed that ants interfered with the searching behavior ofthis parasitoid. Further laboratory and field studies are recommended to determine this parasitoid's functional response in the presence ofants.

105 References Cited

Bartlett, B. R. 1978. Pseudococcidae. pp. 137-170. /11 C. P. Clausen (ed.), Introduced

parasites and predators ofarthropod pest and weeds: a world review. U.S.D.A.

Handbook. No. 480.

Beardsley, 1. W. 1959. On the taxonomy ofpineapple mealybug in Hawaii with the

description ofa previous unnamed species (Homoptera: Pseudococcidae). Proc.

Hawaii. Entomol. Soc. 17: 19-37.

Beardsley, 1. W. 1976. A synopsis ofthe Encyrtidae ofthe Hawaii. Islands with keys to

genera and species (Hymenoptera: Chalcidoidea). Proc. Hawaii. Entomol. Soc.

22: 181-228.

Berlinger, M. 1. 1973. Biological studies ofClauselliajoseji (Hym. Encyrtidae), a

parasite ofPiallococclis vilis. Entomophaga 18: 279-286.

Carter, W. 1937. Importation and laboratory breeding oftwo chalcid parasites of

Pselldococcus brevipes (Ckl.). 1. Econ. Entomol. 30: 370-372.

Carter, W. 1944. Biological control ofPseudococclis brevipes (Ckl.). Proc. Hawaii.

Entomol. Soc. 12: 15.

Carter, W. 1945. Encyrtid parasites ofPselldococcus brevipes (Cockerell). Proc.

Hawaii. Entomol. Soc. 12: 489.

Carter, W. 1967. Insect and Related Pests ofPineapple in Hawaii. A manual for

Fieldman. Pineapple Research Institute (Restricted publication). pp.36-53.

Chapman, R. N. 1938. Biological control (Project 26). Experimental Station ofPineapple

Producers Cooperative Association. Honolulu, Hi. Monthly Report, December,

pp.319-333.

Clancy, D. W., and H. N. Pollard. 1947. Further experiments in parasitization of

mealybugs. 1. Econ. Entomol. 40: 578-579.

106 Collins M. D., S. A. Ward, and F. G. Dixon. 1981. Handling time and functional response

ofAphelil1l1s thomsol1i, a predator and parasite ofthe aphid Drepal10siphum

platal1oids. 1. Anim. Ecol. 50: 479-487.

Compere, H. 1936. A new genus and species ofEncyrtidae parasitic in the pineapple

mealybug, Pseudococcus brevipes (Ckl.). Proc. Hawaii. Entomol. Soc. 9: 171­

174.

Cross, A E., and D. Moore. 1992. Developmental studies on AllagyntS mallgicola

(Hymenoptera: Encyrtidae) a parasitoid ofthe mealybug Rastrococclls illvadellS

(Homoptera: Pseudococcidae). Bull. Entomol. Res. 82: 307-312.

de Jong, P. W., and 1. 1. M. van Alphen. 1989. Host size selection and sex allocation in

Leptomastix dactylopii, a parasitoid ofPlallococcUs cUrio Entomol. expo appl. 50:

161-169.

Funasaki, G. Y., P. Y. Lai, L. M. Nakahara, 1. W. Beardsley, and A K. Ota. 1988. A

review ofbiological control introductions in Hawaii: 1890-1985. Proc. Hawaii.

Entomol. Soc. 28: 105-160.

Holling, C. S. 1959. The components ofpredation as revealed by a study ofsmall­

mammal predation ofthe European pine sawfly. Can. Entomol. 91: 293-320.

Houck, M. A, and R. E. Strauss. 1985. The comparative study offunctional response:

experimental design and statistical interpretation. Can. Ent. 117: 617-629.

Jahn, G. 1992. The ecological significance ofthe big-headed ant in mealybug wilt disease

ofpineapple. Ph.D. Dissertation, University ofHawaii, Honolulu, Hi.

Mani, M., and T. S. Thontadarya. 1988. Biology ofthe grape mealybug parasitoid,

Allagynls dactylopii (How.) (Encyrtidae: Hymenoptera). Entomon. 13: 211-213.

Murdoch, W. W. 1973. The functional response ofPredators. 1. Appl. Ecol. 10: 335­

342.

