Nutritional and chemical ecology of select noctuid caterpillars with emphasis on the velvetbean caterpillar

MISSING IMAGE

Material Information

Title:
Nutritional and chemical ecology of select noctuid caterpillars with emphasis on the velvetbean caterpillar
Physical Description:
xi, 212 leaves : ill. ; 29 cm.
Language:
English
Creator:
Wheeler, Gregory S., 1955-
Publication Date:

Subjects

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

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1989.
Bibliography:
Includes bibliographical references (leaves 184-211).
Statement of Responsibility:
by Gregory S. Wheeler.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001563076
notis - AHH6802
oclc - 22712369
System ID:
AA00003778:00001

Full Text











NUTRITIONAL AND CHEMICAL ECOLOGY OF SELECT
NOCTUID CATERPILLARS WITH EMPHASIS ON
THE VELVETBEAN CATERPILLAR



















By

GREGORY S. WHEELER


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

UNIVERSITY OF FLORIDA


1989

































Copyright 1989

by

Gregory S. Wheeler


































This dissertation is dedicated to my grandparents,

Sherman C. Wheeler and Ilma Hilton Wheeler for their love and

wisdom.














ACKNOWLEDGEMENTS


I thank my Ph.D. committee, Drs. S. J. Yu, D. C.

Herzog, J. E. Funderburk and K. V. Rao for their support and

guidance through my doctoral program. I am indebted to my

major advisor, Dr. Frank Slansky, Jr., for his dedication to

detail, his insight, his tolerance and patience in

discussing the science of nutritional ecology while still

arousing enthusiasm, and finally for teaching me the art of

describing and synthesizing our results in print. I thank

A. L. Brown, J. Neller, R. L. Wilcox, and A. R. Zimet for

assistance in the initial stages of this research; J. W.

Beach and C. P. Rock for assistance in flavonoid extraction

and analysis; D. H. Habeck and J. B. Heppner for assistance

with Anticarsia systematics; D. B. Ward and D. W. Hall for

assistance with legume systematics; N. E. Cohen, B. M.

Gregory, Jr., D. C. Herzog, M. J. Plagens, S. J. Simpson, W.

A. Timmins and D. J. Waters for sharing with me their

unpublished data; the Agronomy nutrition lab for nitrogen

analysis; S. H. Kerr and J. R. Strayer for the graduate

assistantship; R. E. Lynch, M. M. Martin, S. S. Quisenberry

and S. J. Simpson for critical reviews of Chapter 4.

Finally, I thank Karen Thel for her support and patience

during the high and low phases of my doctoral program.















TABLE OF CONTENTS


ACKNOWLEDGEMENTS . . .. iv

ABSTRACT . . ix

CHAPTER 1
GENERAL INTRODUCTION ..... ... .. 1
Nutritional Ecology of Insect Herbivores 1
Dissertation Overview . 3

CHAPTER 2
CHEMICAL ECOLOGY OF ANTICARSIA GEMMATALIS 4
Coevolution of Herbivores and Host Plants 4
The Anticarsia Species .. ... 6
General Systematics of the Leguminoseae 7
Host Utilizaton by A. gemmatalis . 7
Caesalpinioideae and Mimosoideae .. 20
Papilionoideae . . .. 20
Association of Herbivore and Host Species 20
New World Tropical Tribes .. .. 21
Old World Tropical Tribes . .. 22
Temperate Tribes . .. 23
Systematic Implications of Susceptibility to
Other Legume Pests . .. 24
Legume Chemosystematics . 26
Ecological Function of Plant Metabolites .. 27
Chemical Defenses of the Leguminosae .. 28
Alkaloids . . .. 29
Pyrrolizidine Alkaloids . .. 41
Quinolizidine Alkaloids . .. 42
Erythrina Alkaloids . .. 44
Miscellaneous Ungrouped or Protoalkaloids 44
Non-protein Amino Acids . .. 45
Flavonoids . .. 47
Flavone and Flavonols . .. 48
Isoflavones . .. 52
Anthocyanins . . 53
Terpenoids . . 57
Cyanogenic Species . .. 60
Proteinase Inhibitors . .. 61
Miscellaneous Defenses . .. 63
Lectins . . 63
Simple Phenolics . .. 63








Nutritional Limitations . .. 64
Physical Mechanisms of Defense .. 65
Conclusions .. . . 66

CHAPTER 3
TOXICITY OF NON-INDUCED AND HERBIVORE-INDUCED
EXTRACTABLES FROM SUSCEPTIBLE AND RESISTANT SOYBEAN
FOLIAGE TO NON-ADAPTED AND ADAPTED NOCTUID HERBIVORES 67
Introduction . ... 67
Methods and Materials .. .. .. .. 68
Experiment 1. Induction of Susceptible and
Resistant Soybean . ... 68
Experiment 2a. Preliminary Extract Methods 70
Experiment 2b. Refined Soybean Foliar
Extraction . ... .74
Experiments 3a-3e. Influence of Greenhouse
Induction of Soybean Defenses on
Herbivore Performance . 78
Experiment 3a. Influence of Greenhouse
Induction of Soybean Defense on
Velvetbean Caterpillar Mortality .. 81
Experiment 3b. Effect of Induced Resistance
on Non-adapted and Adapted Soybean
Herbivores . ... .82
Experiment 3c. Sensitivity of Three Noctuid
Species to the Non-benzene Soybean
Foliar Extract Fractions 82
Experiment 3d. Effect of the Petroleum Ether
Extract Fraction on RGR of Non-adapted
and Adapted Soybean Herbivores .. 83
Experiment 3e. Purification of the Active
Fractions . 83
Results . ... .. 84
Experiment 1. Induction of Susceptible and
Resistant Soybean . .. 84
Experiment 2a. Preliminary Extract Methods 85
Experiment 2b. Refined Soybean Foliar
Extraction . ... .86
Experiment 3a. Influence of Greenhouse
Induction of Soybean Defense on
Velvetbean Caterpillar Mortality .. 87
Experiment 3b. Effect of Induced Resistance
on Non-adapted and Adapted Soybean
Herbivores . ... .90
Experiment 3c. Sensitivity of Three Noctuid
Species to the Non-benzene Soybean
Foliar Extract Fractions ... .. 90
Experiment 3d. Effect of the Petroleum Ether
Extract Fraction on RGR of Non-adapted
and Adapted Soybean Herbivores 92
Experiment 3e. Purification of the Active
Fraction . . 95









Discussion. . .
Species Sensitivity to Constitutive Soybean
Allelochemicals . .
Conclusions . . .

CHAPTER 4
COMPENSATORY RESPONSES OF THE FALL ARMYWORM (SPODOPTERA
FRUGIPERDA) WHEN FED WATER- AND CELLULOSE-DILUTED DIETS
Introduction . . .


Materials and Methods . .
Diets . .
Insects . .
Analyses . .
Results . . .
Diets .. . .
Larval Mortality and Developmental Time
Larval Composition . .
Absolute Consumption . .
Relative Consumption . .
Growth and Development .
Feeding Efficiencies . .
Discussion . . .
Water Regulation . .
Regulation of Consumption and Digestion
Energy Cost of Consumption .


* 122
* 122
* 123
* 124
* 125
* 125
* 126
* 127
* 130
* 133
* 139
* 142
* 147
* 153
* 155
* 158


CHAPTER 5
COMPENSATORY DILUTION-INDUCED FEEDING INCREASES BY
ANTICARSIA GEMMATALIS LARVAE AND ITS COST/BENEFIT
WHEN FED SOYBEAN ALLELOCHEMICALS . .
Compensatory Feeding Responses in Larval
Herbivores . . .
Methods and Materials . .
Preliminary Study . .
Determination of Compensatory Ability Fed
Diluted Diet Containing Sublethal Levels
of Soybean Foliar Extractables .
Results . . .
Preliminary Study . .
Determination of Compensatory Ability Fed
Diluted Diet Containing Sublethal Levels
of Soybean Foliar Extractables .
Larval Consumption . .
Growth and Development . .
Discussion . . .

CHAPTER 6
GENERAL CONCLUSIONS . . .

APPENDIX A . . .

APPENDIX B . . .


vii


100

106
118


120


160

160
161
161


163
164
164


165
166
170
173


177

181

182










APPENDIX C . . 183

REFERENCE LIST . . .. .184

BIOGRAPHICAL SKETCH. . . 212


















































viii














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

NUTRITIONAL AND CHEMICAL ECOLOGY OF
SELECT NOCTUID CATERPILLARS WITH
EMPHASIS ON THE VELVETBEAN CATERPILLAR

By

Gregory S. Wheeler

December 1989


Chairperson: Dr. F. Slansky, Jr.
Major Department: Entomology and Nematology

Soybean foliage from susceptible and resistant lines

was extracted in a number of organic solvents to assess

constitutive (undamaged) and induced (damaged) effects.

When incorporated in artificial diet, the benzene fraction

contained the constitutive activity of both the resistant

and susceptible lines, as indicated by reduced relative

growth rate (RGR) of several noctuid soybean-adapted and

non-adapted herbivore species. Induced resistance, detected

only in the petroleum ether fraction, reduced RGR of the

fall armyworm.

Caterpillars of the fall armyworm reared on artificial

diets diluted with cellulose and water, increased fresh

weight (fw) consumption 2.5-fold over those on undiluted

diet. At the moderate levels of water- or cellulose-








dilution, this increased consumption, combined with

increased digestion and absorption of nutrients (ADNU),

sufficiently compensated for the reduced nutrient intake to

achieve pupal biomass equivalent to that on the undiluted

diet. At higher levels of water- and cellulose-dilution, fw

consumption and ADNU increased further but pupal dry weight

declined on the water-diluted diets. At each level of

dilution, fw consumption and ADNU increased similarly on the

water- and cellulose-diluted diets, but weight gain was

reduced on the water- compared with the cellulose-diluted

diets. This was due in part to lowered food conversion

efficiency on the water-diluted diets, possibly caused by

increased costs of metabolizing the wetter diets. My data

support the hypothesis that consumption rates are regulated

by nutrient feedback and possibly further modified by

volumetric feedback mechanisms. The cost of increased

consumption rates on diets of reduced energetic value may

constitute a more significant energy expenditure than

previously believed.

The adaptive significance of this compensatory feeding

was examined in terms of avoiding ingestion of a lethal dose

of allelochemical. Velvetbean caterpillars were fed

sublethal doses of the benzene fraction in water-diluted

diet. The increased feeding that occurred on extract-

containing, diluted diets, was less than that on the

extract-free diets. This difference was probably caused by








the toxic impact of the ingested compounds, possibly

interacting with dietary water, rather than being an

adaptive response.














CHAPTER 1
GENERAL INTRODUCTION


Nutritional Ecology of Insect Herbivores

The paradigm of nutritional ecology (Slansky and

Rodriguez 1987) implies that an insect has a genetically

programmed suite of performance values (e.g., body size,

developmental rate, fecundity) that is attained under ideal

environmental conditions. The environment, however, is

rarely ideal; consequently, natural environments place

constraints on the consumption, utilization and allocation

of food, altering performance values accordingly. Insect

herbivores have evolved the ability to perceive the

conditions of the natural environment that alter their

performance and to respond in an adaptive manner. These

responses usually involve modifications in the consumption,

utilization and allocation of food, along with the

associated behavioral and physiological changes. These

changes may involve inductive (e.g., diapause in response to

decreasing day length) or compensatory (e.g., increased

consumption on a dilute diet) responses that mitigate the

negative impact of the environment. Thus, there are trade-

offs between the costs and benefits of these adaptive

responses; the costs generally include reduced performance








2

and the associated decrease in fitness, whereas the benefits

of the response may be assessed in terms of survival or

production of viable offspring. The combination of the

nature and degree of the environmental constraints, the

species' preprogrammed performance values and their ability

to perceive and respond adaptively to environmental

constraints select for a diversity of herbivore lifestyles.

Knowledge of the factors influencing the rate of food

consumption and its efficiency of utilization may provide a

better understanding of herbivores' ability to respond to

environmental constraints. Non-nutritional dietary

components (allelochemicals) that alter non-adapted

herbivore performance or behavior (e.g., toxins, repellents)

may reduce growth either directly by inhibiting primary

metabolic pathways or indirectly by reducing consumption and

the utilization of the ingested food. Allelochemicals to

which an insect is adapted may be detoxified (Ahmad et al.

1986), tolerated (Berenbaum 1986), used as a host finding

cue (Schoonhoven 1972), sequestered for protection from

natural enemies (Duffey 1980), or may serve as sources of

nutrients (Rosenthal et al. 1977). Reduced nutrient intake

through either nutrient complexes with allelochemicals

(Rhoades and Cates 1976) or incomplete or unbalanced

composition may similarly alter herbivore feeding behavior

(Waldbauer and Friedman 1988).








3

Dissertation Overview

This dissertation is divided into chapters designated

for individual publications. Chapter 2 constitutes a review

and synthesis of the pertinent literature regarding legume

phylogenetics, allelochemistry and the plant species

utilization by the velvetbean caterpillar Anticarsia

gemmatalis Hubner. The third chapter includes experimental

laboratory results of the extraction of constitutive

(existent) and inducible soybean foliar allelochemicals and

their influence on unadapted and adapted soybean herbivores.

Chapter 4 describes further laboratory studies dealing with

the compensatory abilities of the fall armyworm Spodoptera

frugiperda (J. E. Smith) to water- and cellulose-diluted

diets. Although incorporating distinct species of

herbivores into a single dissertation may seem diffuse, each

chapter fits into the overall body of research generated and

in progress in the nutritional ecology research group.














CHAPTER 2
CHEMICAL ECOLOGY OF ANTICARSIA GEMMATALIS


Coevolution of Herbivores and Host Plants

Traditionally, secondary plant compounds were

considered waste products of primary metabolism and of no

benefit to the plant (Harborne 1982). This paradigm was

questioned by Fraenkel (1959), who suggested that secondary

metabolites provide plant defense against herbivores and

that they are responsible for the patterns of herbivore-host

selection. Ehrlich and Raven (1964) proposed a stepwise

coevolution of plants and butterflies in which secondary

plant metabolites select for a diverse pattern of host-plant

utilization. These metabolites, or allelochemicals, may

reduce the palatability of a plant or be otherwise

deleterious to herbivore feeding. In response to these

forces, the herbivores have evolved methods of countering

the chemical defenses through a variety of means, among them

avoidance of the toxic plant tissue (Mullin 1986),

detoxifying the compounds (Ahmad et al. 1986) or developing

tissue insensitivity (Berenbaum 1986). Some well-adapted

herbivore species may use allelochemicals as cues for host

location (Schoonhoven 1972), whereas others may sequester

the compounds for their own protection from natural enemies








5

(Duffey 1980). Other well adapted herbivores may use

allelochemicals as nutrients that are otherwise toxic or

deterrent to non-adapted species (Rosenthal et al. 1977,

Bernays and Woodhead 1982). In response to the breached

plant defenses, the plants diversified the defenses,

producing a diverse array of compounds with a variety of

effects on herbivores.

This chapter includes a compilation of information

regarding the velvetbean caterpillar Anticarsia gemmatalis

Hubner larval host range and the associated plant

allelochemistry. The purpose of this synthesis is to

attempt to determine if A. gemmatalis larval host plant

range exhibits any pattern with regard to legume systematics

or allelochemistry. Included here is a compilation of data

on larvae fed plant species in laboratories, reports of

field records of larvae and a list of the natural products

identified in potential host and non-host species. This

information may assist researchers (e.g., plant breeders)

interested in identifying possible chemical factors that may

impart A. gemmatalis resistance in legume agronomic or

horticultural crops. Thus, I will first propose a scheme of

host utilization through the tribes of the Leguminosae, then

review the literature describing the presence/absence of the

classes of natural products reported from the A. gemmatalis

potential hosts or non-hosts and finally relate these

compounds to the potential larval host range. Other








6

pertinent reviews appear elsewhere: A. gemmatalis general

life history and pest population management (Herzog et al.

unpubl.); the physiological ecology of A. gemmatalis

(Hammond and Fescemyer 1987) and the chemical mechanisms of

soybean Glycine max (L.) Merrill resistance to herbivores

(Smith 1985).


