._ |. flywhvv

 

4 .4. ...,.
{5;}. an" :
.. 3.; . . 1",.-.

:rvfi frat:
Anr3~ .w-v a;

«Inn ('53:.
"5:.

r»- .5351}
.‘nm

:23:— '3'. ’
Sf; ”’7'?

v.

' n. “k" V ~ ' I" ‘~ :6.
r flannwm... .L..,

r
. , - ' . v" ‘
U
Mrf-fi ' ' ~.- .1' ”Li
' ' r, ' r ' .1719:
1"”.

H-

.4_

’ y. _ ”4'32.
'I‘IU’HX‘LW“ greas-
"441m”, "

":33

3.}\

1".“

0-1,
32!}.

N, .1, |;‘

'1 ’71:}!
.1...
z. .

\
JV.”
f. an.
.1:‘

f
.3" 44"(
Ira.

m. .;.- 4......
may. {« u

’f";"— ’0'. ' V'V‘0I’;
mix-z“ ' 'vh‘ml'fimzué
' I 0-. ' éf'r:~"".,’w

.‘.,‘.- .. 7. wk:

Liv-“x"

J!
“Dr-m;
"‘ {'5‘ 5V “Ural.” 7-31
J

 

filllllll lllllllllll WM.

:. ~, a. 3 1293 00992

@010 (215674

 

LIBRARY {
Michigan State
University

 

 

 

This is to certify that the

thesis entitled

CONSEQUENCES OF INTRASPECIFIC HOST QUALITY ON THE
ECOLOGY OF TWO SPECIALIST PHYTOPHAGES, MANDUCA
QUINQUEMACULATA AND MANDUCA SEXTA,‘

IEEPIOOP I ERA: SPHINGIOAE)

presented by

Janice Lynne Bossart

has been accepted towards fulfillment
of the requirements for

M . 3. degree in ENTOMOLOGY

 

 

QM [rag

Major professor

 

Date g‘gq “gg

 

0-7 639 MS U is an Affirmative Action/Equal Opportunity Institution

 

MSU

LIBRARIES
__.

 

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES wilI
be charged if book is
returned after the date
stamped below.

 

 

.1331; 3133

 

 

CONSEQUENCES OF INTRASPECIFIC HOST QUALITY ON THE ECOLOGY
OF TWO SPECIALIST PHYTOPHAGES, HANQQQA QUINQUEMACULATA AND
HANDUCA SEXTA, (LEPIDOPTERA:SPHINGIDAE)

 

BY

Janice Lynne Bossart

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Entomology

1987

ABSTRACT

CONSEQUENCES OF INTRASPECIFIC HOST QUALITY ON THE ECOLOGY
OF TWO SPECIALIST PHYTOPHAGES, MANDUCA QUINQUEMACULATA AND
EAEDUCA SEXTA, (LEPIDOPTERA: SPHINGIDAE)

BY

Janice Lynne Bossart

Field and laboratory experiments investigated the
affect of host plant quality on conspecific host discrimina-
tion by two specialist phytophages, Manduca guinggemaculata
and nanduca sexta, the tomato and tobacco hornworm moths
(Lepidoptera: Sphingidae), and development of n; guinggg:
maculata larvae and adults. Results of these experiments
were utilized to examine the hypothesis that female host
preference is correlated with subsequent larval performance.

Higher rates of fertilization produced plants that
conferred the highest fitness on larvae reared on foliage
from these plants. Increased food efficiency, growth,
survival and fecundity were correlated with mean percent
total plant nitrogen and soluble nitrates. In addition,
moth response, measured as the percent of total eggs deposi-
ted and the percent of plants with eggs, was highest on
plants treated at higher rates of nitrogen fertilization.
Finally, plants most preferred by moths permitted larvae to

obtain the highest overall fitness.

This thesis is dedicated to my daughter Allyson.
Through her support, patience and cooperation, I have been
able to obtain goals that would have otherwise been impos-
sible. Thank you Allyson, I will always remember how much

you gave up to make graduate research possible for me.

iii

ACKNOWLEDGEMENTS

First, I would like to express my gratitude to the
members of my graduate committee, Drs. Stuart H. Gage,
William J. Mattson, James R. Miller and Irvin E. Widders for
their assistance and comments throughout this study. I
would especially like to acknowledge Dr. Gage for providing
the monetary support that enabled me to pursue this par-
ticular study. Secondly, I must extend my appreciation to
Mike Jackson and the other folks at the Tobacco Research
Station in Oxford, NC. Without the assistance of these
people and the prompt shipment of hornworm eggs from their
lab, this study would not have been possible. Finally, I
must acknowledge Bruce Birr, of the USDA Forest Service, and
John Dahl and Roger Deason, of the Soil Testing Laboratory
at Michigan State University, all of whom assisted in the

chemical analysis of foliage collected during this study.

iv

PREFACE

This thesis addresses the interaction between host
plant physiology/quality as it relates to total plant nitro-
gen and soluble nitrates and, subsequently, its affect on
herbivore behavior and development. The analysis and dis-
cussion of this research are preceeded by a review of the
literature presenting information relevant to the investiga-
tion of hypotheses regarding herbivore response to variable
host quality. The literature review begins with a summary
of the biology of nanduca spy. and the factors known to
induce host selection and ovipositional responses in these
insects. It then continues with a discussion of the plant
components generally thought to influence host selection in
female herbivores and subsequent development of their off-
spring. The final section of the review presents a summary
of the literature pertaining to the affect of nitrogen
availability on plant biochemistry and herbivore behavior
and fitness.

Results of the research are presented as three sepa-
rate manuscripts. All manuscripts follow the format guide-
lines set forth in Qegologia. The first paper, "The sea-
sonal occurrence and development of nanduca guinggemaculata
(Haworth) and nandggg sexta (L) (Lepidoptera: Sphingidae) in

V

southwestern Michigan", describes the biology and relative
abundance of nanduca sppy, at the Kellogg Biological Station
in Barry Co., Michigan, in relation to degree day accumula-
tion and investigates differences between the two species.
The second manuscript, "Nutritional consequences of variable
plant nitrogen: Experiments with plant fertilization and
nagguga ggiggggmagglata", investigates the influence of food
quality on larval development and survival and subsequent
adult emergence and fecundity. The final manuscript, "Con-
specific host selection and subsequent larval performance in
a specialist phytophage (Lepidoptera: Sphingidae): the
influence of plant quality on herbivore behavior and physi-
ology”, examines associations between tomato plant quality/
physiology, host preference and larval development and
survival. This manuscript also investigates the hypothesis
that female moths prefer to oviposit on plants conferring
the highest overall fitness on their larvae.

The concluding section of the thesis is devoted to a
discussion of the overall conclusions and implications of
the project. Reference is made to results of the study that
were especially interesting to the researcher and those

aspects thought to warrant future investigation.

vi

TABLE OF CONTENTS

LIST OF TABLES ix
LIST OF FIGURES xi
REVIEW OF LITERATURE 1
History and Ecology of Manduca spp. 1
Overview of Host Selection 6
Chemical Components of Host-Selection 8

Secondary Chemicals 9

Nutritional Factors 10

Nitrogen and Herbivory 12

Direct Effects on Insect Nutrition 12

Indirect Effects on Insect Nutrition 14

Nitrogen, Plant Water and Herbivory 14

Nitrogen, Alkaloids and Herbivory 15

Literature Review Summary 17
References to Literature Review 19

MANUSCRIPT I. The seasonal occurrence and development

of mg: seats and Kandace W (LePidoptera=

Sphingidae) in southwestern Michigan.

Abstract 28
Introduction 29
Materials and Methods 30
Results 34
Discussion 43
Conclusions 47
References 48

vii

MANUSCRIPT II. Nutritional consequences of variable plant
nitrogen: Experiments with plant fertilization and Manduca

ac .

Abstract 51
Introduction 51
Materials and Methods 54
Results 60
Discussion 82
Conclusions 89
References 91

MANUSCRIPT III. Conspecific host selection and subsequent
larval performance in specialist phytophages: the influence
of plant quality on herbivore behavior and physiology.

Abstract 97
Introduction 97
Materials and Methods 100
Results 108
Discussion 133
Conclusions 143
References 145
POSTSCRIPT 152

viii

LIST OF TABLES

MANUSCRIPT I

Table 1. Relative occurrence of M; guinggemaculata and M;
sexta larvae at different sampling times during each year of
the study. Each sampling period consists of approximately 5
separate sampling dates.

Table 2. Means and standard errors for development times of
field and lab cohorts of M. guinguemaculata and fl; sexta
larvae during 1986 study. N = 17 for lab cohorts. N = 13
and 15 for M; guinguemaculata and ML sexta field cohorts
respectively. Means with same letter not significant at
p=.05

 

 

MANUSCRIPT II

Table 1. Means and standard errors for percent water con-
tent, total plant nitrogen and soluble nitrates calculated
during each separate aspect of the study. These values were
utilized in all correlation analyses requiring data on the
concentration of particular plant substances.