107 Nechols,1. R., and R. S. Kikuchi. 1985. Host selection ofthe spherir.al mealybug

(Homoptera: Pseudococcidae) by Allagyrus il1dicliS (Hymenoptera: Encyrtidae):

influence ofhost stage on parasitoid oviposition, development, sex ratio, and

survival. Environ. Entomol. 14: 32-37.

Noyes, 1. S., and M. Hayat. 1994. Oriental Mealybug Parasitoids ofthe Anageryrini

(Hymenoptera: Encyrtidae). CAB International, The Natural History Museum,

London. 544 p.

Rogers, D. 1972. Random search and insect population models. 1. Anim. Ecol. 41: 369­

383.

Smith, H. S. 1917. On the life-history and successful introduction into the United States

ofthe Sicilian mealy-bug parasite. 1. Econ. Entomol. 10: 262-268.

Swezey, O. H. 1939. Recent records ofthe introduction ofbeneficial insects into the

Hawaii. islands. Proc. Hawaii. Entomol. Soc. 10: 349-352.

Tingle C. C. D., and 1. W. Copland. 1989. Progeny production and adult longevity ofthe

mealybug parasitoid Allagynispseudococci, Leptomastix dactylopii, and

Leptomastidea abllormis (Hym.: Encyrtidae) in relation to temperature.

Entomophaga 34: 111-120.

Viggiani, G., and A. Maresca. 1977. Ricerche sugli Hymenoptera Chalcidoidea XXXIV.

Dati morphologici e biologici sulla Leptomastidea abnormis (Grit.) (Hym.:

Encyrtidae), importante parassita del Plal1ococclis citri (Risso). Bolletino

Laboratorio di Entomologia Agraria Filippo Silvestri 30: 55-65.

Vinson, S. B. 1976. Host selection by insect parasitoids. Ann. Rev. Entomol. 21: 109­

123.

Weidemann, R. N., and 1. W. Smith, Jr. 1993. Functional response ofthe parasite

Cotesiaj7avipes (Hymenoptera: Braconidae) at low densities ofthe host Diatraea

saccharalis (: Pyralidae). Environ. Entomol. 22: 849-858.

108 Willink, E., and D. Moore. 1988. Aspects ofthe biology ofRastrococcus illvadells

Williams (Hemiptera: Pseudococcidae) a pest offruit crops in West Africa, and

one ofits primary parasitoids, Gyrallllsoidea rebygi Noyes (Hymenoptera:

Encyrtidae). Bull. Ent. Res. 78: 709-715.

Wilkinson, L. 1990. Systat: The System for Statistics. Evanston, IL, Systat, Inc.

Zimmerman, E. C. 1948. Insects ofHawaii. Vol. 5. Homoptera: Stemorhyncha.

University ofHawaii Press. Honolulu, Hi. pp. 189-201.

109 VI. PREY PREFERENCE AND FECUNDITY OF ADULT NEPHUS

BILUCERNARIUS MULSANT (COLEOPTERA: COCCINELLIDAE), A

PREDATOR OF DYSMICOCCUS BREVIPES (COCKERELL)

Introduction

The coccinellid predator Nephlls billicernarius Mulsant was introduced from

Mexico into Hawaii in 1930 and was reported established in 1939 (Swezey 1939, Leeper

1976, Funasaki et al. 1988). However according to Carter (1942), the prospects ofthis predator and 13 other natural enemy species introduced and established in Hawaii in the

1930's were unclear because ofconstant protection ofmealybugs by ants. Nonetheless, in the 50 years after these introductions no attempts were made to assess the biological parameters ofthese natural enemies and their possible impact on mealybug populations.

Jahn (1992) found that after elimination ofthe big-headed ant, Pheidole megacephala

(Fabricius), from experimental pineapple fields, there was an increase in populations of generalist natural enemies (Clirinus coenlleus Mulsant (Coleoptera: Coccinellidae) and undetermined spider species). In a survey ofDysmicoccus natural enemies conducted from 1992 to 1993, N. bilucernarius was found in pineapple fields on the islands ofMaui and Oahu (Chapter III). In some pineapple fields, this predator was found associated with mixed populations ofD. brevipes and D. neobrevipes. In Hawaii, the known prey range ofN. bilucernarills includes the mealybug species D. brevipes, D. neobrevipes, and

Nipaecoccus vas/ator (Maskell) (Funasaki et al. 1988, Chapter III this publication). The importance ofcoccinellids in controlling coccids is supported by 12 cases ofsuccessful

110 biological control introductions, including the most famous example ofRodolia cardina/is

Mulsant introduced from Australia to control the cottony cushion scale, Icerya plirchasi

Maskell (De Bach 1964). Most coccinellid species have a high searching capacity and are

easy to rear (Hodek 1973). These attributes and the sedentary behavior ofthe Coccoidea

makes some coccinellid species good candidates for biological control programs ofsuch

pests. The objectives ofthis study were to obtain basic information about biological

parameters ofN. billicemarius, when reared on D. brevipes, including prey preference,

fecundity and functional response.