The Anticarsia Species

The velvetbean caterpillar (A. gemmatalis) is

distributed throughout much of the subtropical and tropical

new world and seasonally, moths migrate into temperate areas

(Herzog and Todd 1980). Within this range the only area

where A. gemmatalis has not been reported includes the

Amazon basin. In addition to A. gemmatalis, 12 other

Anticarsia species exist worldwide. These have been

reported from Venezuela (A. acutilinea Walker), Ghana (A.

albilineata Hampson), Honduras (A. anisospila Walker), Peru

(A. disticha Hampson), Australia (A. distorta Hampson), Cuba

(A. elegantula Herrich-Schaffer), India (A. irrorata

Fabricius), Brazil (A. parana Guende), China (A. renipunctum

Berio, A. tiqris Berio), Mexico (Veracruz) (A. suffervens

Dyar), Cameroon (A. unilineata Gaede) (Poole 1989). The

species A. repuqnalis Hubner reported from south Florida

(Tietz 1972) was changed to Azeta repuqnalis (Hubner) (Poole

1989).








7

Apart from A. gemmatalis, considered a legume

specialist (see below), information on host-plant

utilization by these other species was recorded only for A.

irrorata, with larvae were collected on the following legume

species: Phaseolus mango; Doclichos (sic) lablab (Poole

1989); Arachis hvpogaea; Medicago sativa; Canavalia

bonariensis and Wisteria sinensis (Biezanko et al. 1957).

Although the host range data regarding the entire genus are

incomplete, they suggest that other Anticarsia species may

also be legume specialists.


General Systematics of the Lequminoseae

The Leguminosae comprises 650 genera and 18,000 species

and constitutes the third largest family of flowering plants

after the Compositae and Orchidaceae (Polhill et al. 1981).

The family is divided into three, more or less distinct

subfamilies, the Caesalpinioideae, Mimosoideae and

Papilionoideae. The general divergence pattern of the

tribes within the Phaseoloideae is presented in Fig. 2-1.

The tribes of this subfamily originated from the ancestral

Swartzieae and diverged, producing the Sophoreae and

subsequently many radiating groups (Polhill 1981).


Host Utilizaton by A. gemmatalis

The data used in this chapter were compiled from a

variety of published and unpublished sources (Table 2-1).

Earlier reviews covered primarily field host records (Herzog








Fig. 2-1. Phylogenetic representation of the divergence of
the tribes of the Papilionoideae (from Polhill 1981).
Large numbers represent the predominant chromosome base
number.































t O 0
I


d
I


fA r


















r-4 =
04 H
r- -r r-
UmI u

CO1 3Q |
3rl C 41 C::l


m C'-to

U) < O u ) 1 Q)t U
OH I O Or-
U (^OQ fUA f


<1 *






0




4-)
"0





(0


>i>
MP
0o,-4


OH '-
O-I
. 4-








>4 t
04
0 H



0



a) ,1:









O
00
0


*0 >
r-I d











Nc
o




S0
0 ,-I
0
U
i-l (
H- (


cU
()
0)
4.4
0)






CO
O)






XI
i)





















'4
*4


Ea






























44
-1
co








X)
















U








U)
o,-


0)



-P
-> -C
S



5X N




(a 41 o


nH M a)
U -4 (
U U


. ) O-H l
1- 1 -, 1 -1
-H to .1 Ul i





0


t. H
(0





t
o I1 4 10
C *q-i
U) to i.i -4-I
0 c '- to



1 E 0
-) 4-1
C -H -H
1 00
C0 a l
I0 -H P (C

4- t0 II
1 1
c5-

U)


co.








Hou


SHH


U)
0
-I
*H


WO
4)
-H


) U
U
00


S00+10 00 O00
(M


( i0
SaH


o0 ao

0 0


I 0
0>

0 >
oa-
o p








O


a-i
0 +






O)







e0
O--
C>




Q) U
(H I) 0

a)






u C







toe
~0J4




t -H






10 En
(- 4





04-Ju
pO
>4 0)
-g Q- '












-4-




gk
0 4j a
4 )










a (
0 3












40 U(






c :30
Nd ^ *



















f in ^ x
3o S^ ^ 0
(M


*




I,1
(0
4 >
(0


r-1
U)
H
U


1 0
(0U
o M
t o
c a)
-Hl (U

(U*6
p.4(0


M
c i
*H -H
14 r-1
10 *
C
.U.

i4 (D En


*H o
r-I 0 '0
(4 (0



00-1
o iI n
0 10
0





















0: uo OONOO 44
03
rC



SC 9




a aM a 0 a




C MC CrCOCC M






o co 0 00 90
) S< CAU)
H H4 HHHH '4

o~ oN
H( H


cO

H

C>


(n
r'



z


0 0
H H
c a


3
S4 N4






0)4

o o
















0)
1) U
o U)
O a

C H


U) 0)
0) 0)
-4 .4-4
a,c
C C

(0 m
a aX
U (0


*
ti





(0
U


H< U

H -4

) 0 -
0 0 -H .0


i) C)0 r-
0 4 4
-4
E


0 *
O P4q
3 0
0) 0
*04
a'u
4-i
-H O
0) '-ft CO)
1 0 00 pp
cU P U 0
H U-H-HH ** ) 'U


0 r- 4-HJ H H



04
ca


'4



H
-H








0) ()
C (n


a a >
to 0 U)

a ) a-
H- -a


4l- i- l

-'00 0 H
(40 l H
4-i'0UV *HH4-
1.0 0 0 'U 4-*4

rU) 'U C ( 'U

0)
(0


a) U} )

00 EH


U)U) 0
OO-H o
,C ,C *H


4)
E)
*U
0
O

a

















Hr

0I

a)


C
a)
tr
r<
H


C
1- 0o "1 -Hi r-

0a a 0 a a aH
H---HHH H a



xa) a ax a) 1




0 .0 0 )
O cr-4-4 Or- 0 0 0 0 -p 3
UU~lftlUP<(UUSB


o oooggg00~Ns i





C
*HH
S000
0 0 0 r- LO










a0
0


a)
0)
a)











4-
.a)
U)




a)























4j
/3




















-,-4
CO
V-
E-4













a)




-4
C
U


C



M* -
aH a) H






4 En a
i l 3 L ) 0) 3
X (04i a) (
0 (0 C -
U PQ 34P
O3 H ^
DABA M 0


H an Co



3 0 3 H -



04 U 4 4


V uK O 0 ooo
Va z 0
~ cc H


a) a) a)

-14 -14 *4
N N C

a) a)

a) a -
,- )Q-: r


a) ) ) a) Q) a) )
0 (0 A0 0 (0 a0 (0
.(0 40 .4 .O -4 4
0 0 0 0 0 0 0)
d In
u) In In n n In 10
0) 0) r 0 ) () q0 .)
Q Q Q Q Q 0 P0








*



a X -) U o



i i > 4 !I 11I4i l qI
a 4 9 a .

C 1f 3 U


0 1 009I 0I m.

S1 1 0 l
0 0 0


r 0 0
a) a)


a) a) a) a)


in In 5o Fa
) a) C C











r-l
4-) 4-)
S -(
S 3

M n


o0 4 '11 -HiS
(0 0) M(IS (I ) r- I a)
*H -H 0 4-1 0 4-1
C *-IC H 0 r0



U) H HU


- -
















H 0 -P r. O 90 4




0 4) w IA N 0 Mr) 0 %D
I p



0 1 0 0 00 O
r0 oAN ON OD
** Am r-









S H H H HH CO 0
-( i- rl i- r-l C (7> H









3 4 @@| ^ | A Xt 5wm |
H +- 0 m+ H o o




N~ cc o\oo(


1)
0

4-4
0)
cc
0










-u

-)


10
CC

.H --H
00
-H Hr
o3Ci


0 0
C C
--I -*-
C C
H- -H
o 0
,.-4
COC


O O OO O )
0 0 0 0
g (d M (

0 0 00 0
4) Q) 4) () Q)
m 4 44 4
0 0 0 0
CA CA CAd A C








O 4* 3 .
r-H O ) U 0 ,
"H 0 ,Q C


- C 0 E -


p *0 r-4 ) s ( *
--I (a H >
00C () H
0 0 3 H (U-H H
po o
1) 0u ,--1 1 -1

4 n > o0 >l (1 MII>
3 It 0 -i (O pU (0
u 1 ^ &c 0 1


0 00 0 0
0 00 0 0
O OO O O
rl rlrl r rl
0 00 0 0
a( Q)e) Q) (a


04 C 4 4 a


4J r0



U Z OM p 0 a 0)

H 1 3 r-I X
S* '0 000 0 -
*- r-I 43] a ).
r-H Ha
* WH CA -4m 4 t

(0rd 0 r- (- ( ( 0 (0 4

S-1 r-I -HH H 3 4J (0" -1 I "1)
CO34J4-H-H+ o1Q-Ho1 1 OC
0 4.-)-1 ) 4-) 4-) :l $4 (d 4 -)40
- u ( U ) u (o )-i (0 0 -i
0 (-1 (0 4 ( 0 P 0 3 C P
p q2r-q r- > ) r-l ( >
U UU U & '-' &<


00


00
H H
o I
0 0
*-H -H
QQ Q














ON



r A

0 -H \ 0
P aP co 04

-5 -5
0 1ONM
0 co N
H- r' C &

r- rl 0 H-1 rl


CS --
CNN (N s
CO N CO
4 r- N-
co r, co
0) 0) 0)


4->|la '0


0 q
1 -H
C C
-0 00



0.04


* H
)|I' '


MI n
S )l N
2C 04 -

*r 0 w ( CO
0 4) 0I w-4 r
3 0 k 3 -
BMUC -0U -


4-)

HH- 4

a a

v 3


La M

-4 '-'
rc
-H-4


m0

10 r-1
r-1 -Hl
( Z


-4 O' iO
*H W > CO

0 s ^lf)
0 r-
to r-I 0O
4 (z l-0 --i
HW .i ,C
&a4

4 P'0 *
0 00

0) Ir 104)'P
4) t *'-I 0 4I 1 r
10) 00
5 AM U n


ta c Co co


aNm N


0 0




4 4
CC
*Hrl *Hr

00
w a)





0)0)


00

(0(0(
ft ft


Oi
In
H
0
In
in
(01
43
II


S-

J N M -

3 0 0 qI
En 0 3 O




-4
0 W o




In 0-' -I
U il 1 1
aJ 0 1 l 0 00 p



SP 0 (a l P P 4-- 4 3
3 -4 -11 1 0 I E- 01

r. H > I 01I ll C 1 4 II
a) c0 H -4 t7 0
0 0Q > >
(I 0 I *>I
U Q >|


*


C
q







SI

4-J
0
r-1 M





3 I


. 1-

0

4J

'0 IH

OH* I

I03
4 In 0


-04
".5f1













i- CV
o-t on co


m



0i--l W V H0 4-0






LO = +1 0 i-
V HN


0 0 0 0)
0000






C))) C
HW C&


0) ) 0 Q
M MM M
(0 10 (0 1


0 :j
*] W-
Sl 0)1 l-4l
4-I1) HI *V)
* P 0r4


3 () Q 3
rl M M >


r-l r-l r-l
0 0 0 0

(0 (0 (0 (a
4,1 r. .
OOOOC)C
V]V]WV
(V(V(V(


,-I ,--I
-. .S




CCP C






00 +1, 0
HH O--


0 0+1 0


14

.0
H 1
A% w


0


O 0
C U

Co
b1,-


C--
co
0-t
H


0 oI:L
iC-


i0 0
CC
'H *H1
C C
(0 0






0 0
W) ]
MV(M


0) M
U (4
Qr<


> >
-* U U
.-- C) C
(- (V1 l
*(V l U 1


M l 0 0l0


* II


U d
0



Cr)
(oe
*H(--I C

S0 -HI 0

U HC (0 r 4

l (
B Q <


(0


.4
(0


r-q 0 r-4 W- r-l r-I
0 tAt o0 C




oC 0U -

-,-H
0 4.
C C CC C













0 C)


/] >1 41-I
C) 4 ) 4- ) -
SO
f o






o 4 0




r:
g4 M
C ( rr










C0 C) 0
U I> fa
C


















>*
W 4 X 3 'o E nj
- U
0




C)








O -



> UH ( M
X! 43 6 0(0
U H rl
M > (0 (0 4- 4->
(U h X;
< <


Cl)~















co q
1-1
- -% -~ -
. .. f*-4
0-A.0 .0 -HC-qH
N 0" : 1 H r-H rl o 4-4
0a000c 4 04 404 o'0co 0 0



CCMCCCrCCC-o A4
C CA > CA CA Cl] C A: CA< >

(0 (0 (0 (U (0 (0 (0 0 ( C'O r (04-)
H-HH i r- r- r-HHHi r- o0 i- (


ooMooooooo 0oco
%0 0 -- H -4j N -ac r- %o co to ko c)


.H
-,
E-4






0)






.d


0)









-r4
5-


O N *t 4



H Z Hr-iA3-0H

(U > O HM C CH M -4

TA Or.-ri 0) Q ol 0 a) ONa)
0 r -i H c o- 0 0 0 0 0 -i
Q40 CO 1-CA U Ua4e
CA~~ H 7 ) )b'U
U) ( (UH (U m (U m( U. m(-
5- H H: H H H 0C H H H


0
0c WC
0"


4
U Q)


0 -1


OE-


) a)

00
4-4 4-
*r -1

E-4-4
p-4 1-
BB


HOOOOOO
co


)aQ)

a)a)
4) W
*H -*4

HH
i-l ir-

00
O 1
U

.0
E




1-4
a
0














-4C
n 0
..e
Q,

0













a



mo

p,
--


.>, .
r3 0) 1- +
a) O
(0 i- (

(a p .y 4
S-H
3 0 ( (a r-0 (


0 0 0) PQ-Pp d*
(U (. Ho 0 XH
0- 00 (U(U1<0(0 (0C
z 10- 00
a 0 34 1-4 Q 414 U

S*4 44 004
'0 C0 O'- 0 0O
() 0a) l M p
SE z E-1 u u U


-4 -H *- -H
-H-H -H -Hl


1*-
0


Ha



S4 0
-H
04 04( >
C-H -
Sln r-IH (0
0 ;j z in
4 4

S> M
411 P|


*



H
0

4-)
C



0 (
H-4


>-1


*C

in
14 a
lU)

a




0
U


OH
-H U
















00
00 co




roo
0 0
E-4E-



oO
N N
0 0
k)(1
a w)


0 00 0


vvvC

Q) D) )

,-I t ,-


0o a 000oo
'-I


Q)
U
0)

44
0)











Q)








'4




-4
a'-












4I
-4
a)













-4




0









CO


Q) 4-)


4






a )
,Q
U1
J (0







0
O


0 0)





"4 -4 *l
c c








-H
-4 -4 5-1






-*<1ilf


*
1-1



0 4-




0q
*H g
C 3



*4 >


0 -4


4-)

-I r-I
EHl r-l
.C --4-H
E- l ,
rC 3

t0






-P


a) a)
Q) cu
0 0

CP


a 0








18

and Todd 1980, Moscardi 1979, Gregory 1986, Gregory et al.

in press). Laboratory rearing data are emphasized here,

with virtually all methods involving the non-choice feeding

on plant foliage by laboratory strains of A. gemmatalis to

determine larval survival. In a few cases larval

consumption, feeding efficiencies and development were

recorded (e.g., Slansky 1989, Waters and Barfield in press).