Table 2. Correlation of mean percent dry weight plant
nitrogen and soluble nitrates with larval performance param-
eters calculated from the General Growth Study. Top row of
numbers show Pearson's coefficient of correlation. Numbers
in parentheses demonstrate level of significance.

Table 3. Analysis of variance of larval performance in-
dices. Larvae reared on foliage treated with four levels of
nitrogen fertilizer. Significance of F-statistics (n.s. =
not significant at p = .05). See Materials and Methods for
explanation of statistics and definition of indices.

Table 4. Means and standard errors of larval performance
indices. Larvae reared on foliage treated with four levels
of nitrogen fertilizer. Means with the same letter are not
significantly different at p = .05 (SNK). See Materials and
Methods for definitions of indices and explanation of sta-
tistics.

ix

Table 5. Correlation of mean percent water content, dry
weight nitrogen and soluble nitrates with larval performance
parameters calculated from the Detailed Growth Study. Top
row of numbers show Pearson's coefficient of correlation.
Numbers in parentheses demonstrate level of significance.
See text for definition of indices.

MANUSCRIPT III

Table 1. Means and standard errors for percent water con-
tent, total plant nitrogen and soluble nitrates calculated
during each separate aspect of the study. These values were
utilized in all correlation analyses requiring data on the
concentration of particular plant substances.

Table 2. Distribtuion of eggs on sampled plants. Data

separated by treatment and year. The top row of numbers,
within each treatment, represents the frequency of occur-
rence and the bottom row represents the percent of total.

Table 3. Total number of hornworm eggs sampled during each
year of the preference study. The number of eggs found at

each nitrogen level was converted to a percent of the total
eggs found during both years. See text for explanation of

calculations. N = 357 and 576 total plants sampled in 1985
and 1986 respectively.

Table 4. Means (1 standards errors) for all measured param—
eters of larval fitness at each level of nitrogen fertilizer
applied.

Table 5. Linear regression statistics of larval fitness
study. Numbers depict R2 value and probability of a greater
F-stastic (n.s. = not significant at p = .05).

Table 6. Correlation of mean percent dry weight plant
nitrogen and soluble nitrates with measured parameters of
insect performance. Top row of numbers show Pearson's
coefficients of correlation. Numbers in parentheses demon-
strate level of significance (n.s. = not significant at

p = .05).

Table 7. Correlation of female preference with larval fit-
ness. Top rows of numbers show Pearson's correlation coef-
ficients. Numbers in parentheses demonstrate level of sig-
nificance (n.s. = not significant at p = .05).

LIST OF FIGURES

MANUSCRIPT I

Figure 1. Seasonal trends of Manduga s99; (ML guingue:
maculata and u; sexta) egg and larval means per half plant.
Data combined from field and orchard site. Circles = eggs,
stars = larvae. In 1985, 60 half plants were sampled per
day. In 1986, 96 half plants were sampled per day until
August 25, at which time, 48 half plants were sampled.

Figure 2. Seasonal egg means of ML guinguemaculata and M;
sexta. Circles = M; in emaculata, stars = M; sexta.

Data combined from field and orchard site. In 1985, 60 half
plants were sampled per day. In 1986, 96 half plants were
sampled per day until August 25, at which time, 48 half
plants were sampled.

 

Figure 3. Seasonal larval means of M; guinguemaculata and

M... seat..- Circles = n... W. stars = IL. sexta-
Data combined from field and orchard site. In 1985, 60 half

plants were sampled per day. In 1986, 96 half plants were
sampled per day until August 25, at which time, 48 half
plants were sampled.

Figure 4. Seasonal development of ML guinguemaculata and_ML
sexta field cohorts monitored from egg hatch to prepupa
during 1986 (N = 13 and 15 for M; guinguemaculata and M;
sexta respectively). Each data point represents the percent
of total larvae for a particular stage of development on
each monitoring day. Triangles = eggs, closed stars = let
and 2nd instars, open circles = 3rd and 4th instars, squares
s 5th instars, open stars = prepupa.

MANUSCRIPT II

Figure 1. Seasonal trends of percent dry weight nitrogen.
Foliage collected during 1986. Closed stars = 0 lbs. N,
open circles a 30 lbs. N, closed circles = 120 lbs. N and
open stars = 240 lbs. N applied per acre.

xi

Figure 2. Seasonal trends of percent soluble nitrates.
Foliage collected during 1986. Closed stars = 0 lbs. N,
open circles = 30 lbs. N, closed circles = 120 lbs. N and
open stars = 240 lbs. N applied per acre.

Figure 3. Mean percent plant water at each fertilization
level. Foliage collected during Detailed Performance Study.
Closed stars = 0 lbs. N, open circles = 30 lbs. N, closed
circles = 120 lbs. N and open stars = 240 lbs. N applied per
acre.

Figure 4. Linear regression of male pupal dry weights,
p < .0001, R2 = .40, N = 9 at each nitrogen level. Mean
values depicted by stars. See text for explanation of
statistics.

Figure 5. Linear regression of female pupal dry weights,
p < .0002, R2 = .34, N = 9 at each nitrogen level. Mean
values depicted by stars. See text for explanation of
statistics.

Figure 6. Percent successful emergence of adult males from
larvae reared on tomato foliage treated at four different
nitrogen levels. A linear regression of arcsine transformed
data produced a slepe that was not significantly different
from zero (R2 = .23, p < .46). See text for explanation of
statistics.

Figure 7. Percent successful emergence of adult females
from larvae reared on tomato foliage treated at four dif-
ferent nitrogen levels. A linear regression of arcsine
transformed data produced a slope significantly different
from zero (R2 = .97, p < .02). See text for explanation of
statistics.

Figure 8. Linear regression of female fecundity. R2 = .17,
p a .009. N = 10 at each nitrogen level. Mean values depic-
ted by stars.

Figure 9. Percent survival of larvae reared from egg hatch
to premolt fifth instar on foliage treated at four rates of
nitrogen fertilization. A linear regression of arcsine
transformed data determined that survival was associated
with fertilization level (R2 = .95, p = .02). N = 30 larvae
per treatment.

Figure 10. Means and standard errors for days to premolt
fourth instar. F = 14.04, p < .0001. N a 26, 25, 30 and 30
for 0, 30, 120 and 240 lbs. nitrogen applied/acre respec-
tively.

xii

Figure 11. Linear correlation between ECD and Approximate
Digestion. Pearson's correlation coefficient = -.90. N =
15, 18, 20 and 22 for 0, 30, 120 and 240 lbs. nitrogen
applied/ acre. See text for definition of indices. Closed
stars = 0 lbs. N, open circles = 30 lbs. N, closed circles =
120 lbs. N and open stars = 240 lbs. N applied/acre.

Figure 12. Means and standard errors for days to premolt
fifth instar. F = 9.52, p < .0001. N = 15, 18, 20 and 22
for 0, 30, 120 and 240 lbs. nitrogen applied/acre respec-
tively.

MANUSCRIPT III

Figure 1. Field design layout for conspecific host prefe-
rence study. Large dots represent dwarf apple trees, small
dots represent tomato plants. See Materials and Methods for
further explanation of design.

Figure 2. Plot design depicting fertilization of single
block. White circles represent treated tomato plants.
Black circles represent untreated tomato plants. See Mate-
rials and Methods for further explanation of design.

Figure 3. Seasonal trends of percent dry weight nitrogen.
Foliage collected during 1986. Closed stars = 0 lbs. N,
open circles = 30 lbs. N, closed circles = 120 lbs. N and
open stars = 240 lbs. N applied per acre.

Figure 4. Seasonal trends of percent soluble nitrates.
Foliage collected during 1986. Closed stars = 0 lbs. N,
open circles = 30 lbs. N, closed circles = 120 lbs. N and
open stars = 240 lbs. N applied per acre.

Figure 5. Plot of mean plant volume at each level of nitro-
gen fertilization. Y—axis represents mean plant volume, X-
axis represents date sampling conducted and Z-axis repre-
sents nitrogen treatment applied. N = 9 - 21 plants sampled
per treatment per day in 1985 and 24 plants sampled per
treatment per day in 1986. See Results for further explana-
tion of plots and calculation of plant volume.

Figure 6. Seasonal hornworm egg densities. Mean eggs
calculated per half plant. Reference bars denote period of
maximum egg deposition. Data combined from field and or-
chard site. N = 60 and 96 half plants sampled per day in
1985 and 1986 respectively.

xiii

Figure 7. Percent tomato plants with eggs at each nitrogen
level. Data combined from field and orchard site--1985
(heterogeneity Chi-square = 1.50, p > .50). A G-test analy-
sis for independence determined that the proportion of
plants with eggs is associated with the level of nitrogen
applied (G = 10.168, p < .025). See text for explanation of
statistics.