Materials and Methods

For all laboratory tests, D. brevipes and D. neobrevipes were obtained from one year old colonies growing on butternut squash under laboratory conditions (26 ± 20 C and photoperiod of 14 hr). These mealybugs were initially obtained from a pineapple planting in Kunia, Oahu Is., Hawaii. First instars ofboth mealybugs species were separately reared on fruit ofbutternut squash. Immature stages ofN. bilucernarius were obtained from mealybug-infested pineapple plants, and subsequent N. bilucemarius adults were reared on D. brevipes colonies. Mealybug colonies were fed on butternut squash (Cucurbita moshata Duchesne), for one year before they were used in this study. Alilaboratory experiments were conducted under temperature conditions of26 ± 20 C and photoperiod of 14 hr. All mealybugs exposed to predator attack were transferred from squash to newly grown Smooth Cayanne pineapple leaves. Leaves were approximately 3 cm wide at the base and 15-20 cm in length and were cut from 2 month old pineapple plants

III maintained under greenhouse conditions. Infestation ofpineapple leaves with mealybugs was done 24 hr before an experiment was started.

Prey Species Preference. Choice! no-choice experiments were conducted to determine which species ofpineapple mealybug was preferred by N. bilucemarius adults.

In the choice experiment, pineapple leaves were individually infested with 5 crawlers each ofD. brevipes and D. neobrevipes and kept in a 5 liter polyethylene container. One day later, mealybugs were exposed for 24 hr to a 2-5 day old adult female N. bilucemarills that was reared on aD. brevipes colony. The number ofconsumed prey was determined by counting the remaining live mealybugs and subtracting the total from the initial prey number. In the no-choice experiment, pineapple leaves were individually infested with 10 mealybug crawlers ofeither D. brevipes or D. neobrevipes. The time ofexposure and the method ofdetermination ofthe number ofconsumed prey were identical to those used for the no-choice experiment. Data were analyzed with ANOVA and when necessary with a

Tukey mean comparison test (Wilkinson 1990).

Prey Stage Preference. Choice/ no-choice experiments were conducted to determine which life stages ofD. brevipes were preferred by N. billicemarius adults. In the choice experiment, fresh-cut pineapple leaves were infested with mealybugs, and exposed 24 hr later to the adult female predator simultaneously in groups ofthree individuals from each ofthe four prey stages. In the no-choice experiment, fresh-cut pineapple leaves were infested with 10 mealybugs ofeither first, second, and third instars and adult females. For this experiment, time ofexposure and assessment ofnumber of prey consumed were determined as in the prey preference experiments. Each treatment

112 was replicated 20 times. Data were analyzed with ANOVA and when necessary with the

Tukey mean comparison test (Wilkinson 1990).

Fecundity ofN. bilucernarius. A laboratory experiment was conducted to determine the number ofeggs N. bilucemarius oviposites when feeding on D. brevipes, the period ofoviposition, and longevity ofall immature stages and adult males and females. Fresh-cut pineapple leaves were infested with 100 mealybug crawlers. Each infested leafwas maintained in a 1.9 liter polyethylene container. The next day, a two-day old adult female predator was introduced to each container. A male was assigned to each female predator and kept in a separate 37 ml plastic vial and was provided a drop ofpure honey as a carbohydrate source. Mating was facilitated by placing both sexes together in a plastic vial for 24 hr before the initial day offemale predator introduction into the mealybug containers. This mating procedure was replicated every two days for one hr and for a period of20 days. Ten replications ofthis study were initiated, but three ofthese replications were dropped from analysis because the female predators never produced eggs. Predators were transferred daily to a new infested leaf, having the same mealybug density, until the predator female died. Numbers ofeggs produced and number of mealybugs consumed daily were recorded.