However, all the laboratory studies included percent

caterpillar survival, and therefore this is the larval

performance parameter used to compare the suitability of the

different potential host species. A plant on which less

than half the individuals reared survived is considered a

non-host, as supported by data for well known host plants G.

max, Indigofera hirsuta and Melilotus alba. This value was

arbitrarily chosen and may be too conservative considering

in some cases laboratory survival was only 49% on G. max

foliage, a very common A. gemmatalis host plant (Slansky

1989).

Field and laboratory records are supportive of one

another but both should be treated with caution as they may

not directly overlap (Wiklund 1982). The field data may

include plant species on which caterpillars were collected

following host devastation, although oviposition occurred on

a different field host. Field data may also include what

may be ovipositional mistakes on plant species poorly

utilized by larvae. On the other hand, laboratory results








19

imply consumption and utilization under very unnatural

conditions where no choice in host plant species is

available. Also the condition (e.g., phenophase, fitness,

fertilization) of the plants used in the laboratory study

may be different from the field plants utilized.

Furthermore, the laboratory insect strains, often inbreed

for many years and from single sources, may not be

representative of field populations in terms of tolerance to

host plant allelochemicals or nutrition (Pashley 1986).

Ovipositional data of female A. gemmatalis also are valuable

but are rather scarce (Gregory et al. in press). Ideally,

the combination of all three types of data (laboratory,

field and ovipositional) would be most useful in determining

the A. gemmatalis natural host range, but they have yet to

be recorded extensively.

All reported host plants from field observations occur

in the Leguminosae except single records from the malvaceous

Gossypium hirsutum and Hibiscus esculentus, the Begoniaceae

Begonia sp. and the two host records on rice and wheat

(Gramineae; Herzog and Todd 1980); however all are regarded

with suspicion (Gregory et al. in press). The results of

the bioassays with non-legumes suggest that only members of

the Leguminosae support A. gemmatalis growth and

development.










Caesalpinioideae and Mimosoideae

All Caesalpinioideae and Mimosoideae species tested

(Table 2-1) failed to support larval growth and development.

These species included (% survival in parentheses):

Parkinsonia aculeata (NH); Chamaecrista fasciculata (0%); C.

nictitans (0%); Senna obtusifolia (0%); S. occidentalis

(0%); Schrankia microphylla (0%), and Albizia julibrissin

(0%).


Papilionoideae

Larvae of A. gemmatalis were reared successfully on

only some of the tested species of the Papilionoideae. Non-

hosts within this subfamily include (% survival):

Aeschvnomene viscidula (0%); Apios americana (0%);

Centrosema virginianum (0%); Cercis canadensis (0-20%);

Chapmania floridana (0%); Crotalaria lanceolata (0%); C.

retusa (40%); C. spectabilis (0%); Dalea pinnata (0%);

Desmodium paniculatum (0%); D. tortuosum (< 5-40%);

Ervthrina herbacea (0%); Lathvrus sp. (10-20%); Lens

culinaris (0%); Lupinus villosus (40%); Phaseolus lunatus (<

5%); P. vulgaris (10-27%); Rhynchosia reniformis (0%);

Sesbania vesicaria (0-90%); Vicia anqustifolia (13-90%);

Wisteria frutescens (0%).


Association of Herbivore and Host Species

The Leguminosae is believed to have originated

primarily in Africa and diversified further in South America








21

(Raven and Polhill 1981). The phylogeny of the

Papilionoideae tribes is depicted in Fig. 2-1, which

indicates that all tribes diverged from the ancestral

Swartzieae and Sophoreae (Polhill 1981). Compilation of

laboratory feeding data (Table 2-1) suggests the association

between A. gemmatalis caterpillars and their host species

may be traced phylogenetically back to the tribe Tephrosieae

(Fig. 2-1). The geographic distributions of Papilionoideae

tribes in Fig. 2-1 (e.g., new world tropical) is believed to

be their regions of origin and is not intended to represent

their current ranges (Polhill 1981).

No feeding data are available for species of the most

ancestral tribes, Swartzieae or Sophoreae; one visual search

in Texas of Sophora tomentosa failed to reveal A. gemmatalis

larvae (Gregory et al. in press). Two to three species of

the Tephrosieae were tested, Tephrosia florida, Wisteria

frutescens and Wisteria sp. Larval survival was 90% on T.

florida, but no larvae survived on W. frutescens, whereas

Wisteria sp. (possibly a different species) was listed as a

potential host. Thus, A. gemmatalis caterpillars apparently

continue to maintain the ability to utilize at least some of

these possibly ancestral host species.


New World Tropical Tribes

Following the New World tropical tribes (Fig 2-1), A.

gemmatalis larvae successfully completed development on








22

nearly all the Robinieae species with the exception of

Sesbania vesicaria; however, many other members of this

genus were acceptable hosts. The single member of the

Amorpheae tested was not a host, while half of the four

Aeschynomeneae species tested (one of each subtribe) were

suitable hosts. Thus, the host range in the neotropics

apparently extends through all the tribes.



Old World Tropical Tribes

Four Old World tropical tribes diverged directly from

the Tephrosieae. One, Phaseoleae, is represented by 25

species in these feeding trials, of which six (24%) did not

support A. gemmatalis caterpillar growth and development.

These included two of the three Erythrininae, Ervthrina spp.

and Apios spp., two species of Phaseolus, Centrosema sp.,

and one of the three Rhynchosia species (one of the five

Cajaninae) tested. With the possible exceptions of the

Erythrina and Phaseolus, whose allelochemistry is reviewed

below, most genera of the Phaseoleae served as hosts and the

species that did not probably represent divergences from the

general pattern of host suitability.

Additional Old World tropical tribes include the

Indigofereae, Psoraleeae (not represented here by feeding

trials) and the Desmodieae. Both of the two species tested

of the Indigofereae served as hosts, while two of the four

Desmodieae species (both Desmodium spp.) did not.








23

Although A. gemmatalis larval survival was relatively

high when reared on three Lespedeza species (80-100%),

larval developmental time was significantly greater and

larval growth was reduced (both absolute growth rate and

relative growth rate, with up to 10 molts required compared

with the typical 5-7) relative to most other host species

assayed (Waters and Barfield in press).


Temperate Tribes

The temperate Papilionoideae tribes extend primarily

from the Tephrosieae through the Galegeae and into (for our

survey) the Vicieae, Cicereae and Trifolieae. Only a single

member of the Galegeae was tested (Astragalus villosus) and

all larvae tested survived (100%). Two of the four species

of the Vicieae tested supported A. gemmatalis growth and

development. Larval survival when reared on foliage Vicia

anqustifolia was high in one study (60-90%, Plagens unpubl.)

but low in another (13%, Slansky 1989), suggesting that

variation in plant material may influence the acceptability

of this species. The sole member of the Cicereae and all

four members of the Trifolieae served as hosts.

Another divergence of temperate species radiates

directly from the Sophoreae. The Thermopsideae is believed

to be ancestral to the Genisteae and A. qemmatalis rearing

data suggest that species of both tribes serve as hosts.

The single species tested from each tribe supported A.








24

gemmatalis larval growth and development. Thus, A.

gemmatalis larval feeding specialization may be traced in

the temperate zone from members of the ancestral

Papilionoideae tribe, Tephrosieae, through the Thermopsideae

and Galegeae, to the more advanced Vicieae, Cicereae,

Trifolieae and Genisteae.

The remaining tribes originated in New Zealand,

Australia and Africa. Although no legume species from New

Zealand or Australia were tested, data are available on five

species of a single African tribe, the Crotalarieae. The

rearing data suggest that all members of this tribe, with

the possible exception of Crotalaria retusa (40% survival),

are non-hosts. This may be due to the phylogenetically

isolated position of the tribe, with no known links to the

ancestral host tribes and/or the presence of toxic alkaloids

(see below).


Systematic Implications of Susceptiblity to Other Legume
Pests

The Leguminosae may be divided on the basis of

susceptibility to Uromyces spp. rusts; the Caesalpinioideae

and Mimosoideae are relatively immune to attack, whereas the

Papilionoideae represents about 95% of the host species of

the family (El-Gazzar, 1981). The phylogenetic distribution

of susceptible tribes includes roughly half those listed in

Polhill (1981) in a nearly contiguous arc from the

Phaseoleae to the Crotalarieae (Fig. 2-1). However,








25

susceptible members of the Sophoreae are isolated from more

advanced tribes by the non-susceptible Tephrosieae, and the

Desmodieae and Phaseoleae are isolated from the temperate

tribes Trifolieae, Coronilleae, etc. by the non-susceptible

Carmichelieae. It may be speculated that the susceptible

Sophoreae was at one time linked to the other susceptible

tribes through the Tephrosieae, but this bridge no longer

exists (or has not been found). The distribution of

Uromyces susceptibility is roughly similar to the A.

qemmatalis host range, except that the Robineae to

Aeschynomeneae New World tribes are immune to Uromyces

infection and the Galegeae to Coronilleae temperate tribes

are susceptible.

Bruchid beetles are known to feed on several plant

families but most (84%) of the hosts are legumes (Johnson

1981). Unlike the Uromyces, the bruchids as a group have a

wide host range, feeding on members of all three subfamilies

of Leguminosae. Individual species of bruchids, however,

are restricted in their host range, most often to a single

genus (Johnson 1981). Legume seeds are well protected from

herbivores (Janzen 1969) and the bruchid species that

utilize select legume genera have adapted to their host

defenses (see below).

This suggests that the factors (i.e., nutritional

requirements, chemical defenses) determining Uromyces

susceptibility and Bruchidae attack are similar to the








26

factors influencing A. qemmatalis larval host range. Both

groups are specific in their host utilization; only select

Papilionoideae tribes are susceptible to Uromyces, whereas,

the bruchid beetles are very species specific in their host

utilization perhaps due to the detoxication of species

specific chemical defenses. It may be proposed that

Papilionoideae protection from attack by Uromvces or bruchid

herbivores may be imparted by alkaloids or phytoalexin

isoflavones in the leaves and various seed toxins, non-

protein amino acids, alkaloids, lectins, proteinase

inhibitors.


Legume Chemosystematics

Apart from morphological and cytological distinctions

among the subfamilies of the Leguminosae (Polhill et al.

1981), the Caesalpinioideae and Mimosoideae may be separated

from the Papilionoideae by their anti-herbivore defensive

strategies. The structurally complex chemical defenses

present in the Papilionoideae consist of alkaloids,

isoflavonoids, and non-protein amino acids (Polhill et al.

1981; El-Gazzar and El-Fiki 1977). Mimosoideae and

Caesalpinioideae chemical defenses most commonly include

tannins and terpenoids (Langenheim 1981). Additionally,

many of the species of the Mimosoideae and Caesalpinioideae

maintain ant-attracting extra floral nectaries that function

indirectly as herbivore defenses in place of the








27

allelochemistry of the Papilionoideae (Polhill et al. 1981).

The enzyme systems necessary to produce the complex

allelochemicals are considered relatively expensive

metabolically (Feeny 1976, Rhoades and Cates 1976). The

presumed relatively inexpensive allelochemicals used by the

Mimosoideae and Caesalpinioideae, (e.g., tannins and

terpenes) suggest that both the synthesis of sophisticated

chemical defenses (e.g., alkaloids and non-protein amino

acids) and the maintenance of an ant symbiosis may be too

metabolically expensive (Janzen 1966). Alternatively, the

complex defenses may be unnecessary due to the effective ant

defenses or foliar nutritional deficiencies.


Ecological Function of Plant Metabolites

It may be difficult (if not impossible) to

unequivocally attribute a specific role to a particular

plant metabolite. The ecological roles of many classes of

compounds may vary, performing possibly several purposes

simultaneously or perhaps at different times in the plant's

life cycle (Janzen 1981). As an example, the plant

flavonoids are generally believed to function as UV filters,

anti-herbivore and anti-pathogen metabolites (Swain 1975).

Furthermore, the concentration or presence/absence of the

plant compounds may vary among plant populations (Jones

1972, Nass 1972), parts of the same individual (Whitham

1983), seasonally (Feeny 1970) or with the nutritional








28

status of the plant (Chew and Rodman 1979). Despite several

extensive and thorough reviews (e.g., Mears and Mabry 1971,

Harborne et al. 1975, Kinghorn and Smolenski 1981, Harborne

and Mabry 1982, Harborne 1988a), chemical information

regarding specific plant species (and genera) remains

fragmentary and incomplete; a few allelochemicals from a

relatively small number of legume species have been

described (Gomes et al. 1981). Further aggravating the

situation, only a few of the A. gemmatalis host or non-host

legume species have been chemically described. However,

there is a growing body of literature that suggests many

plant-derived allelochemicals have a profound influence on

insect herbivores. The chemical participants include many

classes of compounds among them are the alkaloids,

flavonoids, non-protein amino acids, and terpenoids

(Harborne 1982), although rarely is a plant species

simultaneously protected by several of these chemical

classes (Harborne 1982).


Chemical Defenses of the Lequminosae

The vast majority of legume allelochemistry information

describes compounds in the species of greatest economic

importance (e.g., Glycine, Phaseolus, and Pisum).

Frequently, the information is used to describe taxonomic

affiliations (chemotaxonomy) within the family (Harborne et

al. 1971, Polhill and Raven 1981). This review compiles








29

information on alkaloids (Willaman and Schubert 1961,

Willaman and Li 1970, Mears and Mabry 1971, Smolenski and

Kinghorn 1981, Kinghorn and Smolenski 1981), flavonoids

(Gripenberg 1962, Hattori 1962, Seshadri 1962, Harborne

1971a, Ingham 1983, Hrazdina 1982, Wollenweber 1982,

Harborne and Williams 1982, Bohm 1982, Chopin et al. 1982,

Dewick 1982, Harborne and Grayer 1988, Chopin and

Dellamonica 1988, Dewick 1988, Wollenweber and Jay 1988,

Harborne and Williams 1988, Bohm 1988), terpenoids (Harborne

1971b, Langenheim 1981) and non-protein amino acids (Bell

1971, 1981, Rosenthal 1982) for those legume species

represented in laboratory feeding trials (Table 2-1). Only

references that pertain to the compounds identified from

foliage (or in some cases, 'aerial parts') are included.

When complete information is provided by a reviewer, space

is saved by citing the review article; however, where the

source of the plant material (i.e., plant part) listed in

the review article is unknown, the specific reference is

cited if it meets the above criteria.


Alkaloids

Alkaloids, heterocyclic nitrogenous compounds, are

restricted in their taxonomic distribution to one-seventh to

one-third of the families of flowering plants, the majority

of which are dicotyledonous (Robinson 1979). With the

exclusion of the orders Magnoliales and Ranales which








30

contain many alkaloid bearing species, the Leguminosae

contain (as of ca. 1974) about the same percentage of

alkaloid-producing species (41% of 1643 species surveyed) as

other dicot families (40% of 4432 species surveyed) (Levin

1976). The distribution of alkaloids within the family is

strongly biased toward the Papilionoideae where only 2 of

the 15 alkaloid classes analyzed phenylalaninee and bicyclic

and tricyclic tryptophane-derived alkaloids) occur in the

Casaelpinioideae and Mimosoideae (El-Gazzar and El-Fiki

1977). Nearly 100 compounds have been reported from the

Caesalpinioideae and Mimosoideae (Smolenski and Kinghorn

1981) while over 350 have been reported from the species of

the Papilionideae (Kinghorn and Smolenski 1981).