Figure 8. Percent tomato plants with eggs at each nitrogen
level. Data combined from field and orchard site--1986
(heterogeneity Chi-square = 3.03, p > .40). A G-test analy-
sis for independence determined that the proportion of
plants with eggs is independent of the level of nitrogen
applied (G = 5.02, p < .20). See text for explanation of
statistics.

Figure 9. Percent contribution of each nitrogen treatment
to total eggs found. Data fit to a quadratic regression, R2
= .78, p = .02. Circles = 1985 data, stars = 1986 data.

See text for explanation of statistics.

Figure 10. Ovipositional response to treated plants. Cor-
relation of percent nitrogen dry weight with total number of
eggs deposited (a) and proportion of plants with eggs (b) at
each treatment. Data collected during 1986 only. Total
number of eggs deposited = 108. N = 576 total half plants
sampled. Closed stars = 0 lbs. N, open circles = 30 lbs. N,
closed circles = 120 lbs. N and open stars = 240 lbs. N ap-
plied per acre.

Figure 11. Ovipositional response to treated plants. Cor-
relation of percent soluble nitrates with total number of
eggs deposited (a) and proportion of plants with eggs (b) at
each treatment. Data collected during 1986 only. Total
number of eggs deposited = 108. N = 576 total half plants
sampled. Closed stars = 0 lbs. N, open circles = 30 lbs. N,
closed circles = 120 lbs. N and open stars = 240 lbs. N
applied per acre.

Figure 12. Correlation between percent plants with eggs and
larval performance potential. Percent plants with eggs
based on 144 plants sampled per treatment. See text for
definition of larval performance potential. Closed stars =
0 lbs. N, open circles = 30 lbs. N, closed circles = 120
lbs. N and open stars = 240 lbs. N applied per acre.

xiv

REVIEW OF LITERATURE

to n c l of Mandu

Manduca sexta (L.), the tobacco hornworm, and Manduca
guinguemaculata (Haworth), the tomato hornworm, (Lepidop-
tera: Sphingidae) are widely distributed throughout North
America. The northern range for both species extends into
Southern Canada (Madden and Chamberlin 1945, Baumhover
1985). The southern range however, is more variable. For
M; guinggemaculata, the limit appears to be Northern Florida
and Southern Georgia (Madden and Chamberlin 1945), while M;
sexta has been documented as far south as Southern Brazil
(Madden and Chamberlin 1945, Baumhover 1985).

The adult moths are crepuscular, i.e., active prior
to 9 p.m. and just before dawn, and remain hidden during
daylight hours. These moths are large, strong fliers with a
wingspan stretching approximately 12 cm. At the beginning
of active periods, the moths can be found hovering over the
tubular blossoms of jimsonweed, petunia, evening primrose.
and hollyhock, feeding on the nectar from these flowers.

The female oviposits an average of 200 - 300 eggs during her
lifetime, the majority of which are generally deposited on
the upper 1/3 of the host plant. The spherical, light

1

2
yellowish-green eggs are usually laid singly on the lower
surface of a host leaf. Hatch occurs in approximately four
to five days (Madden and Chamberlin, 1945).
The larval stage generally consists of five instars
although, six instars are not unknown. The two species are

readily distinguishable by the third instar. ML sexta

 

larvae are characterized by a scarlet horn and a soft,
pubescent covering on their body. M; guingpemaculata have a
black horn and lack pubescence. Larvae are fully developed
in approximately 20 days and reach a final weight of about 8
- 9 grams and final length of approximately 75 - 85 mm
(Gilmore 1938). The mature larvae leave the plant, burrow
into the ground and hollow out an earthen cell (Gilmore
1938, Madden and Chamberlin 1945). Pupation occurs in the
soil at a depth of 10 - 12 centimeters (Madden and Cham-
berlin 1945). Eclosion takes place, during the night,
approximately 17 - 25 days after the onset of pupation.

Both species overwinter as diapausing pupae.

M; sexta and E; guinguemaculata are both host speci-
fic, utilizing only members of the Solanaceous family--
which includes, among others, tobacco, tomato, eggplant,
potato, jimsonweed and nightshade. Laboratory reared horn-
worms can be induced to feed on some non-solanaceous species
(= non-host species) (Waldbauer and Fraenkel 1961, Waldbauer
1962, 1964), thus suggesting that development can occur on a
much wider range of plants. It is interesting to note how-

ever, that natural larval populations are almost entirely

3
restricted to tobacco and tomato (Gilmore 1938, Yamamoto et.
al. 1969). Since it has already been determined that larvae
tend to remain on the same plant selected by the adult
female (McFadden 1968), it appears that egg deposition
occurrs almost exclusively on these tobacco and tomato
plants.

Captive adults are known to exhibit three types of
flight behavior (Yamamoto et. al. 1969). The first, disper-
sal flight, is characterized by being a highly erratic and
accelerated flight. Following this flight is a normal fora-
ging flight characterized by moths flying in a circular path
at mid-level in the cage. During this stage, objects may be
approached and landed upon. The third type of flight, low
altitude foraging flight, occurs at or near ground level.
Again, acceptable objects are approached and landed upon.
The first two types of behavior have also been noted during
experiments conducted during this study at the Kellogg
Biological Station (RES) in Hickory Corners, Michigan.

The moths approach objects of various physical attri-
butes nearly equally from a distance but do not land on them
equally. Instead, they tend to veer away from non-host
plants at distances of from 4 to 15 cm (Yamamoto et. al.
1969). This observation suggests that host searching is
comprised of random or directed random movements until the
host odor stimulus is encountered by the moth (a close-range
stimulant). Upon contacting this odor, the moth is stimu-

lated to approach the host and, if an appropriate "pattern"

4
of stimuli is present, is induced to land. Once a suitable
plant is located, the hovering female places her fore tarsi
on the upper surface of the ovipositional site and deposits
a single egg on the lower leaf surface (Sparks 1973).

Single-choice tests, conducted by Yamamoto and Fraenkel
(1960a), determined that moths highly prefer solanaceous
plants over non-solanaceous ones. In fact, moths failed to
oviposit at all on the non-solanaceous plants. In addition,
results from a two-choice test demonstrated that tomato
plants were significantly preferred over other choices,
including other solanaceous plants. These results led to
the hypothesis that at least two different chemical stimuli
influence host selection and ovipositional responses by the
moth. The first stimulus serves to orient the moth to the
plant (=orientational stimulus) and appears to be particular
for a given plant species. The other stimulus, an oviposi-
tional stimulus, appears to be common to all solanaceous
plants.

Further experimentation was conducted to elucidate the
chemical components that elicited these behaviors (Yamamoto
and Fraenkel 1960b). Ethanol extracts of tomato leaves,
which were known to contain the "ovipositional stimulus",
were tested for their ability to generate a response by the
moths. Since egg deposition occurred only in those areas in
which the ethanol extract had been sprayed, it was concluded
that this stimulus was perceived upon contact.

Steam distillants of leaves were examined to

5
determine if the stimulatory chemicals could be detected
through olfactory processes. Results demonstrated that
volatile components did, indeed, act as "attractants" and
served to orient the moth to the host. It was also deter-
mined that the steam distillant and ethanol extract used
together elicited a synergistic effect, dramatically in-
creasing the number of eggs deposited. As a result of these
studies, Yamamoto and Fraenkel (1960b) proposed that these
chemical stimuli must act in sequence, i.e., the steam
distillable or volatile component attracts the moth to the
host and the ethanol extractable component induces the moth
to oviposit.

In a more recent paper, Tichenor and Seigler (1980)
assayed chemical fractions of tomato and tobacco leaves for
electroantennogram (EAG) activity. Steam distillable por-
tions of both plant species were found to be EAG active. An
acid-base extraction demonstrated that the volatiles could
be separated into an acidic and neutral fraction both of
which generated strong, positive EAG responses. The basic
fraction of this extraction did not give a positive EAG
response. Tests to examine ovipositional responses demon-
strated a strong response to both the acidic and basic
fractions but showed almost no response to the neutral
fraction. These results suggest that the two-step process
of host selection and oviposition as described by Yamamoto
and Fraenkel (1960b) holds true. Certain chemical com-

ponents serve as ”attractants" and orient the gravid females

6
to the host (steam distillable portion and neutral fraction)
while other components elicit ovipositional behaviors in the
moths (basic fraction) and are probably perceived upon
contact.

At the Tobacco Insect Investigation Laboratory at
Oxford, North Carolina, Sparks (1973) performed a study to
examine the influence of free water, water vapor, touch and
host-plant odor on ovipositional behavior in Manduca. Re—
sults of a two-choice test demonstrated that both odor and
moisture induce ovipositional behaviors and that a combi-
nation of these two factors is significantly more attractive
to the moth than either component separately. Further
testing revealed that contact of the ovipositor with a
bristly surface elicited a greater response than contact of
the ovipositor with a smooth surface. The results of these
tests strongly suggest that physical factors, as well as

chemical factors, influence behavioral responses in Manduca

s22.

gygzgigg 9; Most Selection

A number of factors are involved in the selection of
a host species. These behavioral stimuli can generally be
divided into those that are important during the prealigh-
ting (host searching) phase and those that exert an in-
fluence during the post-alighting (host testing) phase.