To determine the developmental times for all N. bilucemarius immature stages, seventeen newly deposited eggs were selected and held in individual 37 ml plastic vials capped with organdy. Three to four D. brevipes third to fourth instars were introduced to each vial and were replaced daily after eclosion ofeggs. Sex ratio ofemerged adults was determined.

113 Results

Prey Species Preference. In the choice experiment, no significant differences among the number ofD. brevipes and D. neobrevipes crawlers consumed by N. bilucemariuswere recorded (F= 1.73; df= 1; P= 0.20). A mean of3.0 and 3.7 crawlers of D. brevipes and D. neobrevipes, respectively, were consumed. Similarly, in the no­ choice experiment, no significant preference was exhibited by N. iJillicemarius for either mealybug species (F = 0.05; df= 1; P = 0.82). A mean of7.3 D. brevipes crawlers were consumed as compared with 7.4 D. neobrevipes crawlers.

Prey Stage Preference. In the choice experiment, N. bilucemarius attacked all D. brevipes developmental stages (Table 6.1). There was greater preference for first (Tukey

HSD; P < 0.001) and second (Tukey HSD; P = 0.023) instars than for the adult stage (F

= 7.35; df= 3; P < 0.001). When only one stage ofD. brevipes was present (no-choice experiment), N. billicemarius adults preferred the smaller D. brevipes stages (Table 6.1).

The predator had greater preference for the first instar than for all other mealybug stages

(F= 29.7; df= 3; P < 0.001). The mean mealybugs consumed by this predator was 4.25 individuals in the first instar, and 0.95 in the second instar. When alone, third instars and adult females were not attacked or consumed by N. bilucernarius adults.

Fecundity, Longevity and Sex Ratio. N. bilucernarius females oviposited a mean total of 100.3 (SD ± 53.04) eggs over a mean oviposition period ofl02 (± 51.03) days. A pre-oviposition period of9 (± 1.3) days occurred following emergence and mating. Following the preoviposition period, some females oviposited up to 6-7 eggs per day, but after 60 days oviposition decreased to 2-3 eggs per day. Mean daily oviposition

114 Table 6.1. Choice and no-choice experiments ofthe predator stage preference ofN bilucemarius when reared on the pink pineapple mealybug D. brevipes.

Mean number ofattacked mealybugs •

Nymph instars

Experiment I II III Adult

Choice 1.45 a 0.95 a 0.85 ab 0.10 b

No-choice 4.25 a 0.95 b 0.00 b 0.00 b

• Means followed by the same letter in a row are not significantly different as determined by Tukey HSD mean comparison test.

115 per predator was 1.01 eggs (± 0.29). During the first 60 days 68 % ofthe total eggs were

produced (Fig 6.1). Females deposited eggs singly or in groups of2 to 3 and distant from

the mealybug infested pineapple leaves (at the top ofthe plastic cup and underneath the

organdy mesh). N. bilucernarills eggs hatched 6.4 (± 0.71) days after oviposition. First,

second, third and fourth instars required a mean developmental time (± SD) of4.0 (±

0.94) , 2.8 (± 0.39),2.4 (± 0.50) and 3.1 (± 0.49) days, respectively, at 260 ± 20 C.

Completion ofthe pupal stage required a mean of 18.4 (± 1.63) days to emerge as an

adult. Thus, development from egg to adult was completed in 33.3 (± 2.78) days. The

sex ratio ofthis predator was 0.8: 1.0 (females to males). Mean adult longevity was 143.1

(± 44.20) days for females and 67.7 (± 32.6) days for males.

Discussion

N. billicemarilis adults preyed on D. brevipes and D. neobrevipes equally well.

On both mealybug species this predator exhibited similar mean attack rates. Results from no-choice experiments with D. brevipes indicated that N. bilucernarius adults preferred to

attack the smallest mealybug stage (first instar). This preference may be related to the

small size ofthe first instars (crawlers) that is more easy to handle. In contrast, when all stages are present, N. billicemarius attacked all mealybug stages, with more preference for the first and second instars than the adults stage. Third instars were not preferred over the younger instars or the adult stage. These findings indicate that N. billicernarilis is able to attack any stage present. This predator attack behavior would be an advantage when prey have discrete generations and only one stage is present. Another advantage is that if

116 3.5,....------, ~ 3t------+------r------! elL. Q) 0 2.5 t---t--t--+--+------+-+------! oeo --~ 2 -g +--t-tlln-It-+-+t-lr------+---;-+------j .oL. EO. :::]:C 1.5 -t--_.- CQ) c::: 1 eo tIJ Q)&' ~·>O.5 o o 1 11 21 31 41 51 61 71 81 91 101 111 121 131 141 151 161 171 IBfs

Fig. 6.1. Mean daily fecundity ofthe coccinellid predator Nephlls bilucernarius (n= 7) reared on DysmicocclIs brevipes.