Although most alkaloids reported from the Leguminosae

occur in seeds, a few occurrences of alkaloids in foliage

were reported (Table 2-2). Alkaloids represent one of the

most consistently toxic classes of compounds to herbivorous

insects (Janzen et al. 1977). Classes of alkaloids reported

in foliage of Leguminosae species included in Table 2-1

include erythrina alkaloids from the Ervthrina spp.,

pyrrolizidine alkaloids (e.g., monocrotaline) from

Crotalaria spp. and the quinolizidine alkaloids from Lupinus

spp. Only the alkaloid classes represented in the species

list (Table 2-1) are represented here. For a description of

the alkaloid classes see Robinson (1979).











w O0

l l -c r-4 c
H .r-- r- % r%
H H -- va OO
O* N N H i-H
4. +|I H ON 0)
-I H O 4 A
OO | 0) -- 0) *H i HrC *H >

W *H0 4= 0 4 ,>
S4 .H Q H.40 -) Ul NHC
0P. r % 44 4 0
fO C +J- j r-l U O Nr-I U l ^a
(DM m k 4 00 %4 Mo rr-4 r; 0

4- o 4 (n N r'I U a
%14 (l t, 0 N ,--, En 41

-0 I
4 -

C C
0 4 r r- Hr

>1 0 a ) -i -



0 0 0
a ) 10 1 0 c q
- > 0
0 %D a4 rx 4.) a) 4- 0
EO O) i u >1 -


Sl 4. r 0 ) 0 0 O 4-


00 ) *Hr 4 H
-4 i 4 >i- r 4 c 0 iO
OC vO Q l ^ C O



SI M ) r
0 C0 1 r a) C w 0
I )-I 4 'P.10r c az) 0- >i4 0P
0 Hl 0 0) Q) r *'-I C ) Q) 5 *d '*
UN Q crO -A 0 a 4- 0 rq 1)o rd C

09 O Cn >C U3 a ) ) r-l
CHio oe >i W c

u t a ct i c Q UC a+ 10 C1
t-lO c* -* H (O Qr 0 C 0--l 4>U 0' S (
Q*r-j ) 4 Ep) 5 -r5 >
H1) 0 O ir-l '0 )



M 0 z *O 1 -C1
0) C u t U d'- U f i c
4-a)S ;
o 5 c a Os M CO -H

> (0r-I Bl > 4O H Q


O 1: 4 0
:J 4J 04 0 N4


H H U0 > .O
-P4 I- N <0
'0 4 0 0 C

S0 4)
4. -4)H Q aP
049 () AP





IQ 0 0O
H Q9C tl OI


t Ci U -i 4. 5 -
H-4 0) 0 4
r4i & ) Oa '-o

0)0) 0 geS 0)
(tS. -.4r
U












0

%N rH % rH r%

,- '.O Hp ,.0 '3
o -C 01 0i NC
r -(7 ia 10 (- (^ +J a cr> O 1
NHN r q H H
ONo co co O oo

H t o 4 -1 MO 4-) ON M -0-4J
$-4 m H H r H Hm
0 0 o -)-.C )a)--- C O)-' 0
C *4 CO (.oQ I A A -
m r m co e 3 t3 r- 6 o
0 0 < 0l O- ON M 01 -,0 to (0-4
4J 04 O 0,. 'Hr-0 Xr r-4 0( 4 o r-
-4 4J *qHHwH, 0- Mb r-4 Hr- n
0) 10 -H O C C -0 C C r
pO a a H H s H HE




*H *-

SM-H C
S% 04 0


0- 4- r M c

00 00 *O C
V 04 0 )4, Wi-r-4o 04V
H H 0 O 0 M C ~ -H
(I 0 0 41 *cl *0 mO I *) O
--P (0 d ia 9 a -Hq to
0 H C-H0 4 14T$- >4 OH
.H t 0W OM ('-H *- m-4o
ta 0 I 0 1 -* I o
h0 r. H )04 CO -H U r-l
>4 U) 04 o o4 o (1 i 0to
-iC I EC NC t *Hr- --H rH
(o -cH 9l ) H-HO 0 ) O(> 0 Q)i
4- *r < Ha O -r-i 0 4) 0 Q% ) 0)c M e
S0 0 I t ---H > 0 >to -, O 0) O
(0 0 U) in j r C a) Qt C C9-:r-q M e t C0
U0 0 C o -4 0 a 1-i ) -r r () (0 (0 -0 )4 C
z -HG0 I >0 1-i- WH 0 Uo 3.3l-0



(n ** ** ** ** ** *. ** **
> > ** > > > ** > >.*

(o a 0 M
W 4-l 9 0
0 A 4 %4 4 a (o
H

0 p0 H a 0 4 0 t a
V tO 4- o 0 ) ar V

C a 0 4 U-HO OA 0-
*- o V -rH ( ( 0 M .
4P C 3) *Hx-J
C 4 ) 0 00 0*
0 0 0 0 4i U (- 4r' C 4 0 p -
0q *-l D 0 0 (1) E O I-li Q) V- *H>
I *- *0 1 1 -- 3 0 *W W 1 4
0-HN OH V tO O a 3
I C CC 0 b 000i0 4 C >M
N 0 0--H H 5 5 O-4'0 + 3 ,CO
*r *r4 ,Q 0 3 U)- > O 05-
S 0 HO a0 C Q)- ( 3O)
H C1 ) *rIM H H Q 1l M
S4
t0 ot
^ fc

















0n
Am
,-


o



H


r-4
-N
co ,
co00-
0)

OO-
- a)



0
Q4 a


r-
,-t


,-4

SN
a)N
4m
IPP
3p


coco




E)
-H








0
HQ













C






O0

4
0 0

b)
CO
i-l (


0 u


0
-4

O
o i i i

oCo u




4-. '0 4- 0 4-) -4
*H N N In


(U
SP 0 P-4 0 -4 C
km ma)




C C r-C r-l C
(v 0 >4 (o >4 a0 M
0 0 o 0o u 0




:1 p< 3 re 3 4 (u
0000009







0) )) a r-4 1) rl
a a a da)a
0 0) 04
r4 "0 (
'' O h '
00000





ca# cd (


0
I -H
0 -4
CH
0 -)
-H 4 >- )
CH rl H
- rn aC c ca


O C*-l (f ( ^O U U U
0 H-0 0 0 0 U
C H a >r C a) a )
Or-q (0 44 3 4 t3 0 4 r.
e000 00 r t qC J3
WOU OOi-H & &'0&
0 (1 O Q) 0 -r '0'0
c U O O U O 0() O O1 --l o
o >U >>u -t C' C C -
*H0O>4 >4.40) r- OO r.-i 'r
ro r> a4r r- o o o -
0> 10p PO U- 04 0 E

C H -H 4 (>O -4 r- p 3 p O
*H1 0 > 4J 0 U 0 r Q O )-H
Q) C H C w' 0 dF1 4- t3)4l n 44
N'4> r4 40 >0) --W-r4 a a) a4
ra U -ro I U u-A 04 F 0
r >4H -el0 3 >4O lEn (1) (D <
0 r-I Hi-I 1 Or-I w-H 0 (o 0
Oa MrH Ht 0) U M-ri x! r% r 1


*H

0









4M
-H






(0
r04
0),--
.,.












0
N-1.








o.w m
3\00 -
S1-4o o
H 0) r4





C -4 -
t> r-i| r= 1





CM() P) CO

4 4) 09 )
0 >4 lh4 m

m 14 r r


I



OQ)
9-H

H 0
00
-4 -,-
0) 4.)

z -ri




0 U 0 4

Or OH
0).4 04) r

0 04
r ) 0'








EMA
w ME
i- E-i <
< Mt CO


r-I
H

0)
r H
a,




m N

r. 0
44 ,-
co


0H O


f: N
tr N
C O
H 2




C -

S04





PH HIQ
O *H








S--f-Ea 0 0
0) r r-


TI r 4l 0) 7
*H C
M -- H -
C a0 >irH



N U 0) I Q) i
3-*HM 0 0 C

0 ) ,C O >
'0 E 0aNO u


O



0 0

oo

H r-4 U
*>-
03 IJ
HH C.


1-1
r-4


o


CO


0 0o


.H N
tp N
C 0
H z


0)
c c
0 -H
-H -H

.0H
-4 *p


CO
(0 F4
> 4-


CO=
-.44 0
*- 0- 0
Hr -H
r-I C
0 -
0)H1 CC
)M r-1 ( -I
.0 0 > Q

(0 4-4 -H
-. 0 r


0) -H ,d -H X
S4J r- 0
0H- 0)H r- 4
.P 0 0 '
* 0- I 0) >N


X a( ta
.ANA


Z 43
C H
0) ( H

*
E J





C 4.
t4 H

0 C 0
O-l 0
1 4 1 ( d
1- >1











r-

-




) 0



Q. *H N
S..-- 2 -4







rl O0
= -c l
Q) C Or


I





OI'-
O
(1)00 rl


-4 "

O 0
U C>>


a)
CD
a)

























>4
h4
E)
-)
r(























&

-1-1
()
r
O


..
& <

v Ere


(0
in |

CU
<

(0


H
ca



HD k
-) 0 H

r- 4 C0
wH r-
(0 tl


co
m.
LO






H H




-m C
co
M O






o u
u- u






4 -rl 04
C .a



0 z

.Q )






M-H c -H Q)
0 () 4 0
-,u C -
*>1O 0 (r
(Cm z pH3d
M -H 4 1



> 0 -w a w
D -H ) I I0
*<54 ,


SfE Z X r
0*
4
CU

u M
wi -ra
M0 04-4
,y H- I


0 r-4 -

> n > 3 in


r-l COD
00 aO 0





C Q) c









-,-t
HC

C-H l




a-
-4 0
* M 0

0-H >>
O a

U H-l

H H
01 g



a) 4
a)>Ot
>o0 ^1 ^
( o 0(
M 0 > >


a0)


0



C.)
H -

m -H
I H
,-44
4-) U
H)
U >
4 44J
w0
QH U


a
CO







H
i-x


I0

ro-

SX4



C0
C P

ard





(0
rc
QC

c 4
0 04









44 M
u -r
Mc i



O -
4-O(


-E

-4 O I

,M -
-,n r-i -
0 90 C
c Q) -- -
*4'- c C
HC C 0
C ) M *- 01
*r 0t1 C 4
J >-H -UV

Co0 4 0 4
M 0C -H 41 0 H



W 4-4 CM t W u


U






C

O
.,-I
-4
4-4
.,-I
.p
4-)
*H
















C- 0- H o
co oo I o
--4 *




*- H
oo r o




S(o0 0
QmU 0


a) -0



.H M a a-
IQ r- H Lo
r-4 0"- LOn 4-)
-4 r-4 ON (





r- | -H -0
4J H r- 4J 4J r-
o ^P- a a) -) 0

a 1-1 z (Q


,I-
o .-H I
0'1

0) (0 tv


M0 --q Dc"
0 4 (o -o

0 r4 0 I 10
xc: O HO *


0) C 40 -H0
r to a) a) 0 r-O
() r r. 4 r-it
I *rH z X I 4J P
SH 4J) W I H 0)H
( c o nP I C



Pr Q) >a 4 ) 4-)



U W Q--l' 0





**


I -

>1 c
I -H
H 1-
i a
-H 0

r-H 0
S -H
a) a)1 >


4 0 > 0
4P44 O4

C C13-4 0

H -H a )
0a)EI a)1-
to 0 0 --H In
S>i-HU(" -H 0

I >1I r 0
09n 4- ( A I


0
I


C- ) 0
OC -.

-e 44


(d r. .,- V-
- -H 0


O4- O 4) -ri 9 N
) -o-H ~ -O 4- 0
4(0 0 *H 0X-H
r o 0 -COC 4(
04U r) 1- ) -A Q..

,.He -d'f-e- ,c-) -I
UH C I O *H *H *H C
0 0- 404 1 0 r' 0 H -4 0
0C) ) a oH-H-H tc

O4 U C I -Hm-H N O
MO-H-HOH -H -H
1.0' U I O r-0 H'0 *

O*O>a a c te
Maat eHe


t2


0


ozor
LJ 0





A 0m
0o -,

0 e4


co


LM
H-r
m


I)








0

A O
A 0













*-4 ,-et" -tO H- -
-H Q % DA -Q o 'o
H0 H H O-H 0oH 0O
rr L M H 2 0 i-l r
CO -.. O -n N O--N a H -l 10 "--
CO Hl CO H H CO
HGI H 1-O r-IH

-oco ) 0 (04o -
SO UO 0 0 W O C 00

) A0 ., 0 -, O -4 U 0





oO o r.H O < 0 o o 4 o o 4
-f- 0 0 0 .14 O O 00 I
-4 0 0 O 4 4
I En) r-H C-i.- rto
i r I -iE O n 0 In u to

C0 > 0 44 -4 I 0 U r0 >> 3
0 -H -H *) 0 a)s N O 0
oU -0 OP O j i o % o 0 0
u o h0 r c 41 M n o-o 0 o c- t o -.


0> 5q H 1 W 0 a)- %
b O- C -H -r- M i a C O O I O%)4 W O -4


S-l EO- q-H* 9+ON 1N I I u H 0 4 OO C0 WON
o U 0.) 0 ) r 0 04-1 0c 4P 0 4- >
S( 4-4 o0 0 -0 --1 1 P 44 r O U Q44 1
0-s w u)u o n0 41 g0 >4r i i 0) 0) ( z O c -p
0d 01 Co U r l Z ( 0 C40 a)i 0 0, X OH I4rXZ
4J ou) H ) 'C04 -) -) r-ii 04) JO
ucr1 0 H0 0 0 0 4 1) 1 WH 0 4 r-I 7 WO dZ l r-j C 4J ( d 4J

Prr- 0 r-q N W> > WPI 4J0 0W I I-H UU> W N> > OI
or-i (a 0 l oAr 1 -- o-r aor)p i (o op 0 O WM >
A 01 1o 0 H U r -H 1ln -r-l X '0 (4-0 Ir.:
a)1 a) (d Q) (q 0 0 O i0 0 0 -P p 0 a ) 0- -P 0 (a 4) 0
M $4 0 C M 0-4 CD4- 0 r O U W r 0 N O

















0 U
U 2






E-4
C (




















$4 4
'-H C'



000
A 44-)
4 4J


-C
*H W
I
H- O
00 u

- C -
0 w1 H0
HH) N(0 0 ?
O> m)
r-. +0J C 0 H
0 0 I P .p-. (x
Hl C ra a
O4->IO M mm

*H IO X m r- o
0- o r0 -Hi 0 U R

-0) >X 0) (




) 0) 4W 00 0 --1
4 > Nr- 4 -H (0 UH


ON
0\
4-) '-04
-H -*
8 *u



-r. O W r
mO O

H > -H >
'- i -'.Q r.1
0 0


0)
-H



a) 0
9 010
01 0
0) 0)

(0 r1
3 H -
4) O -H



0D 0o4
5 O i


4
4J
0

(r
.r-H

4-)

*H 4i
(0 )



0p


PI

l o


HH





0 0 U
4O 1
O > o O
00



*s--

01 0)




GOc H
O Og I O

t0orl 0 0 0 4)
Z d 04 0 OO rd
O S 0- 0 0014-( C)


-H I 0 H -M M -i
- 0 0 0 I H 0 Q)
IO OI J 4J 0 I U
90W C 0pe


SOO t I 0-HO ) 1



Ut. 30 0 1) 1 p
A r-I m c 0 -i U- U0
S 0 *4 *4P4 > -0lM *&)
CCOH00WCOOCO
-H *I7 C (0 5 a U *HO


C OCOM 0OQ0
U-H Mj M I 1
tr OM 0 ^ i 4-i f : mcn


4-i-
0)
*p

EH







0 1
H


0
0)

0)
4-
0)
















A

Cr.