During the pre-alighting phase, behavioral responses

7
to olfactory and visual stimuli enable the insect to selec-
tively approach a particular plant species. Visual cues,
such as plant height and size and leaf shape, area and color
(Rausher 1978, Myers 1985) may provide the insect with a
"search image" of the preferred host (Rausher 1978, 1983).
Females responding positively to particular visual stimuli
will preferentially alight on plants matching this image
(Rausher 1978, Courtney 1983), thus increasing the proba-
bility that they will alight on the "true” host species.

Certain insects perceive the chemical odors produced
by their host plant from a distance (Hawkes 1974, Thibout
et. al 1982, Visser and Thiery 1985). For these species,
long range odoriferous compound induce an anemotactic re-
sponse, i.e., molecules of "attractive" chemicals stimulate
the insect to fly upwind into an odor plume until reaching
the source of the odor (Rausher, 1983). Kennedy and Moor-
house (1969) reported that locusts moved upwind in response
to wind borne grass odor. Hawkes (1974) determined that the
cabbage fly, Delia brassicae, was able to detect and subse-
quently locate crucifer crops from distances up to 15 meters
upwind.

In other species, host searching is largely made up
of random and directed random movements. For these insects,
non-hosts are approached at the same rate as hosts species.
However, once the insect enters the area immediately sur-
rounding a host plant and encounters the associated chemical

stimuli, subsequent movement becomes directed towards the

8
odor source. Manduca sexta is known to approach objects of
various physical attributes nearly equally, but tends to
veer away from non-hosts at distances of from 4 to 15 cm
(Yamamoto et. al. 1969).

It is well known that plant chemicals also exert an
influence during the post-alighting phase of the selection
process (see Rausher, 1983). These stimuli are often re-
ceived through chemoreceptors located on the tarsi or on the
proboscis (Stadler 1984) and provide the insect with infor-
mation on the suitability of a plant as an ovipositional
site. Bieris brassigae females have been noted "drumming"
the host plant with their front legs prior to ovipositing
(Ilse, 1956). Delia antique exhibits pre-ovipositional
behaviors that include a number of "test runs" on the onion
plant (Miller and Harris 1985). Manduca gpp. are known to
place their front two pairs of tarsi on the upper leaf
margin of the host before curling the abdomen around to the
lower surface and depositing an egg (Sparks 1973). These
behavioral responses suggest that females ”test" potential
hosts for the proper array of chemical and physical stimuli

prior to oviposition.

Chemical Components of Host-Selection

A vast amount of research has been devoted to deter-

mining the major chemical constituents that enable an insect

to distinguish among potential hosts. Two basic theories

9
have been proposed concerning the mechanisms that mediate
host-plant selection. During the 1950's and 60's, each of
these theories was debated quite extensively and, while not
as intensely argued today, still remain a source of con-
troversy.

Segondary_§hemigal§: Many phytophagous insects are
associated with particular groups of plants. For these
specialists, acceptable host plants are generally limited to
species that are taxomonically related and contain similar
classes of chemical compounds (Fraenkel et. al. 1962,
Schoonhoven 1968). These secondary or "key" compounds,
include the plant glycosides, essential oils, alkaloids,
etc. It has been hypothesized that these chemical sub-
stances were originally synthesized by the plant as a defen-
sive mechanism against herbivory, but later, came to be
utilized in the host selection process by those insects that
were able to adapt to the chemical "flavors" of their hosts
(Dethier 1947, Finch 1978). Thus, for insects attracted to
these "flavors”, selection is dependent upon a behavioral
response towards this chemical stimulus (Hawkes 1974,
Thibout et. al. 1982).

As early as 1910, Verschaffelt reported that host
selection was largely influenced by the presence of certain
chemical compounds. Lending support to these early find-
ings, similar types of behavioral responses have been noted
in more recent studies. Gupta and Thorsteinson (1960)

determined that isothiocyanates stimulated egg-laying by

10

Elaaella maeallpeaala. Isothiocyanates have also been shown
to induce host-finding and ovipositional behaviors by Della
Dzaaeleae, the adult cabbage maggot (Nair et. al. 1976,
Finch 1978). Manaaea aexea and Lepelneeaaaa geeemlineata
are known to be stimulated by the steroidal glycoalkaloids
associated with the Solanaceous plant family (Yamamoto and
Fraenkel 1960a and 1960b, Mitchell and Harrison 1985).

ut a a o 3: Since the nutrition of the
host-plant can affect the nutrition of the insect (House
1969, Fox and MacCauley 1976, Rough and Pimentel 1978, Hare
1983), it is reasonable to hypothesize that insects are
equipped with a mechanism that enables them to distinquish,
in part, the kinds and quantities of nutrients that a host
contains (Thorsteinson 1958). Thorsteinson (1960) was one
of the first to argue that although secondary chemicals
might be important in host recognition, nutrients also
stimulate positive behavioral responses by phytophagous
insects. Plant sugars, water and nitrogen are nutritional
components known to influence insect performance.

Sugars exert a direct chemotactic influence on phyto-
phagous insects (Dethier 1947). Fermenting sugar baits have
been used for collecting nocturnal moths (Utrio 1983) and
for monitoring populations of tortricids (Green et. al.
1960). In addition, these compounds are also known to
interact as stimuli with several other groups of substances
(Thorsteinson 1960, Utrio 1983).

The positive affect of plant water on behavioral

11
responses in herbivorous insects is widely documented
(Dethier and Chadwick 1948, Painter 1953, Shorey 1964, Myers
1985). Wolfson (1980) determined that the number of eggs
deposited by Diegis zapae was positively correlated with
leaf water content. Sparks (1973) demonstrated that free
moisture and water vapor induced ovipositional behaviors in
WERE-

Insects are well known to respond to several different
nitrogenous compounds (see Thorsteinson 1960). RaageleEla
pelaeaella is attracted to traps containing a bait composed
of powdered egg albumin (Hewitt and Connor 1965). More
recently, Myers and Post (1981) found a positive correlation
between host plant quality (measured as percent nitrogen)
and the number of eggs oviposited by the Cinnabar moth per
gram of plant material.

It is now generally recognized that both secondary
chemicals and nutritional compounds influence host
searching and acceptance (Kennedy 1965, Schoonhoven 1968,
Wolfson 1978, Miller and Strickler 1984). Kennedy (1965)
proposed the dual-discrimination theory whereby both secon—
dary chemicals and nutritional substances elicit oviposi-
tional responses by insects. It is likely that, at a given
time a particular chemical may serve as an allelochemic or
as a nutrient depending on its concentration in the plant
and the physiological state of the insect (Scriber and
Slansky 1981). A current hypothesis suggests that secondary

chemicals may be utilized to discriminate among particular

12
plant families or species while nutritional factors are
mostly used to discriminate among individual plants within a

particular host species (Wolfson 1978).

He i

An herbivore must find and ingest food that is "capable
of being assimilated and converted into energy and struc-
tural substances" as well as meeting the nutritional re-
quirements of that insect (Beck and Reese 1976). Nitrogen
is a key mineral that is associated with a number of these
metabolic processes (Mattson 1980). Not only does this
substance influence the physiology of an insect, but it also

affects the general behavior of the animal (Beck 1974).

Direct Effects on Insect Nutrition

Acquiring a balanced supply of amino acids and proteins
from a particular food source may present a major obstacle
for many herbivores (McNeill and Southwood 1978). First,
plant nitrogen concentrations vary throughout the host plant
and throughout the growing season (Beck 1972, Mattson 1980,
White 1984). Secondly, a large portion of the total nitro-
gen found in plant tissue is not readily available to an
herbivore (McNeill and Southwood 1978, Reese 1978, Mattson
1980). Finally, the nutritional value of proteins injested

by the herbivore is dependent upon the ratio, pattern and

l3
quantity of amino acids that make up those proteins (Rock
1972, House 1969).

Increased levels of plant nitrogen are often bene-
ficial and attractive to an insect. Growth rate, longevity,
fecundity, and/or survivorship are known to increase (Shaw
and Little 1972, van Emden and Way 1973, Auerbach and Strong
1981, McNeill and Prestidge 1982, Al-Zubaidi and Capinera
1984). Plants with high nitrogen concentrations often have
higher infestations of herbivores (Kennedy 1958). White
(1979) proposed that increased levels of plant nitrogen
cause a subsequent increase in the numbers of insects found
on that plant. House (1966a) reported that resistant plant
varieties are those having the lowest concentrations of
amino acids, implying that susceptible varieties are those
having the higher levels. Regan et. al. (1978) and Semtner
et. al. (1980) found that nitrogen fertilization increased
populations of Manduea aexea.