117 this predator is mass reared and released to control initial field infestations ofmealybugs, it

will attack first instars as well as adult females producing crawlers. Preference for

attacking smaller prey stages has been observed in some coccinellid species. Adults of

Nephus reunioni Chazeau preferred to feed on larvae ofjunior instars and on eggs of

Planococcus cUri Risso and P. ficus Signoret (Izhevsky and Orlinsky 1988). However,

adults ofCryptolaemlls montrollzieri Mulsant attack late third instars and adult

mealybugs, while the larvae feed on eggs and young nymphs (Bartlett 1978, Heidari and

Copland 1992). In the same way, adults ofRodolia cardinalis (Mulsant) prey on adults of

the cottony cushion scale, Icerya purchasi Maskell (Hale 1970).

N. bilucemarills had a fecundity of 100.3 eggs during a mean total oviposition

period of 102 days. This is comparable to that observed in other effective coccinellids.

For example, C. montrouzieri can produce between 100 to 200 eggs (Sweetman 1958).

Cryptognatha nodiceps Mshl., a predator ofthe coconut scale, Aspidiotus destructor

Sign., produces approximately 100 eggs (Sweetman 1958). N. reunioni that is adapted to more temperate regions ofMiddle Asia has an average fecundity of 177.1 eggs with an oviposition period ofabout 140 days (Izhevsky and Orlinsky 1988). However, considerable variability exists in the fecundity and size ofegg clusters ofcoccinellid females (Hodek 1967). In this study, N. bilucemarius oviposited up to 6 to 7 eggs singly or in groups oftwo or three per day. Single egg deposition is characteristic of

Chilocorini, Scymnini and Hyperaspirini (Hagen 1962). Immature development ofthis predator was approximately one month and related to that ofits mealybug prey, between

33 to 34 days (Ito 1938). The developmental period ofN. bilucernarius was similar to

118 Hyperaspis pumila, a predator ofPseudococclis comstocki (Kuwana) (eggs: 6-9 days; larvae: 10-12 days; and pupae: 9-14 days) in Palestine (Rivnay and Perzelan 1943). Egg production for a long period by N. bilucemarius is a good attribute for natural enemies used to maintain multivoltine mealybug populations at low densities.

119 References Cited

Bartlett, B. R. 1978. Pseudococcidae. pp. 137-170. In C. P. Clausen (ed.), Introduced

parasites and predators ofarthropod pest and weeds: a world review. U.S.D.A.

Handbook. No. 480.

Carter, W. 1942. The geographical distribution ofmealybug wilt with notes on some of

other insect pests ofpineapple. 1. Econ. Emomol. 35: 10-15.

Carter, W. 1967. Insects and Related Pests ofPineapples in Hawaii. Pineapple Research

Institute ofHawaii. 105 pp.

De Bach, P. 1964. Biological Control ofInsects Pests and Weeds. Chapman and Hall,

Ltd. London, 844 pp.

Funasaki, G. Y., P. Y. Lai, L. M. Nakahara, 1. W. Beardsley, and A. K. Ota. 1988. A

review ofbiological control introductions in Hawaii: 1890-1985. Proc. Hawaii.

Entomol. Soc. 28: 105-160.

Hagen, K. S. 1962. Biology and ecology ofpredaceous Coccinellidae. Ann. Rev.

Entomol. 7: 289-326.

Hale, L. D. 1970. Biology ofIcerya purchasi an its natural enemies in Hawaii. Proc.

Hawaii. Entornol. Soc. 20: 533-550.

Heidari, M., and M. 1. W. Copland. 1992. Host finding by Cryptolaemus montrouzieri

(Col., Coccinellidae) a predator ofmealybugs (Hom., Pseudococcidae).

Entomophaga 37: 621-625.

Hodek, I. 1967. Bionomics and ecology ofpredaceous Coccinellidae. Ann. Rev.