( 0


H ff


a La




44 0




Xt- > 0
0 X X-
SC O r-

1- 0 0
) 0
O


, V o ^i E



0 0 M 0 -P
0 > 0 >O z > >

%- (a 0 O 04 0
- > 01- W I- 0
* O- > 0 -) O- -


- 0 0 0-i oU o >




MC >I I > I 1 > I
Orl -I I -H -HI 0

P-4r4%M U4 UM r- C O 0 4r -rI CO 0 0-


> I I >- l >-H

1- on m ni4-i u ('d


O




Os'
C.)
E -
r-I
4C -
r -
*iz


A



a-
0

H
1*
C Q)

e


C c
--4H --H C
g o m) go
S.4 0000 --H
WH 04 *H 4J



M ar004 o -r~ -H
S) 4- 0 a( a(0 1



H'd -- >w 04-4 'd-
0* -CCH r 0
o e c *<- *H (



q0 -




M Zl 0 DM .*
3 1 -i : Cr- a4w



r0 O 0 O 1 H -le o (l
! (1) I) >C U r(Orl a) *l



r-1 tP0I0 Wd W O 1
*0 I go *4
c a O-HM *.H m 7 ao)


0C 3 Q)e 0a3


9P O) 0 1>i>o a c) o


*H
rl
-H








C
0




































o I

c$
04






*H
0)
CO



rc%

-o













Pyrrolizidine Alkaloids

The pyrrolizidine alkaloids reported from the plant

species included here were monocrotaline, retusine and

retusamine from C. retusa and monocrotaline and spectabiline

from C. spectabilis. No reports of foliar alkaloids were

found for C. lanceolata. The compilation of feeding trial

information indicates that, no A. gemmatalis larvae survived

on C. spectabilis, whereas 40% survived on C. retusa.

Neonate larvae (40%) apparently survived on C. retusa

foliage until late instars (Plagens unpubl.) yet only a

slight amount of feeding was observed on C. spectabilis

(Cohen unpubl., Plagens unpubl.).

Pyrrolizidine alkaloids are sequestered by herbivorous

larvae (e.g., Utetheisa) and metabolized for an adult

courtship pheromone (Connor et al. 1981), for the

morphogenesis of scent glands (Schneider et al. 1982) and

for defense against spider predators (Brown 1984). Their

antiherbivore activity may be due to their bitter taste

(Glendinning 1989) and thus, they may be repellent to non-

adapted herbivorous species. It is difficult to distinguish

between toxicity and repellency with these mortality data

(Table 2-1), yet I speculate that the foliar pyrrolizidine

alkaloids retusine and retusamine present in C. retusa may

be only moderately toxic or repellent whereas spectabiline,








42

present in C. spectabilis, may be very toxic to A.

gemmatalis larvae. As monocrotaline was present in both

Crotalaria spp. it is probably not responsible for the

mortality of A. gemmatalis larvae.


Quinolizidine Alkaloids

Species within the Genisteae and Thermopsideae (e.g.,

Lupinus spp., Baptisia spp.) are distinguished from most

other Papilionoideae tribes by the occurrence of

quinolizidine alkaloids (Polhill 1981). Examples of these

alkaloids include (among many quinolizidine alkaloids)

epilupinine, hydroxylupanine, lupanine, lupinine, and

sparteine from the foliage of Lupinus spp. and sparteine

from Baptisia lactea foliage (Table 2-2).

The quinolizidine alkaloids (if not all

allelochemicals) may be quite species-specific in their

activity toward herbivorous species. High concentrations of

quinolizidine alkaloids (i.e., 17-oxosparteine, sparteine,

12,13-dehydrosparteine and lupanine) impart aphid (Aphis

cytisorum Hartig) resistance in Cytisus and Lupinus (Wink et

al. 1982). Sparteine was also toxic to the bruchid beetle

Callosobruchus maculatus (Fabr.) (Janzen et al. 1977) while

the same compound is a repellent of the cabbage butterfly

Pieris brassicae (L.) (Schoonhoven 1973). The grasshopper,

Melanoplus bivittatus (Say), is deterred from feeding on

lupinine containing plants (Harley and Thorsteinson 1967).








43

However, other species may be tolerant of alkaloids or

use them as cues for host finding. For example, sparteine

from Lupinus spp. and other legume species is not toxic to

the aphid Aphis rumicis L. (Tattersfield et al. 1926) yet is

a feeding stimulant of the aphids Acyrthosiphon spartii

(Koch.) on broom (Smith 1966). In lupines, sparteine

comprises one component of a complex of alkaloids that

protect against the flower-feeding Glaucopsyche lyqdamus

Doubleday (Dolinger et al. 1973). These workers also found

that several alkaloids occurred simultaneously in lupines

and collectively protected the plant as they were in high

concentrations but individually the compounds occurred at

low levels. The authors speculated that this variability

imparted more permanent resistance in the plant as the

herbivores would not be likely to metabolize all defensive

compounds. Additionally, relatively low concentrations

(0.6% dry weight) of lupinine and sparteine incorporated

into the artificial diet of the armyworm Spodoptera eridania

(Cramer) reduced survival and larval growth (Johnson and

Bentley 1988). However, the influence of these alkaloids on

the A. gemmatalis host range is uncertain as only two

potentially quinolizidine-bearing species were tested (i.e.,

Lupinus villosus and Baptisia lactea) resulting in 40 and

70% larval survival, respectively.










Erythrina Alkaloids

With regard to the erythrina alkaloids, only an

unidentified alkaloid (possibly erysotrine, or a-

erythroidine as reported from the foliage of other Ervthrina

spp., Hargreaves et al. 1974) has been reported from

Ervthrina herbacea foliage. Additionally, the seeds of E.

herbaceae contain hypaphorine, erysodine, erysopine

glucoerysodine (Mears and Mabry 1971, Hargreaves et al.

1974). The erythrina alkaloids are known to block

acetylcholine receptors (Robinson 1979) but as no other

chemical information is known for this local species, more

work is required to investigate the relationship between the

low survival (0%) of A. qemmatalis larvae and E. herbacea

allelochemistry.


Miscellaneous Ungrouped or Protoalkaloids

Several other alkaloids and protoalkaloids (non-

heterocyclic nitrogen compounds) were reported from plants

tested for A. gemmatalis host utilization. These plant

species include the non-hosts, Senna occidentalis (0%

survival), Albizia julibrissin (0% survival), P. vulgaris

(10-27% survival), and the hosts Indigofera hirsuta (87-100%

survival), Pachvrhizus erosus ( field host), G. max (49-

100% survival), P. sativum (40-70% survival), and M. sativa

(field host). No clear relationships are apparent from the

alkaloid compounds reported and the host status data.








45

Because several species of alkaloid-bearing

papilionoids in the genera Erythrina and Crotalaria did not

serve as laboratory hosts of A. gemmatalis (Table 2-1),

these alkaloids may be important in determining the host

range extension of this herbivore. Collaborative data

(Slansky and Wheeler unpubl. data) suggest that A.

gemmatalis larvae may be sensitive to alkaloids in general,

as the larvae suffered 100% mortality when the purine

alkaloid caffeine was incorporated in artificial diet at

between 0.1 and 0.5% fresh weight.

A few alkaloids have been reported in the

Caesalpinioideae (e.g., cassinine in the Cassia s.l.,

Mulchandani and Hassarajani 1977) but far more

Papilionoideae species produce alkaloids and far more

alkaloids are produced by the members of this subfamily.


Non-protein Amino Acids

Non-protein amino acids have been reported primarily

from the Leguminosae (Bell 1981) and may be found in all

plant tissues but generally are concentrated in the seeds

(Rosenthal and Bell 1979). Over 80 non-protein amino acids

are known from the Leguminosae; among them, canavanine has

been surveyed extensively and has been found only in the

Papilionoideae (Turner and Harborne 1967). Although

extensive surveys of the remaining compounds have not been

conducted, they may be distributed widely in all subfamilies








46

(see below). The toxicity is generally due to incorporation

of non-protein amino acids into proteins and their

subsequent failure to function normally (Pines et al. 1981).

Several non-protein amino acids incorporated into

artificial seeds were toxic to larvae of the seed-eating

bruchid C. maculatus (Janzen et al. 1977) and inhibited

feeding in Locusta (Navon and Bernays 1978). The non-

protein amino acid L-canavanine was toxic to the tobacco

hornworm, Manduca sexta (L.) (Dahlman and Rosenthal 1976).

The non-protein amino acid, B-cyanoalanine, disrupts the

water balance of Locusta migratoria L. (Schlesinger et al.

1976). However, bruchid beetles (Caryedes basiliensis),

adapted to feed on seeds rich in L-canavanine, may use the

toxic metabolite as a source of nitrogen (Rosenthal et al.

1982).

Although much information is available on seed non-

protein amino acids, very little could be found on their

foliar levels because either they rarely occur in foliage or

foliar non-protein amino acids have yet to be thoroughly

investigated. However, I did find literature citations

describing non-protein amino acids (Table 2-2) occurring in

Arachis hvpoqaea, P. sativum, M. sativa, all considered A.

gemmatalis hosts (Table 2-1).

Also included here are the plants that sequester

selenium from soils and incorporate it in amino acids.

These frequently occur in the legume genus Astragalus (among








47

others) (Rosenthal and Bell 1979), which often grow where

selenium is abundant in soils (Shrift 1969). Although these

accumulator plants are toxic to humans and livestock (Shrift

1969), few examples of insect toxicity were found. Bruchids

and seed chalcids, adapted to feeding on high selenium-

containing Astragalus seed pods, were very tolerant of

levels toxic to various mammals (Trelease and Trelease

1937). However, the foliage of selenium-accumulating

Astragalus sp. was toxic to aphids and various mite adults

and eggs (Gnadinger 1933, Hurd-Karrer and Poos 1933).

Only a single species of this genus, A. villosus, a

non-accumulator (Barneby 1964), has been tested as a

potential host plant of A. gemmatalis with 100% larval

survival (Table 2-1). Thus, the information reviewed

suggests that non-protein amino acids have little influence

on the host range of A. gemmatalis larvae.


Flavonoids

The flavonoids comprise more than 4000 unique chemical

structures (Harborne 1988a) and are widely distributed

throughout the angiosperms (Swain 1975). This estimate may

represent only a fraction of the flavonoids described,

considering the small percentage of the plant species

thoroughly investigated (Harborne 1988b). Flavonoids

frequently serve as either herbivore attractants or

deterrents depending upon the compound and/or herbivore








48

species (Harborne 1988b). Flavonoids may be involved in

regulation of rhizobium root nodulation, or protection from

fungal pathogens (Harborne 1988b), nematodes (Kaplan et al.

1980) or insect herbivores (Hart et al. 1983).

Among the many classes of flavonoids (Harborne 1988a),

I will discuss only those that have been shown to influence

herbivore performance, namely the flavones, flavonols,

flavonoid glycosides, isoflavones and anthocyanins.


Flavone and Flavonols

Flavones and flavonols are generally localized in

surface foliar tissues or structures (e.g., glandular

trichomes) (Wollenweber and Dietz 1980) and are associated

with lipophilic substances (e.g., terpenoids and waxes)

(Wollenweber 1982). These natural products are not widely

known from members of the Leguminosae (Wollenweber and Jay

1988).


Subfamily differences in flavonoid allelochemistry.

Despite the occurrence of flavonoids in all tribes of the

family (and all flowering plants examined) some groups of

compounds predominate in select legume subfamilies. The

common flavonols myricetin, quercetin and kaempferol are

almost exclusively restricted to the arborescent

Caesalpinioideae and Mimosoideae (Harborne 1971a). The

tannins of the leucoanthocyanidin group (proanthocyanidins)

are widely distributed in these two subfamilies, whereas








49

they occur in only a few woody Papilionoideae species

(Haslam 1982). On the other hand, two common flavones,

luteolin and apigenin, have been recorded only from the

Papilionoideae (El-Gazzar & El-Fiki 1977).

This raises the possibility that the mentioned

flavonols and tannins are toxic or deterrent, thereby

limiting the utilization of the Caesalpinioideae and

Mimosoideae, whereas the Papilionoideae flavones attract or

are otherwise innocuous to A. qemmatalis larvae. Kaempferol

was reported from many species of the non-host genus Cassia;

however, kaempferol and quercetin glycosides were reported

from foliage of the Papilionoideae species G. max, P.

sativum, P. vulgaris and Baptisia lactea (Table 2-2). Of

this group, only P. vulqaris is regarded as a non-host

(Table 2-1). Additionally, kaempferol and quercetin have

been reported from many Lupinus spp. (Table 2-2), and L.

villosus is a non-host of A. gemmatalis. Furthermore, both

the flavones luteolin and apigenin commonly occur in Lupinus

spp. Thus, as the flavonols were reported from both the

Caesalpinioideae and Papilionoideae and from host and non-

hosts they apparently do not have an influence, or do not

occur at concentrations that influence the A. gemmatalis

host range.

A survey of the flavonoids of 73 species of Lupinus

from western North America revealed some generalities of the

genus. Although our local species, L. villosus was not








50

included, representative flavonoids of the genus may include

the flavones (and many glucosidic forms) apigenin, acacetin

(also present in the A. gemmatalis potential host R.

pseudoacacia), luteolin, chrysoeriol, and the flavonols

kaempferol, quercetin, vitexin and orientin (Nicholls and

Bohm 1983). The relatively low percent survival (40%)

observed when A. gemmatalis larvae were reared on L.

villosus (Table 2-1), suggests that one or more of these

flavonoids may be toxic or repellent under these

circumstances. Furthermore, the lupines contain many active

alkaloids (see above) that may have contributed to the

observed larval mortality.

Insect herbivore bioassays of several of these

compounds revealed considerable activity. The flavones

vitexin, luteolin and the flavonols quercetin, quercitrin,

rutin, morin, myricitrin and tricin extracted from wheat,

deterred feeding of one or both aphid species Schizaphis

qraminum (Rondani) and Myzus persicae (Sulzer) (Dreyer and

Jones 1981). Quercetin, catechin (a flavan-3-ol), and

naringenin (a flavanone), when incorporated into artificial

diet, were all toxic to the aphid S. graminum (Todd et al.

1971). Larval growth of Heliothis zea (Boddie), H.

virescens (Fabr.) and Pectinophora gossypiella (Saunders)

was reduced when their artificial diet contained the

flavonol quercetin or two of its glycosides, quercetrin or

rutin (Shaver and Lukefahr 1969, Elliger et al. 1981). In








51

contrast, rutin added to the artificial diet of the cricket

Acheta domesticus (L.) accelerated growth (McFarlane and

Distler 1982, Neville and Luckey, 1971).

The glycosides of myricetin and quercetin from Rhus

spp. and Schinus stimulated feeding by the flea beetles

Blepharida spp. (Furth and Young 1988). Other flavonoid

glycosides have stimulated feeding in several herbivore

species: luteolin-7-glucoside, isolated from Salix

gracilistyla foliage, stimulated feeding of the leaf beetles

Chrysomela vigintipunctata costella (Marseul) and Lochmaea

capreae cribrata Solosky (Matsuda and Matsuo 1985), and the

beetle Agasicles sp. uses 7-a-L-rhamnosyl-6-methoxyluteolin

as a feeding attractant to its host plant alligatorweed

(Alternanthera phvlloxeroides) (Zielske et al. 1972). The

six flavonol glycosides of kaempferol and quercetin,

isolated from soybean leaves (Buttery and Buzzell 1975), may

similarly stimulate feeding in A. gemmatalis larvae.