For other insect species though, plant nitrogen con-
tent appears to have little influence on overall perfor-
mance. Often those individuals developing on a nutritional-
ly inadequate diet can regulate consumption (House 1965,
Slansky and Feeny 1977). Scriber and Slansky (1981) deter-
mined that a reduction in plant nitrogen had little or no
effect on growth due to an increase in the rate of consump-
tion and/or an increase in the amount of time spent feeding,

digesting or developing.

14

Indirect Effects on Insect Nutrition

It is obvious that the availability of nitrogen af-
fects the pattern and quality of amino acids and proteins in
the plant and hence, directly influences herbivore nutri-
tion. However, because of its influence on other metabolic
processes, plant nitrogen may also indirectly affect insect
fitness.

Nitregen, Plan; Water and Merblvoay: Plant water
content is one aspect of metabolism affected by nitrogen
availability. The association between plant nitrogen con-
tent and leaf water is well documented (Scriber 1977, Slan-
sky and Feeny 1977, Mattson 1980, Mattson and Haack 1987).
Positive correlations between concentrations of plant nitro-
gen and levels of plant water have been demonstrated by a
number of researchers (Benebal and Hall 1967, Slansky and
Feeny 1977, Manuwoto and Scriber 1985). Similarly, White
(1978) and Mattson and Haack (1987) reported that drought
increased the levels of nitrogenous compounds found in plant
tissue.

The effect of plant water on ovipositional responses
by insects has already been established (see discussion
under host selection). Additionally, however, concentra-
tions of plant water can also influence insect fitness. Al-
though once overlooked as a nutrient, it is now well docu-
mented that growth and development of an herbivorous insect

is dependent upon the water content of its diet

15

(Pratt et. al. 1972, Beck and Reese 1976, Feeny 1976, Reese
1983, Hagen et. al. 1984). Scriber (1978) determined that
various insect growth parameters are positively correlated
with percent leaf water. Larvae consumed, grew and utilized
their food faster and more efficiently when fed diets con-
taining higher concentration of leaf water (see also Scriber
1977, Scriber and Slansky 1981, Reynolds et. al. 1985).

en kal 'ds a e b : Approximately,
one-seventh of the total number of flowering plants or
approximately 100 families (McKee 1962, Robinson 1979)
contain alkaloids. Many of these alkaloids are derived from
amino acids and can form through simple metabolic proces-
ses. Other alkaloids are more closely related to steroidal
precurssors--steroidal glycoalkaloids--(McKee 1962, Geissman
and Crout 1969) and are thought to arise through the degra-
dation of cholesterol (Geissman and Crout 1969). Generally,
alkaloids accumulate in young, actively growing tissues.
However, these compounds have also been documented in vas-
cular sheaths, latex vessels and epidermal and hypodermal
tissues (Robinson 1979).

Within a given plant species, the concentration of an
alkaloid varies depending on the particular tissue examined
and the supply of minerals, soil moisture, temperature and
light that the plant receives (Waller and Nowacki 1978,
Robinson 1979, Mattson and Haack 1987). Since alkaloids are
compounds containing nitrogen, it is reasonable to assume

that the availability of this element will affect alkaloid

16
metabolism in the plant (Waller and Nowacki 1978). This
response has, in fact, been demonstrated in several cases.
Waller and Nowacki (1978) cite several studies which show
that the addition of nitrogen affects the production and
subsequent level of alkaloids in a plant. A review of the
literature by Mattson (1980) found that "optimal growth
conditions and high soil nitrogen levels favored maximal
alkaloid synthesis”. Geissman and Crout (1969) determined
that increased plant nitrogen was paralleled by an increase
in alkaloid production. However, high levels of plant
nitrogen do not necessarily mean that alkaloid concentra-
tions will increase. For the steriodal glycoalkaloids,
commonly found in many species of Solanum and Lycopersicon,
biosynthesis is dependent upon the ratio of carbon to nitro-
gen: therefore, unless sufficient stores of carbon are
present, an increase in the availability of nitrogen will
not increase total alkaloid levels (Nowacki et. al. 1976,
Waller and Nowacki 1978).

Many of the alkaloids are noted for their physio-
logical effects on animals and are known to prevent normal
metabolic activities from occurring by disrupting membrane
transport, enzyme inhibition and activation, DNA replica-
tion, RNA transcription and protein synthesis in the animal
(McKee 1962, Robinson 1979). In insects, alkaloids may in-
hibit growth and development (Harley and Thorsteinson 1967,
Nowacki et. al. 1976) or act as a repellent. Leptine I, an

alkaloid found in potatoes and tomatoes, deters feeding by

17
the Colorado potato beetle (Tingey 1984, Sinden et. al.
1986).

It has been hypothesized that insects adapted to
particular plant families utilize the secondary chemicals
associated with those families as host searching cues.
Although other secondary chemicals, such as sinigrin and
isothiocyanates are known to induce host searching behavior
in certain phytophagous insects (Gupta and Thorsteinson
1960, Nair and McEwen 1976, Finch 1978), no studies are
known that demonstrate a similar response with plant alka-
loids, although, glycosides, a structural component of
glycoalkaloids, are thought to induce feeding in Manduea
eeata (Yamamoto 1960b).

Literature Review Summagy

Manduca app. are specialist insects that utilize the
solanaceous plants (glycoalkaloid containing plants) as
hosts. Adult females respond to short range odors and
chemical "flavors" of particular plants. The moths also
respond to physical attributes associated with their hosts.
Both plant moisture and leaf texture influence ovipositional
behaviors. When Maaaaea moths are presented with combina-
tions of host stimuli, a synergistic effect is induced,
implying that host selection and oviposition are influenced
by an array of factors that, in combination, provides the

moth with "unique" information regarding its host species.

18

In general, host selection by a specialist phytophage
may be dependent upon particular ”key" plant compounds or
certain nutritional components. However, it is likely that
combinations of these plant factors interact during dif-
ferent stages of the host selection process to produce a
"pattern" of chemical stimuli that elicit a response in the
insect.

Nitrogen is one plant component known to directly
influence plant physiology and subsequent insect perfor—
mance. Depleting or increasing the nitrogen pool alters the
pattern and concentration of plant substances which incor-
porate this element as a structural component. These al-
terations can, in turn, affect host finding, ovipositional

response and larval development.

REFERENCES TO LITERATURE REVIEW

Al-Zubaidi PS, Capinera JL (1984) Utilization of food and
nitrogen by the beet armyworm, Spodoptera exigua
(Hubner) (Lepidoptera: Noctuidae) in relation to
food type and dietary nitrogen levels. Env. Ent.
13:1604-1698.

Auerbach MJ, Strong DR (1981) Nutritional ecology of
Meleiceaia herbivores: experiments with plant
fertilization and alternative hosts. Ecological
Monographs. 51:63-83.

Baumhover AH (1985) Manduca sexta. In: Handbook of Insect
Rearing. Vol. 2. (Pritam S, Moore RF, eds.). Elsevier
Science Publishers B.V. Amsterdam.

 

Beck SD (1965) Resistance of plants to insects. Ann. Rev.
Entomol. 10:207-232.

Beck SD (1972) Nutrition, adaptation and environment. In:
Insect and Mite Nutrition. North-Holland Publishing
Co. Amsterdam.

Beck SD (1974) Theoretical aspects of host plant specificity
in insects. In: Proc. Summer Institute on Biological
Control of Plant Insects and Diseases. (Maxwell FG,
Harris FA, eds.). Univ. Press of Miss. Jackson,

Miss.

Beck SD, Reese JC (1976) Insect-plant interactions:nutrition
and metabolism. In: Biochemical Interactions Between
Plants and Insects. (Wallace JW, Mansell RL, eds.).
Rec. Adv. Phyto Chem. 10:41-92.

Benebal PS and Hall CV (1967) Biochemical composition of
host plants of Drassica oleracea var. capitata (L)
as related to resistant to TIiQhQElESQi al Hubner

and Eleala zapae Linne. Proc. Am. Soc. Hort. Sci.
91:325-330.

Bernays EA (1982) The insect on the plant-a closer look.
Proc. Fifth Int. Symp. Insect-Plant Relationships.
PUDOC, Wageningen.

19

20

Courtney SP (1983) Models of hostplant location by but-
terflies: the effect of search images and search ef-
ficiency. Oecologia. 59:317-321.

Dethier VG (1947) Chemical Insect Attractants and Repel-
lents. Blakiston Co. Philadelphia.

Dethier VG, Chadwick LE (1948) Chemoreception in insects.
Physiol. Rev. 28:220-54.

Feeny P (1976) Plant apparency and chemical defense. Rec.
Adv. in Phytochem. 10:1-40.

Finch S (1978) Volatile plant chemicals and their effect on
host plant finding by the cabbage root fly (Della

Drasslcae). Ent. exp. appl. 24:350-359.

Fox LR, MacCauley BJ (1976) Insect grazing on Eucalyptus in
response to variation in leaf tannins and
nitrogen. Oecologia. 29:145-162.