Entomol. 12: 79-104.

120 Hodek, I. 1973. Biology ofCoccinellidae. Academia, Publishing House ofthe

Czechoslovak Academic ofScience, Prage.

Ito, K. 1938. Studies on the life history ofthe pineapple mealybug, Pseudococcus

brevipes (Ckll.). 1. Econom. Entomol. 31: 291-298.

Izhevsky, S. S., and A. D. Orlinsky. 1988. Life history ofthe imported Scynmus

(Nephus) reullioni (Col.: Coccinellidae) predator ofmealybugs. Entomophaga 33:

101-114.

Jahn, G. C. 1992. The ecological significance ofthe big-headed ant in mealybug wilt

disease ofpineapple. Ph. D. Dissertation, University ofHawaii, Honolulu, ill.

Murdoch, W. W. 1973. The functional response ofPredators. 1. Appl. Ecol. 10: 335­

342.

Leeper, 1. R. 1976. A review ofthe Hawaiian Coccinellidae. Proc. Hawaii. Entomol.

Soc. 22: 279-306.

Rivnay, E., and 1. Perzelan. 1943. Insects associated with Pseudococcus spp.

(Homoptera) in Palestine, with notes on their biology and economic status. J.

Ent. Soc. S. Africa 6: 9-28.

Sasaji, H. 1971. Fauna Japonica. Coccinellidae (Insecta: Coleoptera). Biogeographical

Society ofJapan. Academic Press ofJapan. Tokyo, Japan.

Sweetman, H. L. 1958. The Principles ofBiological Control. Interaction ofhosts and

pests and utilization in regulation ofanimal and plant population. WM. C. Brown

Co. Dubuque, Iowa., 560 pp.

121 Swezey, O. H. 1939. Recent records ofthe introductions ofbeneficial insects into the

Hawaiian islands. Prec. Hawaii. Entomol. Soc. 10: 349-352.

Wilkinson, L. 1990. Systat: The System for Statistics. Evanston, IL, Systat, Inc.

122 VD. CONCLUSIONS

This work included the identification and field evaluation ofintroduced natural

enemies relative to suppression ofDysmicocclis brevipes and D. neobrevipes in Hawaii;

determination ofthe impact ofPheidole megacephala (F.) on the activities ofnatural

enemies ofDysmicoccus spp.; and quantification ofspecific biological parameters ofthe

encyrtid Anagyrlls ananatis Gahan and the coccinellid Nephlls billicernarius Mulsant.

Conclusions and recommendations drawn from these studies follow.

Identification of natural enemies. Dysmicocclls brevipes was the dominant

mealybug in pineapple fields surveyed, and P. megacephala was the most common ant

associated with mealybugs in surveyed plantings. Approximately 20 percent ofthe almost

20 natural enemy species imported and released against Dysmicoccus spp. from 1924 to

1958 established and persist today in Hawaii's pineapple fields. Ofthese natural enemies,

two were primary parasitoids, Anagyrlls allanatis Gahan (Hymenoptera: Encyrtidae), and

Euryrhopailis propillqulIs Kerrich (Hymenoptera: Encyrtidae). The remaining were

predators, Lobodiplosispselldococci (Felt.) (Diptera: Cecidmomyiidae), Nephlls

bilucemarius Mulsant and Stich%tis ruficeps Weise (Coleoptera: Coccinellidae).

Additional entomophagous species found included the hyperparasitoid, Prochiloneunls sp.

(Hymenoptera: Encyrtidae), associated only with the encyrtid parasitoid E. propinqlllls,

and the parasitoid Homalotylus sp. (Hymenoptera: Encyrtidae) found attacking larvae of

both coccinellid predators, N. billicernarius and S. nificeps. All natural enemies found

during the survey attacked DysmicocclIs spp. tended by ants.

A. Gllallatis was the most common natural enemy found associated with mealybugs

in pineapple. However it attacked only D. brevipes, the dominant mealybug in the pineapple fields surveyed. In this survey, parasitization was observed on D. brevipes tended by ants. Therefore, good potential exists for use ofA. allallatis in augmentative programs against D. brevipes in Hawaii. 123 E. propinquus was present in almost all localities, but in low numbers compared to

A. allanatis. Nevertheless, this parasitoid may be a good candidate for use in biological

control programs because it attacks both D. brevipes and D. neobrevipes . Although,

capable ofparasitizing both mealybug species, it appeared to prefer D. neobrevipes as a

host when mixed mealybug populations were present in pineapple plantings. Ofadditional

concern with this wasp is its susceptibility to the encyrtid hyperparasitoid Prochiloneunls

sp.