Flavonoids utilized as ovipositional stimulants have

been reported for two swallowtail butterflies, Papilio spp.,

both on Citrus. The flavanone glycosides naringin and

hesperidin from sour orange, Citrus natsudaidai, were

synergistic in their activity, only active when combined or

when included in an unidentified mixture (Honda 1986). The

extracts of fresh leaves of C. unshiu containing 6,8-di-C-B-

D-glucopyranosylapigenin were also synergistic with other

compounds in an unidentified mixture (Ohsugi et al. 1985).








52

The aggregations of A. gemmatalis adult males on sweep nets

repeatedly swept through soybean foliage suggest a similar

stimulant (Gregory 1986). Perhaps A. gemmatalis males

locate suitable habitats by flavonoids released when foliage

is injured by herbivores or sweep nets. However, the

attraction of adult Lepidoptera to salts (e.g., puddling) is

well documented (Arms et al. 1974) and may also explain

these aggregations.


Isoflavones

The isoflavones constitute one of the most active

classes of legume compounds. Isoflavones are synthesized de

novo (phytoalexins) when elicited by topical application of

cupric chloride (Burden and Bailey 1975), irradiation with

UV light (Hart et al. 1983) or pathogen infusions (Keen et

al. 1971, Ingham et al. 1981). Among their many activities,

these compounds are oestrogenic in livestock, antifungal,

antibacterial (Ingham 1983) and deter herbivore feeding

(Hart et al. 1983). Additionally, their toxicity to insects

is well known, as demonstrated by the insecticide rotenone,

extracted from Derris roots (Krukoff and Smith 1937). The

fractions containing the isoflavones phaseol and afrormosin

(also present in P. sativum and Centrosema spp. foliage)

extracted from resistant (PI-227687) soybean foliage were

highly toxic (71-98% mortality) to the soybean looper

Pseudoplusia includes (Walker) (Caballero et al. 1986).








53

Many other examples of isoflavones are distributed

widely in the Papilionoideae (Table 2-2, e.g., citations by

Dewick 1982, Ingham 1983, Dewick 1988) but their activity

toward insects is largely unexamined. However, a few

surveys have been conducted primarily with the root-boring

scarab beetle, Costelytra zealandica White, with which

feeding stimulation or repellency was assessed (Table 2-3).

The results suggest that both A. gemmatalis non-host and

host species contain isoflavones that deter scarab feeding.

Some of these repellent isoflavones that occur in A.

qemmatalis non-hosts include phaseollin, phaseollidin,

genistein, kievitone, formononetin and coumestrol.

Commercial coumestrol, however, had no impact on P.

includes when incorporated alone in artificial diet,

however it may be but one component of herbivore resistance

in G. max (Smith 1985). The picture is far from clear

regarding the impact of the majority of the isoflavones in

A. gemmatalis host or non-host species. However, the

greatest activity (towards P. includes) occurs in G. max

foliage, namely the isoflavones phaseol and afrormosin.


Anthocyanins

The anthocyanins are generally regarded as plant

pigments located in flowers fruits, leaves and storage

organs (Harborne and Grayer 1988). However, anthocyanins

are also produced in response to stress and form the red











Table 2-3. The influence of various isoflavones on the
feeding of several insect herbivores 1



Isoflavone Source (Table 2-2) Feeding 2


phaseollin

phaseollidin

vestitol

2'-hydroxy-

formononetin

2'-hydroxy-

genistein

medicarpin







maackiain



pisatin

kievitone





luteone

genistein



biochanin A


V. unguiculata, P. vulgaris,

V. unquiculata, P. vulgaris,

M. sativa



M. deeringiana, T. reopens,



P. vulgaris, C. cajanus, T. repens,

M. deeringiana, V. unguiculata,

A. hypogaeae, Lathyrus spp.,

C. arietinum, M. alba, M. sativa,

T. repens,

M. deeringiana, Lathyrus spp.,

P. sativum, C. arietinum,

Lathyrus spp. P. sativum,

Desmodium spp., Lespedeza spp.,

M. deerinqiana, V. unquiculata,

P. lunatus, P. vulqaris

Lupinus spp.

M. deeringiana, G. max, B. lactea,

M. sativa, C. calan, B. lactea

C. arietinum, M. sativa, B. lactea


3



3

3





3



NS3

NS3








Table 2-3.--Continued. 55


Isoflavone Source (Table 2-2) Feeding



formononetin C. virginianum, C. calanus, NS3

P. sativum, C. arietinum, M. sativa,

T. repens, B. lactea

coumestrol G. max, V. unguiculata,

P. vulgaris, T. repens, M. sativa NS3&4




1. Sources: Russell et al. (1978), Lane et al. (1985),
Dreyer et al. (1987), Lane et al. (1987).
. Feeding deterrent (-), or no change in feeding (NS)
. Feeding tests conducted on Costelytra zealandica; or
4 Aphid feeding test.








56

color of autumn leaves and the halo surrounding the site of

pathogen infection (Hrazdina 1982). Although not

implicated, the poor performance (low pupal biomass) of A.

gemmatalis larvae fed senescent soybean foliage (Moscardi et

al. 1981) may be attributed to the accumulation of

anthocyanins; however, the impact of reduced nutritional

leaf quality (e.g., nitrogen, Egli et al. 1978) during this

soybean plant phenophase can not be ruled out. Furthermore,

anthocyanins have yet to be reported from mature or even

senescent soybean foliage. However, the anthocyanin

malvidin has been reported from the hypocotyl and stem of G.

max seedlings (Nozzolillo 1973) and is possibly translocated

to senescent foliage.

Further evidence for the toxicity of anthocyanins

occurs in the non-host Albizia julibrissin, found to contain

an unidentified anthocyanin in foliage. Additionally,

anthocyanins were reported from the non-hosts Lens culinaris

(stems; delphinidin), Lupinus sp. (hypocotyl, cotyledon,

petiole; cyanidin), P. vulgaris (hypocotyl and stem;

malvidin), P. lunatus (hypocotyl and stem; malvidin) and

Sesbania punicea (hypocotyl and stem; cyanidin and

delphinidin), Wisteria frutescens (stem; unidentified) and

from the hosts Medicago sativa (hypocotyl and stem;

unidentified), M. lupulina, (hypocotyls; cyanidin) Vigna

unguiculata (stem, petiole and leaf; cyanidin) and V.

luteola (stem, unidentified) (Nozzolillo 1973, Nozzolillo








57

and McNeill 1985). Because many of these do not occur in

the foliage they are not noted in Table 2-2.

Only three studies were found evaluating the effect of

anthocyanins on insect herbivores. The cotton leaf

anthocyanin cyanidin 3-glucoside inhibits larval growth,

possibly by reducing nutrient digestibility, in the cotton

pest H. viriscens (Hedin et al. 1983). However, the tomato

pest H. zea is not adversely affected by the tomato

anthocyanin petanin (Isman and Duffey 1982, Harborne and

Grayer 1988). Thus, no clear interpretation of these

limited data may be made regarding the influence of

anthocyanins on A. gemmatalis survival. The fact that they

occur in leaves of the non-host A. julibrissin suggests they

may be toxic or repellent. In some circumstances (e.g.,

foliage senescence) the compounds may accumulate in tissues

normally fed upon. However, at this point I can not make a

definitive statement regarding their influence on A.

gemmatalis host range.



Terpenoids

The terpenoids are widely distributed throughout all

living organisms and comprise many biologically important

classes of compounds (Mabry and Gill 1979). Examples of

this diverse group include monoterpenes, such as the

pyrethroids from Chrysanthemum (Compositae) and the resin

constituents of coniferous trees (e.g., Pinus ponderosa,








58

Abies qrandis); sesquiterpene lactones frequently associated

with leaves (e.g., glandular trichomes); and the

triterpenoid cucurbitacins and saponins (Mabry and Gill

1979). Although legumes (Caesalpinoideae) containing foliar

terpenoids have been reported to affect the performance of

associated herbivores (Langenheim et al. 1986), terpenoids,

with the exception of the saponins, have not been reported

in the leaves of potential A. gemmatalis host species (Table

2-2).

Saponins occur in the foliage of several forage

legumes, including alfalfa and ladino clover (Applebaum and

Birk 1979). Alfalfa and soybean saponins may have several

modes of action against insect herbivores; they may inhibit

the enzymes a-chymotrypsin and cholinesterase in Tribolium

midguts (Ishaaya and Birk 1965), they may be toxic due to

sterol inhibition (Birk and Peri 1980, Shany et al. 1970) or

herbivore consumption may be reduced due to their bitter

taste (Applebaum and Birk 1979). Most of the examples of

saponin activity have been directed toward the Coleoptera,

however, a few examples were found including aphids and

stinkbugs.

Saponins from soybean seeds were toxic to the weevils

Callosobruchus chinenisis L. and Sitophilus oryzae (L.)

(Applebaum et al. 1965, Su et al. 1972). Horber (1964,

1972) found non-preference and antibiosis against the white

grub (Melolontha vulgarus F.) in high saponin content roots










of resistant alfalfa strains. Pea aphids, Acyrthosiphon

pisum (Harris), were the only insect herbivore of 5 tested

species to be affected by the high foliar saponin content of

alfalfa cultivars imparting resistance (Pedersen et al.

1976). Saponins, when included in an artificial diet,

caused 35% mortality and increased developmental time ca. 3-

fold in the aphids M. persicae. Additionally, in choice

tests, 100% of the aphids (n=40) preferentially fed on an

unadulterated diet instead of a saponin treated diet

(Schoonhoven and Derksen-Koppers 1976). However,

developmental times of nymphs and adults of the southern

green stinkbug, Nezara viridula (L.), were not affected by

saponin (from Gypsophylla sp., Caryophyllaceae) added to

artificial diet (Bowen 1988). Furthermore, saponins

apparently function as a feeding attractant to the Mexican

bean beetle in P. vulgaris foliage (Nayar and Fraenkel 1963,

citing Lippold 1957).

Saponins have been reported from the foliage of four

potential legume hosts of A. qemmatalis (G. max, P.

vulgaris, M. sativa and T. repens) (Table 2-2). Although

there are no larval rearing data on M. sativa (Table 2-1),

only P. vulgaris is a non-host (10-27% survival, Table 2-1).

Saponins do not appear to explain the A. gemmatalis

host range observed in Table 2-1. Considering that saponins

are most active (foaming) at pH's of 4.5-5.0 (Mangan 1958,

1959), their activity may be minimal in lepidopterous larval








60
midgut where pH's are generally in the range of 8.3-8.7

(Berenbaum 1980). Although the foaming action of saponins

is considered to be responsible for bloating in ruminants,

the influence of foaming on insect herbivore toxicity is as

yet unknown.


Cyanogenic Species

Cyanogenic plants release hydrogen cyanide gas (HCN)

when tissues containing enzymes and cyanogenic precursors

(e.g., cyanogenic glycosides) are disrupted. Approximately

500 genera in 100 families contain cyanogenic species. The

Leguminosae is noted as one of the major sources of

cyanogenesis, containing 125 cyanogenic species (Conn 1979).

While the toxicity of these compounds to mammals is well

documented, little attention has been given to their

toxicity toward insects (Conn 1979).

Of the potential host species listed in Table 2-1, only

three produce foliar cyanogenic compounds. These include

linamarin and lotaustralin from T. repens foliage and

phaseolunatin from P. lunatus and P. vulqaris. Rearing data

(Table 2-1) suggest that while the two Phaseolus spp. were

poor hosts, A. gemmatalis larval survival on T. reopens

foliage was 81-90%, suggesting this species was a suitable

host.

The difference in toxicity among the different

cyanogenic species may have many potential causes. Possible








61

explanations include the relative toxicity of the different

cyanogenic glycosides found in the plant species (e.g.,

linamarin vs phaseolunatin), their susceptibility to

detoxication or concentration differences among the host

species. Furthermore, populations of T. repens are

polymorphic where both cyanogenic and acyanogenic

individuals occur (Jones 1972, Nass 1972) and possibly

plants used in the feeding trials lacked toxic levels of

cyanide.

On the other hand, the cyanogenic glycoside phaseolutin

at low concentrations in P. lunatus and P. vulgaris foliage

elicited a strong biting response from the Mexican bean

beetle, while at high concentrations phaseolutin acted as a

feeding deterrent (Nayar and Fraenkel 1963). Similarly,

larvae of the southern armyworm, S. eridania, were

successfully reared on cyanogenic Lotus corniculatus

possibly due to detoxication of the cyanide (Scriber 1978).

Furthermore, S. eridania larvae grew as well or better on

diets containing cyanide than on diet lacking added cyanide

(Brattsten et al 1983). It is not known, however, if any of

these cyanogenic glycosides elicit a similar response.


Proteinase Inhibitors

Proteinase inhibitors occur widely in all plant life

and may serve several functions including the regulation of

proteolytic enzymes and the protection of tissues from








62

herbivory (Ryan 1979). Knowledge of the taxonomic

distribution of this group of chemicals is rather

incomplete; the majority of studies surveyed included only

seeds from economically important plant species (Ryan 1979).

Accumulation of proteinase inhibitors, mediated by a

wound hormone proteinase inhibitor inducing factor (PIIF),

occurs when tomato and potato foliage is damaged by the

beetle, Leptinotarsa decemlineata (Say) (Green and Ryan

1972). The wound hormone was translocated throughout the

plant triggering an immunological response to herbivory.

Wound-induced tomato proteinase inhibitors reduced the

larval growth rate of the beet armyworm, Spodoptera exicua

(Hubner) (Broadway et al. 1986).

Proteinase inhibitors were also very active in the

foliage of other legume species to both endogenous inducers

and the known tomato inducer (Walker-Simmons and Ryan 1977).

The most actively induced species was M. sativa, while other

species responded with minor activity to the inducers (L.

culinaris, P. vulgaris, P. sativum, and T. repens). No

rearing data are available for A. gemmatalis feeding on M.

sativa; however Ellisor and Graham (1937) reported

collecting larvae in an alfalfa field. Pisum sativum and T.

repens are hosts of A. gemmatalis, but, L. culinaris, and P.

vulgaris are considered non-hosts (Table 2-1). Thus, the

occurrence of proteinase inhibitors in these species does

not consistently explain the host range data.












Miscellaneous Defenses

Lectins

I have reviewed several other classes of compounds that

have been reported almost exclusively in tissues other than

foliage or have not been reported from species included in

Table 2-1. These include the lectins, or

phytohemagglutinins, that, although they occur in over 600

species of legumes, are reported in highest concentrations

in seeds (Toms and Western 1971). Lectins are probably

synthesized in leaves and rapidly translocated to seeds

perhaps too quickly for detection (Liener 1979). Lectins

are toxic to non-adapted species of insect herbivores

(Janzen et al. 1976). However none of the plant species

listed in Table 2-1 have been reported to contain lectins.


Simple Phenolics

Coumarins, like the cyanogenic compounds, are

hydrolyzed by compartmentalized enzymes and substrates

following tissue disruption (Haskins and Gorz 1961).

Coumarins may act as either insect herbivore attractants or

feeding deterrents (Manglitz et al. 1976). Coumarins have

been reported from Melilotus alba foliage (Table 2-2), and

this species served as an A. qemmatalis host, with 56-96%

survival (Table 2-1). Thus, coumarins may not restrict the

host range of A. gemmatalis.








64

In a search for the chemical nature of resistance in

soybean foliage to insect herbivores, a higher sterol

content was reported from insect resistant foliage compared

with susceptible foliage (Tester 1977). However, the direct

bioassay of the identified soybean foliar sterols failed to

reveal correlations between insect resistance and sterol

content or imbalance (Grunwald and Kogan 1981). The soybean

foliage sterol pinitol (also present in Indicofera

suffruticosa), when incorporated into artificial diet and

bioassayed, reduced growth of H. zea larvae (Dreyer et al.