Fraenkel G, Nyar J, Nalbandov O, Yamamoto RT (1962) Further
investigations into the chemical basis of the in-
sect-hostplant relationship. XI Int. Congr. Entomol.
Vienna, Verhandlungen, Bd. III.

Geissman TA, Crout DHG (1969) Organic Chemistry of Secondary
Plant Metabolism. Freeman, Cooper and Co. San Fran-
sico, California.

Gilmore JU (1938) Observations on the hornworms attacking
tobacco in Tennessee and Kentucky. J. Econ. Ent.
31:706-712.

Green N, Geroza M, Hall SA (1960) Recent developments in
chemical attractants for insects. Advances in Pest
Control Research 3:129-179.

Gupta PD, Thorsteinson AJ (1960) Food plant relationships of

the diamond-back moth (Elutella maculipenaie
(Curt.)). II. Sensory regulation of oviposition of

the adult female. Ent. exp. appl. 3:305-314.

Hagen KS, Dadd RH, Reese J (1984) The food of insects. In:
Ecological Entomology. (Huffaker CB, Rabb RL, eds.).
John Wiley and Sons, Inc.

Hare JD (1983) Seasonal variation in plant-insect associa-
tions:utilization of Solanam dulcamara by Leptiao-

tagsa aecialineata. Ecology. 64:345-361.

21

Harley KLS, Thorsteinson AJ (1967) The influence of plant
chemicals on the feeding behavior, development and
survival of the two-stripped grasshopper, Melanoptus
Dixiegagaa (Say), Acrididae: Orthoptera. Can. J.
2001. 45:305-319.

Hawkes C (1974) Dispersal of adult cabbage root fly (Erics-
eala baassicae (Bouche)) in relation to a brassica
crop. J. appl. ecol. 11:83-93.

Hough JA, Pimentel D (1978) Influence of host foliage on
development, survival and fecundity of the gypsy
moth. Env. Ent. 7:97-102.

House HL (1965) Effects of low levels of the nutrient
content of a food and of nutrient imbalance on the
feeding and the nutrition of a phytophagous larva,

Qeleaie eaphobiae (Linnaeus) (Lepidoptera: Sphin-
gidae). Can. Ent. 97:62-68.

House HL (1965b) Insect nutrition. In: The Physiology of
Insecta. (Rockstein M, ed.). Academic Press. New
York. 2:815-858.

House HL (1969) Effects of different proportions of nu-
trients on insects. Ent. exp. & appl. 12:651-669.

Howitt AJ, Connor LJ (1965) The response of Rhageleela
poaonella (Walsh) adults and other insects to trap
boards baited with protein hydrolysate bait. Proc.
Entomol. Soc Ontario. 95:134-136.

Ilse D (1956) Behavior of butterflies before ovipositing.
J. Bombay Nat. Hist. Soc. 53:486-488: cited by
Miller and Strickler, 1984.

Kennedy JS (1958) Physiological conditions of the hostplant
and susceptibility to aphid attack. Ent. exp. & appl.
1:50-65.

Kennedy JS (1965) Mechanisms of host plant selection.
Ann. Appl. Bio. 56:317-322.

Kennedy JS, Moorehouse JE (1969) Laboratory observations on
locust responses to wind-borne grass odour. Ent. exp.
& appl. 12:487-503.

Madden AH, Chamberlin FS (1945) Biology of the tobacco
hornworm in the southern cigar-tobacco district. US
Dept. of Agric. Tech. Bull. No. 986.

22

Manuwoto S, Scriber JM (1985) Consumption and utilization of
experimentally altered corn by southern armyworm:
iron, nitrogen and cyclic hydroxamates. Journal of
Chemical Ecology. 11:1469-1483.

Mattson WJ, (1980) Herbivory in relation to plant nitrogen
content. Ann. Rev. Ecol. Syst. 11:119-161.

Mattson WJ, Haack RA (1987) The role of drought in outbreaks
of plant-eating insects. BioScience. 37:110-118.

McFadden MW (1968) Observations on feeding and movement of
tobacco hornworm larvae. J. Eco. Entmol. 61:352-356.

McKee HS (1962) Nitrogen Metabolism in Plants. Clarendon
Press. Oxford.

McNeill S, Prestidge RA (1982) Plant nutritional strategies
and insect herbivore community dynamics. Proc. Fifth
Int. Symp. Insect-Plant Relationships. PUDOC, Wage-
ningen. pp. 225-235.

McNeill S, Southwood TRE (1978) The role of nitrogen in the
development of insect/plant relationships. In: Bio-
chemical aspects of plant and animal coevolution.
(Harborne JB, ed.). Academic Press. London.

Miller JR, Harris MO (1985) Viewing behavior-modifying
chemicals in the context of behavior: lessons from
the onion fly. In: Chemical Ecology of Insects.
(Bell WJ, Carde RJ, eds.,). Sinauer Associates. Sun-
derland, MA.

Miller JR, Strickler KL (1984) Finding and accepting host
plants. In: Chemical Ecology of Insects. (Bell WJ,
Carde RT, eds.). Sinauer Associates, Sunderland,
MA.

Mitchell BK, Harrison GD (1985) Effects of Solanum glyco-
alkaloids on chemosensilla in the Colorado potato
beetle. A mechanism of feeding deterrence? Journal
of Chemical Ecology. 11:73-83.

Myers JH, Post BJ (1981) Plant nitrogen and fluctuations of
insect populations: a test with the Cinnabar moth-
tansy ragwort system. Oecologia. 48:151-156.

Myers JH (1985) Effect of physiological condition of the
host plant on the ovipositional choice of the cabbage

white butterfly, Eieais rapae. Jour. of Animal Eco.
54:193-204.

23

Nair KSS, McEwen FL, Snieckus V (1976) The relationship
between glucosinolate content of cruciferous plants
and ovipostion preferences of Mylema Drasaica (Dip-
tera: Anthomyiidae) Can. Ent. 108:1031-1036.

Nowacki E, Jurzysta M, Gorski P, Nowacka D, Waller GR
(1976). Effect of nitrogen nutrition on alkaloid
metabolism in plants. Biochem. Physiol. Pflanzen.
169:231-240.

Painter RH (1953) The role of nutritional factors in host
plant selection. Trans. Ninth Int. Congr. Ent. Vol.
2:101-105.

Pratt JJ, House HL, Mansingh A (1972) Insect control
strategies based on nutritional principles: a pro-
spectus. In:Insect and Mite Nutrition. (Rodriques JG,
ed.). North-Holland Publishing Co. Amsterdam.

Rausher MD (1978) Search image for leaf shape in a butter-
fly. Anim. Behav. 27:1034-1040.

Rausher MD (1983) The ecology of host selection behavior in
phytophagous insects. In: Variable Plants and
Herbivores in Natural and Managed Systems. (Denno RF,
McClure MS, eds.). Academic Press, New York.

Reese JC (1978) Chronic effects of plant allelochemics on
insect nutritional physiology. Ent. exp. & appl.
24:425-431.

Reese JC (1983) Nutrient-allelochemical interactions in
host plant resistance. In: Plant Resistance to In-
sects (Hedin PA, ed.). ACS Symposium Series, No.
208. American Chemical Society. Washington, DC.

Regan TE, Rabb RL, Collins WK (1978) Selected cultural
practices as affecting production of tobacco.
J. Econ. Entomol. 71:79-82.

Reynolds SE, Nottingham SF, Stephens AE (1985) Food and
water economy and its relation to growth in fifth-
instar larvae of tobacco hornworm, Manaaca aeaga. J.
Insect Physiol. 31:119-127.

Robinson T (1979) The evolutionary ecology of alkaloids.
In: Herbivores Their Interaction with Secondary
Plant Metabolites. (Rosenthal GA, Janzen DH, eds.).
Academic Press. New York.

Rock GD (1972) Optimal proportions of dietary amino acids.
In: Insect and Mite Nutrition. (Rodriquez JG, ed.).
North-Holland Publishing Co. Amsterdam.

24

Schoonhoven LM (1968) Chemosensory bases of host plant
selection. Ann. Rev. Ent. 13:115-136.

Scriber JM (1977) Limiting effects of low leaf-water content
on the nitrogen utilization, energy budget and larval

growth of Myalepaeaa eeerepla (Lepidoptera: Saturni-
idae). Oecologia. 28:269-287.

Scriber JM (1978) The effects of larval feeding speciali-
zation and plant growth form on the consumption and
utilization of plant biomass and nitrogen:an ecologi-
cal consideration. Ent. exp. & appl. 24:494-510.

Scriber JM, Slansky F (1981) The nutritional ecology of
immature insects. Ann. Rev. Entomol. 26:183-211.

Semtner PJ, Rasnake M, Terrill TR (1980) Effect of host-
plant nutrition on the occurrence of tobacco horn-
worms and tobacco flea beetles on different types of
tobacco. J. Econ. Entomol. 73:221-224.

Shaw GG, Little CHA (1972) Effect of high urea fertilization
of balsam fir trees on spruce budworm development.
In:Insect and Mite Nutrition. (Rodriquez JG, ed.).
North-Holland Publishing Co. Amsterdam.