The only predators that appeared to have potential in Dysmicoccus biological

control were N. bilucernarius and L. pseudococci. Both predators attack D. brevipes and

D. neobrevipes, and both were present in almost all surveyed areas. N. bilucernarius is

more prey specific, and, therefore, it is a possible candidate for natural enemy

augmentative programs for Dysmicoccus control. Additionally, the immature and adult

stages feed on all the mealybug stages.

Additional attempts should be made to discover and introduce natural enemies of

D. brevipes and D. neobrevipes from South America. Most natural enemies found in this survey, except the coccinellid S. nificeps, were associated with mealybugs feeding on the aerial parts ofpineapple plants. This suggests that these natural enemies were originally collected in other countries attacking mealybugs present on the aerial plant parts. No previous reports exist on natural enemies attacking Dysmicoccus spp. mealybugs feeding on the pineapple roots.

Considering the current distribution ofthe natural enemies found in this study and their feeding behaviors, A. ananatis, E. propinquus, 1. pseudococci, and N. bilucernarius, should be considered for use in augmentative biological programs against Dysmicoccus spp. in pineapple.

Natural enemy effectiveness in the presence ofants. Two methods were used to evaluate the effectiveness ofnatural enemies in maintaining low densities ofD. brevipes

124 on pineapple. These were the biological check method (interference) and the paired cage

technique (exclusion). Elimination ofants during the biological check method showed

that D. brevipes had higher densities when ants were present, and natural enemies could

not access mealybugs as readily. Dysmicoccus natural enemies were less commonly found

when ants were present. Natural enemies, particularly A. ananatis, contributed to the

decline ofmealybug densities following elimination ofants via ant baits.

A study that combined the biological check method and the paired cage technique

revealed similar results, but produced some unexpected observations also. As expected in the absence ofants, higher mealybug densities were observed in closed cages where natural enemies were excluded from pineapple plants as compared to plants that natural enemies were allowed access. This suggests that natural enemies were impacting the mealybug population. However, in the presence ofants, higher densities ofmealybugs were recorded in the open cages as compared with the closed cages. This was an unexpected result for which no explanation was obvious, but several were hypothesized.

Furthermore, densities ofmealybugs in the presence ofants (closed and open cages) were higher than in the absence ofants. This suggests that some factor other than protection of mealybugs by ants was operating to promote mealybug reproduction / survival. A reasonable explanation is that the ants removed excess honeydew (sanitation) from the mealybugs' habitat thereby reducing mealybug mortality associated with honeydew entanglement. Further studies should be conducted to determine ifthis is a suitable explanation. Additionally, comparison ofresults obtained in the biological check method and the combined biological check method and paired cage technique suggests that the former technique is not adequate to evaluate natural enemy efficacy ifincreases in mealybug densities are partially associated with sanitation provided by the ants.

In laboratory experiments, P. megacephala reduced mealybug mortality induced by A. ananatis and N. bilucemarius adults via interference with natural enemy searching

125 behavior. Additionally, ants killed the immature stages ofN. bi/ucernarills IfA. al1anatis and N. bi/licernarius have adapted to avoid ant interference, they have not successfully modified their oviposition or predation behaviors, respectively, to circumvent ant aggression. However, this does not mean that they are not efficient natural enemies in the absence ofants.

Biological parameters ofAnagyrus ananatis. Immature males completed their development (eggs to adult) in 21.2 days, while immature females required 23.3 days at

26°C. A. ananatis preferred to attack adult females ofD. brevipes over D. neobrevipes females. A. ananatis appeared to be host specific in its attacks and has only been reared in

Hawaii from D. brevipes. A. anal1atis preferred adult female D. brevipes to parasitize when given a choice oflife stages. Effective timing ofA. ananatis releases should be based on the presence ofearly infestations ofthird instar and adult females. Over its adult life, A. ananatis parasitized an average of27.7 mealybugs, with a range of 11-51 parasitized mealybugs. At 26°C, A. ananatis females had a mean longevity of9.8 days (±

1.97), while the males lived for a mean of 10.8 days (± 2.80). A. ananatis parasitized most ofthe hosts it was biologically able to parasitize during the first few days after adult emergence.