1979) and the mechanism of this reduced growth was revealed

as inhibition of consumption and reduced feeding efficiency

(ECD, Reese et al. 1982).


Nutritional Limitations

The production of symbiotic nitrifying bacteria (e.g.,

Rhizobium spp.) in root nodules, is generally restricted to

the Caesalpinioideae and Papilionoideae. Root nodules are

generally lacking in non-legumes and the Mimosoideae (Corby

1981). Herbivore species specialization on members of the

Leguminosae may be due to the intake of greater amounts of

root nodule-produced nitrogen present in legume foliage

compared with other forage species (e.g., temperate and

tropical grasses, Lyttleton 1973). Foliage feeders may be

frequently nitrogen-limited; when fed low nitrogen content

foliage, they often exhibit reduced feeding efficiencies and








65

growth rates, and increased consumption rates compared with

foliage of higher nitrogen content (Tabashnik and Slansky

1987).

However, nodule-produced nitrogen in soybean (allantoin

and allantoic acid) may not be as available to foliage

feeding herbivores (e.g., Epilachna varivestis Mulsant) as

amino acid nitrogen (Todd et al. 1972, Wilson and Stinner

1984). The availability of nodule-produced nitrogen in

other legume species to herbivores is not known. Thus,

nodule produced nitrogen does not entirely explain the A.

gemmatalis host range including only the Papilionoideae.

The nitrogen utilization efficiency of nodule produced

versus amino acid nitrogen has yet to be determined for A.

gemmatalis larvae.


Physical Mechanisms of Defense

While I have not exhaustively reviewed the physical

characteristics that impart legume resistance, published

accounts where these mechanisms may be important are listed

here. The mode of action of the Desmodium spp. defenses

involved hooked trichomes that entangled young caterpillars

(Cohen unpubl.) and damaged mandibles and crochets (Plagens

unpubl.). Similarly, mortality (13%) due to glandular

trichomes was found when the cotton pest H. virescens was

reared on the alternate host D. tortuosum (Hallman 1985).

Foliage consumption and A. gemmatalis larval weights were








66

reduced when fed G. max tawny pubescent genotypes (possibly

due to anthocyanins) compared with the gray pubescent

genotypes under field and laboratory conditions (Lurding

1984).


Conclusions

The evolution of the larval host range of A. qemmatalis

within the Papilionoideae may be traced phylogenetically

back to the Tephrosieae. Host utilization patterns follow

the divergences in Papilionoideae taxa in the Old World and

New World tropics and temperate regions. No species of the

phylogenetically isolated African Crotalarieae serve as A.

gemmatalis larval host plants.

The most active compounds thus far discovered toward A.

gemmatalis larvae include the isoflavones afrormosin and

phaseol. Additionally, the pyrrolizidine alkaloids present

in the Crotalaria spp. may limit exploitation of this genus

by A. gemmatalis larvae.

The occurrence of afrormosin and phaseol only in the

host species G. max suggests that many other factors may

have influenced the compiled data (e.g., concentration,

toxicity, detoxication). Considering the degree of

evolutionary interactions occurring at the biochemical level

that may have occurred between A. gemmatalis and its host

plants, the description of presence or absence of compounds

may be misleading.














CHAPTER 3
TOXICITY OF NON-INDUCED AND HERBIVORE-INDUCED
EXTRACTABLES FROM SUSCEPTIBLE AND RESISTANT SOYBEAN
FOLIAGE TO NON-ADAPTED AND SOYBEAN-ADAPTED NOCTUID HERBIVORES



Introduction

Virtually all soybean [Glycine max (L.) Merrill]

germplasm being developed for resistance to herbivores is

derived from the plant introductions PI171451, PI227687 or

PI229358 (Herzog et al. unpubl.). Because these genotypes

lack acceptable agronomic characteristics, they are used

primarily as breeding lines for the development of resistant

cultivars (Hartwig and Edwards 1985). The latter two

genotypes adversely affect nearly all major soybean

herbivores (Smith 1985). The mode of action of these

defenses is a combination of antixenosis, where larval

feeding is reduced, and antibiosis, where high larval

mortality occurs (Reynolds et al. 1984). The chemical

mechanisms of soybean foliar resistance have been discussed

in Chapter 2.

Increased levels of insect resistance following

herbivory, or induced resistance, are correlated with

greater amounts of alleged plant defenses (Schultz and

Baldwin 1982). Induced resistance reduces subsequent








68
herbivore performance (Haukioja and Niemela 1977, Wallner

and Walton 1979, Raupp and Denno 1984) and herbivore numbers

(Karban and Carey 1984, Karban et al. 1987). Following

herbivory, a mobilization and/or de novo synthesis of

allelochemicals may systemically protect foliage adjacent to

the feeding site, and sometimes throughout the plant, from

further herbivory (Schultz and Baldwin 1982, Karban and

Carey 1984). Induced resistance in soybean foliage has been

reported to reduce herbivore growth and consumption (Hart et

al. 1983, Reynolds and Smith 1985, Chiang et al. 1987).

This study describes the chemical mechanisms of constitutive

and induced resistance found in soybean foliage and

determines their impact on non-adapted and adapted soybean

herbivores.


Methods and Materials

Experiment 1. Induction of Susceptible and Resistant Soybean

Aqronomic practices. Two soybean lines (provided by E.

E. Hartwig, USDA Delta Branch Experiment Station,

Stoneville, Mississippi) were planted in late July 1986 at

the North Florida Research and Education Center (NFREC) at

Quincy, Florida. The soybean genotypes consisted of the

resistant plant introduction PI229358 and the susceptible

commercial variety Bragg. Plots comprising 2 rows, 6.1 m

(20 feet) in length with 180 seeds per row, were treated

prior to planting with 561.4 kg/ha (500 Ibs/acre) 0-10-20








69
fertilizer and alachlor at 4.7 1/ha (2 qts/acre). All

plants were maintained insect-free with 2 applications of

methyl parathion 4E at 0.3 1/ha (0.25 pt./acre). The final

insecticide application was made 4 weeks prior to bioassays

of soybean foliage.

Induction methods. Field collected 5-6th instar

velvetbean caterpillar (VBC), Anticarsia gemmatalis Hubner,

larvae were placed in individual organdy leaf cages (15 x 25

cm) covering soybean trifoliates. Four larvae were confined

within each cage until ca. 50% of the contained foliage was

consumed (ca. 24 h). The cages were placed on all alternate

leaves of plants (10-15 leaf cages per plant) between the

R3-R5 phenophases (Fehr et al. 1971). Following larval

damage (48 h), the adjacent undamaged leaves were collected

and immediately stored in an ice chest (ca. 10C) and

brought back to the laboratory for feeding trials.

Uninfested, but similarly caged leaves from soybean plants

similarly treated were included as controls.

Rearing methods. Larvae (third instar) of the

velvetbean caterpillar, the soybean looper (SBL)

Pseudoplusia includes (Walker), and green cloverworm (GCW)

Plathypena scabra (F.) were field collected at the NFREC

from untreated soybean (var. Kirby). Individual larvae were

reared in petri dishes (15 x 150 mm) on the leaves of

undamaged or damaged soybean plants under standardized

environmental conditions (27"C, 50 10% RH and 14:10 L:D).








70

All foliage was presoaked for 30 min. in a 1% bleach (sodium

hypochlorite) solution. Leaf petioles were inserted into

stoppered water vials to maintain leaf turgidity and

replaced once during the first three days of the experiment,

and daily thereafter until larval pupation. Larval dry

weight (dw) gain and relative growth rates (RGR) were

calculated according to a gravimetric method (Waldbauer

1968, Slansky and Scriber 1985). All weights were obtained

using an electronic balance (Mettler AC-100, 0.1 mg).

Similar field collected larvae (n=20) were weighed fresh and

again after oven drying (60*C for 48 h) to estimate the

initial percent dw.


Experiment 2a. Preliminary Extract Methods

Soybean foliage from a commercial variety (Kirby, R-3

plant phenological stage) grown at the NFREC was hand

harvested, frozen (-10C) and extracted with 95% ethanol for

3 days followed by a series of non-polar to polar organic

solvents (Fig. 3-1). The ethanol extract was rotoevaporated

(60*C) to dryness and the residue partitioned between water

and petroleum ether. Following removal of residual

petroleum ether, the water fraction was partitioned between

ethyl acetate and water. The water fraction was

rotoevaporated to remove residual ethyl acetate and passed

through an Amberlite XAD-7 non-ionic (Aldrich Co.) column

(4.2 x 29.0 cm) previously washed with 200 ml of deionized








Fig. 3-1. Soybean foliage extract partition scheme.
EtOH=ethanol, Pet. Ether=petroleum ether, EtoAc=ethyl
acetate, XAD-7=Amberlite XAD-7 non-ionic resin,
Phenolics=phenolic compounds from the water fraction, Hyd.
Aqueous=hydrolyzed aqueous fraction.

























Soy leaves in EtOHI


Pet. Ether


I H 7.0
Water PH 7.0


XAD-7


Hyd.








73

water. Material adhering to the column was washed with

deionized water (200 ml) and eluted with 70% methanol (200

ml). The wash water lacked UV absorbance (265 and 320 nm)

and was thus discarded. The methanol eluate was

rotoevaporated to dryness and a sample (1 g, dw) was

hydrolyzed (to cleave sugars from glycosides) by refluxing

for 30 min ca. 600C followed by extraction with ethyl

acetate. All fractions (petroleum ether, ethyl acetate,

water extractables (phenolics) and hydrolyzed aqueous

extractables (after extraction with ethyl acetate) were

rotoevaporated to dryness and refrigerated (100C).

Diet preparation. Each dry extract fraction was

dissolved in methanol, incorporated into a 5% fresh weight

(fw) agar-water solution at 1, 2 and 5% (fw) and applied

with a fine brush to the upper surfaces of susceptible

soybean leaves (var. Kirby). Approximately 2 ml of each

treatment solution was applied to individual leaves. A

methanol and agar (5% fw) control was also included. All

treated leaves (15 replicates) were air dried and petioles

were inserted into a water vial and placed in petri dishes

(150 x 15 mm).

Larval rearing. Field collected fourth instar soybean

looper larvae were weighed and reared on treated leaves for

48 h under standard environmental conditions (Experiment 1).

Larval percent dw was determined as described above

(Experiment 1) and final larval dw and feces production were








74

recorded. Treatments and blocks were arranged in a

randomized complete block experimental design and the data

were analyzed by a two-way analysis of variance (ANOVA),

where extract fraction and concentration constituted the

main effects. Means were separated with a least square

means test, maintaining the experimentwise error rate at 5%

with the Bonferroni inequality (Sokal and Rohlf 1981), using

SAS/PC (SAS Institute, Inc. 1987).


Experiment 2b. Refined Soybean Foliar Extraction

In contrast with the previous procedure that

fractionated the water layer, this procedure fractionated

the organic solvent extractables (Fig. 3-2) as activity was

previously detected in the organic solvent fraction (see

Experiment 2a results) and ethyl acetate extractables

contained activity (Chiang et al 1987). Soybean foliage

(var. Kirby) was collected and stored as described in

Experiment 2a. The leaves were extracted for 3 days with

ethanol (95%) and the solvent was removed as in Experiment

2a. These ethanol extractables were partitioned between

water and ethyl acetate. The ethyl acetate fraction was

further partitioned between petroleum ether and 80%

methanol. The methanol fraction was partitioned between

benzene and 50% methanol. The aqueous extractables were

prepared as in Experiment 2a. All fractions were dried and

stored as described previously (Experiment 2a).








Fig. 3-2. Soybean foliage extract partition scheme.
EtOH=ethanol, Pet. Ether=petroleum ether, MeOH=methanol,
EtoAc=ethyl acetate, XAD-7=Amberlite XAD-7 non-ionic resin,
Phenolics=phenolic compounds from the water fraction, Hyd.
Aqueous=hydrolyzed aqueous fraction.







76
















Soy leaves in EtOH

EtoAc Latr pH 7.0
XAD-7
et. Ethe MeOH 80%Phen
-Phenolics
Benzenel MeOH 50% olysis

IHyd. Aqueousl








77

Each solvent-free fraction was mixed with cellulose

(alphacel, ICN Biochemicals, Inc.) in an excess of acetone,

rotoevaporated to dryness. The cellulose was incorporated

into an artificial diet at 5% fw (Greene et al. 1976).

Treatment diets consisted of fraction concentrations of 0.5

and 1% dw except the methanol fraction due to a shortage of

material (0.1 and 0.5% dw). Additionally, the active

fraction from Experiment 2a (petroleum ether) was included.

A control diet consisted of cellulose (5% fw) and acetone

mixed and rotoevaporated before combining with the standard

artificial diet. Diet samples (500 mg, n=15) were weighed

fresh, oven dried and reweighed to estimate their initial

percent dry matter.

Insects and rearing methods. Third instar velvetbean

caterpillar larvae (obtained as eggs from the USDA-ARS

Insect Attractants and Basic Biology Laboratory,

Gainesville, Florida) were weighed and reared at standard

laboratory conditions as previously described (Experiment 1)

for 5 d in inverted 30 ml (1 oz) plastic cups lined with a 2

ml layer of Gelcarin HWG (1.5% fw, Marine Colloids, Inc)

along the cup top. Larval consumption, growth, and feeding

efficiencies were calculated according to a gravimetric

technique (Waldbauer 1968, Slansky and Scriber 1985). The

initial larval percent dw was estimated (n=20) as described

previously (Experiment 1). Insect feces were removed daily

after the second day of the experiment and larvae were refed








78

as needed. Larval feces and uneaten diet were collected,

dried and weighed. All data (randomized complete block

experimental design) were analyzed with an ANOVA followed by

the Tukey-Kramer test (Sokal and Rohlf 1981) using SAS/PC

(SAS Institute, Inc., 1987), except for the mortality data

which were analyzed with a G-test (Zar 1984).


Experiments 3a-3e. Influence of Greenhouse Induction of
Soybean Defenses on Herbivore Performance

A series of experiments was conducted evaluating the

activity of soybean foliar extract fractions from mite-free

and mite-damaged foliage on larval mortality and RGR

(calculated as in Experiment 2a). Several non-adapted and

adapted soybean folivors were included, the velvetbean

caterpillar, fall armyworm (FAW), Spodoptera fruqiperda (J.

E. Smith), the corn earworm (CEW) Heliothis zea (Boddie),

the tobacco budworm (TBW) H. virescens F. and the cabbage

looper (CL) Trichoplusia ni (Hubner), all obtained as eggs

(from the USDA-ARS, Experiment 2b) and reared to the third

instar before feeding on the treatment diets. With the

exception of the velvetbean caterpillar and the corn earworm

(primarily a pod-feeder), these species are not major

soybean foliage pests (Kogan and Turnipseed 1987) and thus

may not be as well adapted to the allelochemicals contained

in soybean foliage. All studies involved rearing larvae for

3 d on the standard artificial diet (Experiment 2b) that

including various soybean foliar extracts. Mortality data








79

were analyzed either by a probit analysis (Finney 1971) or a

G-test (Zar 1984), whereas the RGR data (randomized complete

block experimental design) were analyzed by ANOVA and means

were separated with the Tukey-Kramer test using SAS/PC (SAS

Institute, Inc., 1987).

Agronomic methods. The influence of piercing-sucking

herbivore damage on the induction of soybean defenses was

studied by evaluating herbivore mortality while feeding on

diet containing extract fractions of the treatment foliage.

Three soybean lines were evaluated, the two previously

mentioned lines (Bragg and PI229358, Experiment 1), plus the

advanced breeding line D75-10169 (whose pedigree comprises

Govan x (Bragg x PI229358) (Hartwig and Edwards 1985).