Shorey HH (1964) The biology of TIiQDQQlB§iQ Bi (Lepidop-
tera: Noctuidae). III. Response to the oviposition
substrate. 57:165-170.

Sinden SL, Sanford LL (1986) Leptine glycoalkaloids and
resistance to the Colorado potato beetle (Coleoptera:

Chrysomelidae) in Solanam chacoense. Env. Ent. 15:
1057-1062.

Slansky F, Feeny P (1977) Stabilization of nitrogen accum-
ulation by larvae of the cabbage butterfly on wild
and cultivated food plants. Ecol. Monogr. 47:209-228.

Sparks MR (1973) Physical and Chemical stimuli affecting

oviposition preference of Maaduca eexta (Lepidoptera:
Sphingidae). Ann. Ent. Soc. Am. 66:571-573.

Stadler E (1984) Contact Chemoreception. In: Chemical Eco-
logy of Insects. (Bell WJ, Carde RJ, eds.). Sinauer
and Associates, Sunderland, MA.

Thibout E, Auger J, Lecomte C (1982) Host plant chemicals
responsible for attraction and oviposition in gage:
° aaeeeeella. In: Insect-plant relationships.
(Visser JH ed.). Minks PUDOC, Wageningen.

25

Thorsteinson AJ (1958) Acceptability of plants for phyto-
phagous insects. Proc. Tenth Int. Congr. Entomol.
Montreal 1956. 2:599-602.

Thorsteinson AJ (1960) Host selection in phytophagous
insects. Ann. Rev. Ent. 5:193-218.

Tichenor LH, Seigler DS (1980) Electroantennogram and ovipo-
sition responses of Maagaea sexta to volatile com-
ponents of tobacco and tomato. J. Insect Physiol.
26:309-314.

Tingey W (1984) Glycoalkaloids as pest resistance factors.
Am. Potato J. 61:157-167.

Utrio P (1983) Sugaring for moths:why are noctuids attracted
more than geometrids. Ecological Ent. 3:437-445.

van Emden and Way (1973) Host plants in the population
dynamics of insects. Insect/Plant Relationships.
Symposia of the Royal Entomological Soc. of London.
No. 6. Blackwell. Oxford.

Verschaffelt E (1910) The cause determining the selection of
food in some herbivorous insects. Proc. R. Acad. Am.
13:536-42.

Visser H, Thiery D (1985) Behavioral responses of the Colo-
rado potato beetle to stimulation by wind and plant
odors. In: Proc. Symp. Colorado potato beetle,
Seventeenth Int. Congr. Entomol. (Ferro DN, Voss RH,
eds.). Chp. 10. Bull. Mass. Agric. Exp. Stn.

Waldbauer GP, Fraenkel GS (1961) Feeding on normally re-
jected plants by maxillectomized larvae of the tobac-
co hornworm, Drotoparce sexta (Lepidoptera, Sphingi-
dae). Ann. Ent. Soc. Amer. 54:477-485.

Waldbauer GP (1962) The growth and reproduction of maxillec-
tomized tobacco hornworms feeding on normally rejec-
ted nonsolanaceous plants. Ent. exp. & app. 5:147-
158.

Waldbauer GP (1964) The consumption, digestion and utiliza-
tion of solanaceous and non-solanaceous plants by
larvae of the tobacco hornworm, Eretoparce aeaea
(Johan.) (Lepidoptera: Sphingidae). Ent. exp. & appl.
7:251-269.

Waller GR, Nowacki EK (1978) Alkaloid Biology and Metabolism
in Plants. Pleunum Press, NY and London.

White TCR (1978) The importance of a relative shortage of
food in animal ecology. Oecologia. 33:71-86.

26

White TCR (1979) An index to measure weather-induced stress
of trees associated with outbreaks of psyllids in
Australia. Ecology. 50:905-909.

White TCR (1984) The abundance of invertebrate herbivores in
relation to the availability of nitrogen in stressed
food plants. Oecologia. 65:90-105.

Wolfson JL (1978) Environmentally induced variation in
Daassica nigga (Koch): affects on generalists and
specialist herbivores. Ph.D. dissertation. State
Univ. of New York. Stony Brook

Wolfson JL (1980) Oviposition response of Dieria zapae to
environmentally induced variation in Dzaeelea nlg_a.
Ent. exp. & appl. 27:223-232.

Yamamoto RT, Jenkins RY, McClusky RK (1969) Factors deter-
mining the selection of plants for oviposition by the

tobacco hornworm, Manduca sexga. Ent. exp. & appl.
12:504-508.

Yamamoto RT, Fraenkel GS (1960a) The specificity of tobacco

hornworm, Baotoparce sext , to solanaceous plants.
Ann. Ent. Soc. Amer. 53:503-507.

Yamamoto RT, Fraenkel GS (1960c) The physiological basis for
the selection of plants for egg-laying in the tobacco

hornworm, Protoparce sexta (Johan.) XI Int. Kongr.
Ent. Wien. 3:127-133.

 

MANUSCRIPT I: The biology and seasonal occurrence of

Maaaaea gainqaemaeulata and Manduca sexta (Lepidoptera:
Sphingidae) in southwestern Michigan.

27

28

ABSTRACT

The seasonal occurrence and development of field

 

populations of Manduca sexta and Manduca gaingaemaculata
were monitored at the Kellogg Biological Station in Barry
Co., Michigan from late June to early September during 1985
and 1986. Both insect species were equally abundant in
field plots during mid-season (end of July to mid August) in
both years of the study. In 1985, Ml gaingaemaculata com-
prised approximately 80% of the hornworm population sampled
early in the season, while Ml eeaea was the more prevalent
species late in the season, comprising 100% of the hornworm
population sampled. In 1986, although Ml SQKLQ was again
the more prevalent species late in the season (73%), both
species were relatively common during the beginning of the
season. These observations suggest that Ml gaingaemaculata

is univoltine in Michigan while Ml sexta may be multivol-

 

tine.

Egg deposition and larval development were found to
be closely associated with degree day (DD) accumulation base
50° F. Peak oviposition occurred on DD 1350 (July 11) in
1985 and DD 1320 (July 10) in 1986. Field and laboratory
experiments demonstrated that development rates were similar

for both species, requiring, on the average, 64 DD per in-

star.

29

INTRODUCTION

Studies conducted in the southern tobacco growing
regions have documented the biology, distribution and occur-
rence of field populations of Maaaaea eeaaa (L.), the tobac-
co hornworm (Lepidoptera: Sphingidae) on two common host
plants, tobacco and tomato (Madden and Chamberlin 1945, Rabb
1966, McFadden 1968, Jones and Thurston 1970, Semtner et.
al. 1980, Kennedy et. al. 1984, Borth and Harrison 1984, de
Boer and Hanson 1984). However, only one such study has
examined this insect in its more northern habitats (Chang
1977) and only on one host, tobacco. Furthermore, although
it is widely known that Ml §§E§Q shares its resources with a
closely related species (Madden and Chamberlin 1938, Gilmore
1938, Borth and Harrison 1984), Ml gaiagaemaculata (Ha-
worth), the tomato hornworm, few studies compare the two
insects (however, see Gilmore 1938).

In Michigan, larvae of Ml EQELQ and Ml galagaeaaea;
laaa readily feed on both tomato and potato foliage. Cur-
rently, growers achieve control of these pests through the
application of one of a variety of insecticides, i.e.,
Daelllaa thureagiensia. Unfortunately, although present
management techniques limit feeding damage by these her-
bivores, the potential for pesticide resistance and there-
fore, economic damage, always exists. As a result, informa-
tion which contributes to our overall knowledge and under-

standing of these two species will prove important for

30
future pest management programs.

Given the lack of cooperative studies of the horn-
worms on tomato and tobacco in more northern habitats, this
study was designed to document the seasonal occurrence of Ml
aexea and Ml gaingaemaculata in fields in southwestern
Michigan and investigate possible differences in development
rate of larvae and adults for both species. Furthermore,
the research was designed to examine the effectiveness of
accumulated degree days as a predictor of egg deposition and
subsequent insect development. We hypothesized that sea-
sonal abundance and insect development would be different
for the two species and that moth activity and larval growth

would be closely correlated with DD base 50° F.

MATERIALS AND METHODS

W

This study was part of a sustainable agriculture
project. Investigations into the affects of intercropped
systems on the abundance and distribution of insects con-
tinues to be an important aspect of this long term project.
In accordance with these goals, two separate sites, a tomato
monoculture, and a tomato/apple tree intercropping, were
included in this particular study. The intercropped site
consisted of a determinant cultivar of tomato, var. Sunny,
planted between tree rows in a dwarf apple orchard. The

tomato monoculture consisted of the same cultivar of tomato

31
planted in an adjacent field. In 1986, the field site was
moved from its original location to a similar nearby area.
The tomato plots within the orchard site remained at the
same location throughout both years of the study.