The curvilinear response obtained by the nonlinear regression equation ofRogers

(1972) did not adequately describe the observed rates ofhost attack by A. ananatis on D. brevipes. A handling time of5.1 (± 1.9) seconds was observed at host densities of 10, 25,

50 and 100 mealybugs per unit, and was about 1800-fold shorter than the handling time estimated by Rogers (1972) technique.

Biological parameters ofNepltus hilucernarius. N. bi/licemarius adults preyed on D. brevipes and D. neobrevipes crawlers equally well. For both mealybug species this predator exhibited similar mean attack rates. In choice experiments with D. brevipes, N. bi/ucernarius adults preferred to attack the smallest stages present (crawlers), but would

126 feed on all stages. This behavior would be an advantage when prey have discrete

generations and only one stage is present. Ifthis predator was mass reared and released to

control initial field infestations ofmealybugs, another advantage would be that it will

attack first instars as well as adult females producing crawlers. Development from egg to

adult was completed in 33.3 (± 2.78) days at 26°C. The sex ratio ofthis predator was

0.8: 1.0 (females to males). Mean adult longevity was 143.1 (± 44.20) days for females

and 67.7 (± 32.6) days for males at 26°C. N. bilucernarills had a mean fecundity of

100.3 eggs during a mean total oviposition period of 102 days.

Recommendations for an augmentation program on pineapple. Although

several species ofDysmicocclIs natural enemies are established in Hawaii, augmentation

programs for all species are not practical due to the expense and logistics required to

develop, implement, and maintain the programs. Thus, one or two species should be

selected initially for consideration. Relative to the natural enemies ofDysmicocclis

brevipes and D. neobrevipes, more information is available on the parasitoid A. ananatis

and the predator N. bilucernarius. Each species has strengths and weaknesses relative to use in augmentation. Releases would be made in environments where ants were present.

The benefits ofusing A. ananatis are numerous. The parasitoid has a relatively short developmental cycle (ca. 23 days) and only requires one mealybug for development.

Females lay most oftheir eggs within approximately six days after emergence. Thus, females would not have to be maintained for long periods to obtain their optimum reproductive capacity. In the field, ants only appear to interfere with the searching behavior ofthe adult parasitoid but do not cause heavy mortality to the immatures contained within mummified mealybugs or the adults (unpublished observation). Ant interference reduces A. ananatis parasitization rates approximately 48%, but does not eliminate parasitoids from the field. Thus, inoculative augmentative releases would potentially have some impact on mealybug populations as a result ofpost-release

127 parasitoid generations. Additionally, no hyperparasitoids have been reported in association with A. anana/is. A major weakness with the use ofA. anana/is in Hawaii is its preference for D. brevipes as a host. However, this may not be a significant problem in plantings where D. brevipes is in high densities and D. neobrevipes is rare. Timing of releases ofA. anana/is may be ofcritical importance given its preference for third instar and adult mealybugs.

The predator N. bilucemarills has an advantage in that it feeds equally well on both Dysmicocclis spp. in pineapple. It can feed on all stages ofthe mealybugs, and thus can potentially impact Dysmicocclis populations no matter what age distribution ofthe mealybugs is present when releases are made. Mass rearing ofthe predator would undoubtedly require more time, space, and numbers ofDysmicoccus than that for A. anana/is given the predator's longer developmental cycle, its reproductive behavior (9 day preoviposition period), and the numbers ofmealybugs required to complete development.

Although N. bilucernarills can lay more eggs than the parasitoid over its 100 day or more adult life, A. ananantis females can complete their entire ovipositional period before female N. billlcemarills initiate oviposition. Relative to inundative releases, N. billlcemarius predation rates may be reduced as much as 70% in the presence ofants.

Although a candidate for inoculative releases, generations subsequent to the release generation would probably be small because ofant predation on eggs and immatures.

Moreover, low daily ovipositional rates would also hamper the impact ofthe predator on the mealybugs.

Based on current information, A. anana/is is probably the more optimal species with which to initiate augmentative trials. However, further research should be directed towards introduction ofnew biological control agents that can maintain low Dysmicoccus densities in the presence ofants. Additionally, other natural enemies should be studied further to evaluate their potential in augmentative programs.

128