Plants were greenhouse grown (275"C, 5010% RH), three to a

25.4 cm (10 in) plastic pot, in a sterilized soil mixture

(soil, sand, vermiculite 2:1:1) until the R3-R4 phenophase

(Fehr et al. 1971). Pots were fertilized (0-10-20) monthly

until foliage harvest.

Induction of defenses. Soybean plants were infested

with twospotted spider mites (Tetranychus urticae Koch) by

moving plants into infested areas of the greenhouse. Only

plants reaching high mite densities (where the webs covered

nearly all the foliage) were used. Plants that did not

support abundant mite populations were discarded. However,

none of the Bragg nor PI229358 plants were successfully

infested. Plants, designated as controls, were kept mite-








80

free by isolating them from the infested plants. Thus, the

plant treatments consisted of mite-free susceptible Bragg,

mite-free resistant PI229358, and mite-free and mite-damaged

resistant D75-10169.

Extraction of foliage fractions. Leaves were hand-

harvested, stored and extracted as described previously

(Experiment 2b, Fig. 3-2). Because results in Experiment 2b

indicated the benzene fraction was active, this fraction

from the mite-free D75-10169 treatment was further purified

by sequentially eluting with 5:95% acetone:benzene followed

by 5:95% methanol:benzene on a silica gel (100-200 mesh,

Fisher Scientific) column (15 x 225 mm). The eluate was

separated into two fractions, the acetone and the methanol

extractables. Each fraction included several bands that

were monitored with thin layer chromatography (TLC,

Kieselgel 60 HF, EM Science) using solvent mixtures ranging

in polarity from 1:99% acetone:benzene to 5:95%

methanol:chloroform. All fractions were rotoevaporated to

dryness and refrigerated (100C).

The active fractions (reported in Results) were further

purified by ionization with 0.5N NaOH (pH=ll-12) followed by

extraction with chloroform. The ionizable fraction was

neutralized (pH=7) with lN HC1. The extracts were divided

into the ionizable and non-ionizable fractions and developed

with TLC (5:95% methanol: chloroform). The developed bands

from the extracts were compared with the isoflavone








81

standards genistein, daidzein (both provided by S. K.

Chattopadhyay, Department of Medicinal Chemistry, College of

Pharmacy, University of Florida), coumestrol (Eastman Kodak

Co.) rotenone and biochanin A and the flavones rutin and

quercetin (Sigma Co.). Following development, flavonoid

spots were viewed under short UV light (254 nm), sprayed

with ferric chloride (5% dw dissolved in methanol) and

heated on a hot plate set at low (75*C) (Krebs et al. 1969)

for phenolic visualization (Egger 1969). The Rf values were

calculated by dividing the distance between the origin and

the center of each flavonoid spot by the distance between

the origin and the solvent front. Final purification was

achieved by scraping individual bands developed with

preparative TLC (25 x 25 cm, Kieselgel 60 HF, EM Science)

run 3-5 times in 5:95% acetone:benzene.


Experiment 3a. Influence of Greenhouse Induction of Soybean
Defense on Velvetbean Caterpillar Mortality

Treatments included the benzene, acetone and methanol

fractions mixed with the standard artificial diet

(Experiment 2b). The benzene and acetone fractions were

mixed at concentrations ranging from 0.05-2% dw. The

methanol fraction was incorporated at only the 2% dw

concentration. The standard solvent + cellulose control

diet was also included (Experiment 2b). Velvetbean

caterpillar larval mortality was recorded.








82

Experiment 3b. Effect of Induced Resistance on Non-adapted
and Adapted Soybean Herbivores

This study was designed to determine the influence of

extracts from mite-free and mite-damaged treatments on non-

adapted and adapted soybean herbivores. The benzene

fractions from the mite-free and mite-damaged D75-10169

soybean line, were incorporated into the standard artificial

diet (0.5% dw) and were fed to third instar larvae (20

replicates) of three noctuid species, the corn earworm, the

fall armyworm and the tobacco budworm. Larval RGR data were

calculated.


Experiment 3c. Sensitivity of Three Noctuid Species to the
Non-benzene Soybean Foliar Extract Fractions

The previous two experiments (3a and 3b) only evaluated

activity in the benzene extract fractions (the active

fraction as demonstrated by the results of Experiment 2b)

against non-adapted and adapted herbivore species. However,

it is possible that an induced allelochemical is extracted

in another solvent. Thus, the larvae of three noctuid

species, velvetbean caterpillar, fall armyworm and cabbage

looper were fed (20 replicates) artificial diets containing

the non-benzene fractions (at 0.5 and 1% dw) from Bragg and

mite-free and mite-damaged D75-10169 foliage. Larval RGR

data were calculated.








83

Experiment 3d. Effect of the Petroleum Ether Extract
Fraction on RGR of Non-adapted and Adapted Soybean
Herbivores

This experiment was designed to confirm the reduction

in RGR caused by the petroleum ether extract. Results from

the previous study indicated that petroleum ether

extractables of the mite-damaged D75-10169 treatment reduced

herbivore RGR values significantly more than the comparable

mite-free treatment. The same three herbivore species

tested in Experiment 3c (velvetbean caterpillar, fall

armyworm, cabbage looper) were fed artificial diets

containing the petroleum ether extract fraction (at 0.5 and

1% dw) from the foliage of Bragg, mite-free and mite-damaged

D75-10169, and an untreated control (20 replicates).


Experiment 3e. Purification of the Active Fractions

The Bragg benzene extract fraction (from Experiment 3a)

was washed with sequentially increased concentrations of

acetone (1:99% to 15:85% acetone:benzene) eluting seven

fractions by following major bands on TLC (as described in

Experiment 3a). These solvent-free fractions were

incorporated into the standard artificial diet (at 0.5 and

1% dw) and fed to larvae (10 replicates) of 5 noctuid

species; the velvetbean caterpillar, fall armyworm, cabbage

looper, tobacco budworm and the corn earworm. Insufficient

material was available to test fraction 21-23 at both

levels, and thus it was tested at only the 0.5% (dw)

concentration.












Results

Experiment 1. Induction of Susceptible and Resistant Soybean

Preliminary results of the effects of induced

resistance suggest changes in larval performance when fed

foliage from velvetbean caterpillar damaged plants.

However, nearly half of the field collected larvae died from

an outbreak of the entomopathogen Nomuraea rileyi (Farlow)

Samson, and thus, no statistical analyses were conducted on

these data. However, the results suggest that larval

relative growth rates [RGR=biomass dw gain (mg)/average

caterpillar dw (mg)/ developmental time (d)] may have

increased for both the velvetbean caterpillar and the

soybean looper when fed the undamaged resistant (PI) foliage

compared with the undamaged susceptible (Bragg) variety

(Table 3-1). Additionally, velvetbean caterpillar and

soybean looper larvae had reduced RGR values when fed

foliage from damaged plants of resistant foliage compared

with foliage from undamaged resistant plants. However,

velvetbean caterpillar, soybean looper and green cloverworm

larval RGR values increased when fed foliage from damaged

plants of the susceptible Bragg cultivar. Furthermore, the

RGR of green cloverworm larvae decreased dramatically on the

resistant compared to the susceptible line. Further studies

are required to determine the significance (statistical

and/or biological) of these differences.












Table 3-1. Average (se) relative growth rates (RGR) of
velvetbean caterpillar, soybean looper and greenclover
worm larvae fed field grown undamaged or damaged
soybean foliage. Damaged treatments consisted of
plants that had prior (48 h) velvetbean caterpillar
larval damage. See Results (Experiment 1) for RGR
formula.



Line1 Spp2 Treat3 n RGR se


BG VBC CHK 3 0.45 (0.16)
BG VBC DAM 3 0.75 (0.47)
PI VBC CHK 2 0.75 (0.19)
PI VBC DAM 3 0.56 (0.12)
BG SBL CHK 7 0.69 (0.13)
BG SBL DAM 8 0.93 (0.21)
PI SBL CHK 5 0.79 (0.24)
PI SBL DAM 6 0.68 (0.10)
BG GCW CHK 4 0.58 (0.40)
BG GCW DAM 3 0.83 (0.59)
PI GCW CHK 3 0.06 (0.04)
PI GCW DAM 4 0.06 (0.50)


1 Lines: BG=Bragg (susceptible); PI=PI229358 (resistant).
2Species: VBC=velvetbean caterpillar; SBL=soybean looper;
GCW=green cloverworm.
Damage treatments: CHK=undamaged plant; DAM=VBC damaged
plant.


Experiment 2a. Preliminary Extract Methods

The average larval biomass gain and feces production

were significantly influenced only by the extract fraction

(e.g., control, water, etc.) but not by the concentration of

each fraction (P=0.40). Thus, concentration effects were

considered as replicates of each extract fraction. Biomass

gain and feces production for larvae fed leaves treated with

the petroleum ether fraction were significantly reduced (39








86

and 36%, respectively) compared with the control leaves

(Table 3-2). Biomass gain of larvae feeding on the water

fraction was also significantly greater than on the

petroleum ether fraction. Furthermore, larvae feeding on

the petroleum ether fraction produced less feces than the

larvae feeding on the other treatments, possibly due to

reduced consumption.





Table 3-2. Average (se) biomass gain (dry weight=dw) and
feces production by third instar soybean looper larvae
fed field grown soybean foliage treated with soybean
(Kirby) foliar extract fractions.



Fraction' n Biomass gain Feces
(dw, mg) (dw, mg)


Control 15 55.8 (4.7) a2 194.8 (26.1) a
Water 45 49.8 (5.4) a 213.7 (15.3) a
Water (Hydrolyzed) 45 45.6 (4.0) ab 173.0 (14.5) a
Ethyl acetate 45 44.2 (5.4) ab 201.8 (25.2) a
Petroleum ether 45 33.8 (4.7) b 124.6 (16.8) b


SSee text for descriptions.
2Means followed by the same letter are not significantly
different according to a least square means test
(P=0.05).



Experiment 2b. Refined Soybean Foliar Extraction

Survival of velvetbean caterpillar larvae feeding on

the soybean foliar extract fractions was lowest on the

benzene 1% (dw) diet (13%) compared with the other








87

treatments (Table 3-3). The benzene 0.5% and methanol 0.1%

treatments were deleted from the G-test analysis because of

zero mortality. Additionally, the RGR, efficiency of

digestion and absorption of dw diet [AD=100*(ingestion-

feces)/ingestion (all dw, mg)], efficiency of conversion of

digested food into insect tissue [ECD=100*biomass

gain/(ingestion-feces) (all dw, mg)] and efficiency of

converting ingested food into insect tissues

[ECI=100*biomass gain/ingestion (all dw, mg)] values were

reduced for the two survivors of the benzene extract (1%).

The petroleum ether fractions, isolated from the preliminary

study (PE1), are included here to compare with the toxicity

of the more refined extract fractions.


Experiment 3a. Influence of Greenhouse Induction of Soybean
Defense on Velvetbean Caterpillar Mortality

Percent mortality data suggest that mortality increased

with greater concentrations of each benzene extract,

regardless of the undamaged or damaged condition of the

plants (Table 3-4). Because zero observations occurred in

many cells (18 of 29 cells where mortality was either 0 or

100%), the fiducial limits of the probit analysis (which

uses a X2 analysis) could not be confidently (P=0.05)

estimated. However, concentration data for each plant

extract could be combined to produce an overall percent

mortality value that was analyzed with a G-test (Zar 1984).











()






(0 4 +H
.-4
H1U C





( 0 0 W.

(U 0- 4




a o-r :> o
CO-H,- C
0 0 0

- u ) (0
W-44i9




SC -- I> (0
W.Q 0 ) Q)
*-i 4J ii









0. -i r-Ho 4.
I- 4 0
() *> O)




(0 44 IP 4













(1 0 4.) -4
' CU) (




,4 0 0 a W



)U ) H


+100 U4 (
r4 4)- Q) 4)
M4. Cm 0 m
5-i-i H -H (0






.H )i 4 -1 .
m 0C-m r
p-H 4 O) 0 4
4-10 .- 0



0 W r C V




p44 P4

-m -i
,. (- 4 1
(U 03II 04-





















a) 4 -4 (U
a) 5-ia U 5- 4




0 a)U 0a a)










Hc a



000 0-00 rU
aQs e < u+
* () r 0 ^1r-






EA


U
N
>
O-
U
0
I U
SUC
0 0
4.3


coo c Occo ooo
-4OOOOOOOOOOr-1

00000000000
+1+1+I +1 +I+1+1+1+1+1 +1 +


000


000


mm~o
Sm 0) N' 040
CO r-H CO LO co

+1 +1 +1 +1 +1
'- 1 %


wDO CO0
%D c%


cn0 D0 %0 LO
1.in o, -. 0( r,0

+1 +1 +1 +1 +1 +1


me
r41.
cn (1)


co I %

C' CO 9
C%3 (M


co o

000


+1 +1 +1 +1 +1 r-4
u v u u+1


\ oo
doy


0r-l
*
ino
LnA


0 100- 0r

+1 +1 +1 +1 +1


oNT


r- ,


M .X,419 1..rI II.;1. ;I I I

+-1 +4 + + co+ +r +r +- +1 + +1
+1 A A A A A A 1 A

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1


0 co
0CO

m> V
^1" *I


N CONl
r O n

.. 11 o


r-o
"a" in


AA AA AAA A A


D CA LO CO a)
0',000

00000
+1 +1 +1 +1 +1



) r-) r- rLO
-4-4-4-4-


Sr- co
000
00ooo
+1 +1 +1

co
r- o co
-4 -1-


r-lO 0

000
+1 +1 +1

0 C CO
CO COC
l r 1r


r-ommmr-ooomm

in rno r-i c.-
L- A I r- A L I -


H
MOW
Pt04


ERM MMM
04 m z 3 co


4J 0
1 4












H OI
-i al ,
0 0
(U








































O n
4 0 r4

-4 0 )




II a)
i) 4$
M u 4.3
r-











t101
a) C

0 a) l
4) 4-4 r-i
6 4-4 4)
5rl '0 C*
4 a)


^ (

o u e^

0 0U

) ) 4 -
*3 -4 10- 4J
4) o M

0) V4
o .4 v .


.43 4J W )

a) *a) +J M


-4 LO >4


. ) 0(U)




o ro- 0
II r4 11 44 4-4

E- 04 a)



- II c M
0 N > +II-



- U) a)












Table 3-4. Velvetbean caterpillar third instar larval
mortality fed soybean foliar benzene fractions
incorporated into artificial diet. Foliar extracts:
BG=Bragg (susceptible); PI=PI229358 (resistant);
DIND=D75-10169 mite-damaged; DCON=D75-10169 mite-free;
ACE=acetone fraction of the BG extract.



Concen- BG PI DIND DCON ACE
tration
(% dw) n % n % n % n % n %
mort mort mort mort mort


0.05 20 5 20 0 20 5 20 0 nt nt1
0.10 10 0 30 3 30 3 30 0 10 10
0.25 10 60 10 30 10 0 10 0 10 80
0.50 23 100 23 100 24 100 25 96 10 90
1.00 35 43 20 100 14 100 14 100 20 100
2.00 9 100 13 100 12 100 12 100 20 100


Overall2
% mortality 50 52 47 45 83


nt=not tested.
2 Overall mortality was significant according to a G-test
(X=17.7, df=4, P<0.005).


Although significant differences in mortality occurred

among the plant extracts, distinction among the treatments

is not possible with this analysis; however, the data

suggest that differences in mortality among the mite-damaged

(DIND), mite-free (DCON & PI) resistant treatments and the

susceptible variety (Bragg) were minimal (45-52%), whereas

the majority of mortality (83%) occurred when larvae fed on

the refined acetone treatment. These data indicate that the

activity contained in the benzene fraction is acetone