Both sites were divided into multiple plots of tomato
plants. Each plot consisted of double rows of plants 50 m
long. Plants in these rows were spaced about .5 m apart.
Each plot contained approximately 200 tomato plants. Plots
in the field site were spaced 5 m apart while those in the
orchard site were spaced 7 m apart and separated by a row of
apple trees. There were a total of 8 plots in 1985 and 7 in

1986.

W

Tomato plants (half plants) were randomly sampled
twice weekly during both field seasons. In addition, in
1986 plants were sampled every other day from July 13 to
July 21. Actual plants sampled were chosen by utilizing a
random numbers table. On each sampling day in 1985, 60 half
plants, approximately 15 per plot, were visually scanned for
eggs and larvae. In 1986, 96 half plants were sampled each
day through August 21. However, beginning on August 25,
only 48 half plants were sampled per day (in 1986, plants in
the orchard site became heavily infected with SEQIQIIQ
lyeepezalel and in late August, sampling was terminated at
this site due to lack of foliage).

Tomato plant stems and flowers and upper and lower

32

leaflets, were searched. The location of all eggs or larvae
was recorded and the host plant tagged. For sampled larvae,
the growth stage and species were also recorded. Since the
two species are readily distinquishable by the third instar
(Ml EEKLQ larvae are characterized by a scarlet horn and a
soft, pubescent covering on their body, Ml galagaeaaealaea
have a black horn and lack pubescence), tagging plants
enabled us to return to specific locations to determine the
species of previously indistinguishable eggs and early
instar larvae. By not sampling tagged plants on successive
days, the procedure eliminated the possibility of counting

the same eggs or larvae more than once.

Ei§l§_QQhQII_§§EQY

Eggs deposited in the field on July 10th and 13th,
during 1986, were monitored every other day to determine
larval developmental period for both species. Monitoring
was continued until a larva could not be located on three
subsequent monitoring days. After this time period, moni-
tored larvae were categorized as follows: 1) if the larva
had previously been documented as having reached the fifth
instar, we assumed it had left the plant as a prepupa, 2) if
the larva had last been observed as an earlier instar and
was then located on a subsequent sampling day at a later
growth stage but falling into the expected sequence of
development, the larva was assumed to be the same hornworm

originally monitored and was retained in the study, and

33

3) if the larva had last been observed as an earlier instar
and could not be located on any of the three subsequent
sampling days, we assumed the larva had "disappeared" and it
was dropped from the monitoring study. A second cohort
study was conducted later in the season based on eggs depo-
sited on August 8th and 10th using the methodology described
above. Larval development time was calculated as the number
of days required for a first instar larva to reach the

"assumed" prepupal stage.

W12

Larvae from both hornworm species were reared during
May of 1986 from eggs received from the Oxford Tobacco
Research Station in North Carolina. Rearing conditions were
maintained at ambient temperature (23° - 25° C) and humidi-
ty, under a 16:8 light:dark cycle. Eggs from each species
were placed into separate plastic petri dishes. To ensure
that neonate larvae would immediately have access to food,
pieces of tobacco hornworm diet (BioServ, Product No. 59783)
were also placed into the dishes in close proximity to the
eggs. Upon reaching second instar, 20 larvae were randomly
selected from each species, removed from the petri dishes
and placed in individual plastic 50 ml containers. As newly
molted fifth instars, larvae were transferred to 1000 ml
containers to provide ample room for continued growth. All
rearing containers were covered with loose weave mesh cheese

cloth to ensure air exchange. Frass was removed from the

34
containers and diet was replaced daily. Larvae entering the
prepupal or ”wandering" stage, were placed into clean con-
tainers and allowed to pupate. Sticks, approximately 1 cm
wide and 25 cm long, were placed upright in each container
to provide a surface for newly emerged moths to crawl up on.
Pupae were checked for emergence early each morning. Larval
development time was calculated as the amount of time re-
quired for a larva to grow from neonate first instar to
prepupa. Development to the adult stage was calculated as
the amount of time required for a larva to grow from neonate

first instar to newly emerged adult.

RESULTS

The number of hornworm eggs and larvae sampled on
field grown tomato plants was low during both years of the
study (Figure 1). Eggs were first observed on July 2 in
1985 (DD 1150) and July 7 in 1986 (DD 1250). The mean
number of eggs/half plant peaked at .62 on DD 1350 (July 11)
in 1985 and .5 on DD 1320 (July 10) in 1986. In 1985, egg
deposition spanned a period of approximately three weeks.
However, in 1986, eggs were almost exclusively laid on July
10 and 13.

Because of the low infestation level of hornworms and
the fact that foliage in the orchard site was virtually
nonexistent by mid August in 1986 (a result of disease),

data were pooled from both the orchard and field sites

35
1.0‘ 1985

0.81
0.6‘

0.4'

0.2“

l

 

0-04 3 I ' I ' I ' I w I ' l ' r
1140 12340 1540 1740 1940 2140 2340 2540

 

1.0j 1986

HALF PLANT MEANS

 

 

0.0‘, . .
1140 1340 1540 1740 1940 2140 2340 2540

ACCUMULATED DEGREE DAYS

 

Figure 1. Seasonal trends of Manduca spp. (M. quinquemaculata and M. Sexta) egg and larval means per
half plant. Data combined from field and orchard site. Circles = eggs, stars = larvae. In 1985, 60 half
plants were sampled per day. In 1986, 96 half plants were sampled per day until August 25, at which
time, 48 half plants were sampled.

36
(Chi-square = 2.176, p > .15 in 1985 and 2.99, p > .10 in
1986) prior to separating eggs and larvae by species.

Figure 2 depicts the seasonal occurrence and relative
abundance of eggs from each species (estimated by monitoring
egg development through the season and by extrapolating from
larvae sampled and identified on a particular date, back to
the probable date that oviposition occurred). Moths from
both species began depositing eggs early in the season.
However, eggs deposition by Me galngaemaealaEa declined
throughout July and by early August (DD 2000), was non-
existent. In contrast, Ml SQKIQ eggs were observed from the
beginning of the season until sampling stopped in early
September (DD 2500).

Figure 3 shows the seasonal occurrence and relative
abundance of larvae from each species. In both years, Ml
gaiagaemaculata larvae were more prevalent during the early
part of the season (Table 1). In fact, in 1985, this spe-
cies was more abundant during the entire first half of the
season. Although the same trend existed in 1986, the dif-
ference between the proportion of larvae from each species
was much smaller (Figure 3). During the middle of the
season, the number of individuals from each species was
almost equal, especially in 1986 (Table 1). Towards the

last 2 1/2 - 3 weeks of the study, Ml sexta became the more

 

prevalent species sampled. In 1985, it was the only species
observed during this entire sampling period (Table 1 and

Figure 3).

37

0-307 1985

‘At.’

0 00 —--- - . - . ‘-—-—v—--

1100 1300 1500 1700 1900 2100 2300 2500

 

0.307

HALF PLANT MEANS

024*}
0.18%
0.12“}

0.06j

 

 

0.004. -. .1 ..
1100 1300 1500 1700 1900 2100 2300 2500

ACCUMULATED DEGREE DAYS

Figure 2. Seasonal egg means of M. quinquemaculata and M. sexta. Circles = M. quinquemaculata,
stars a M. sexta. Data combined from field and orchard site. In 1985. 60 half plants were sampled per
day. In 1986. 96 half plants were sampled per day until August 25, at which time, 48 half plants were
sampled.

 

38

Table 1. Relative occurrence of M. quinquemaculata and M. sexta larvae at different sampling times during
each year of the study. Each sampling period consists of approximately 5 separate sampling dates.

 

 

 

TOTAL NO.
SAMPLING LARVAE PERCENT PERCENT
YEAR PERIOD D Day>50 F IDENTIFIED M. quin. M. sexta
1985 We
SJUL- 18 JUL 1280-1515 18 89 11
11110322590
23 JUL- 2AUG 1649-1817 45 69 31
9 AUG - 20 AUG 1967-2184 10 50 50
late seasgn
23 AUG - 6 SEPT 2228-2522 13 0 100
1986 rl ea
7JUL- 19 JUL 1255-1551 30 77 23
1111111225211
22 JUL - 4 AUG 1625- 1920 64 42 58
7 AUG - 18 AUG 1980-2199 39 46 54
lateseasan

21 AUG - 2 SEPT 2258-2434 26 27 73

 

0.204 1985

 

 

 

O o O O u ‘ F T ' , l - 1 1 .-". '-5-G-5_-'. _3

1200 1400 1600 1800 2000 2200 2400 2600

0'20‘ 1986

HALF PLANT BEANS

0.16‘
0.12“
0.081

0.04J
1

(L00‘ . . .
1200 1400 1600 1800 2000 2200 2400 2600

ACCUMULATED DEGREE DAYS

 

 

Figure 3. Seasonal larval means of M. quinquemaculata and M. sexta. Circles a M. quinquemaculata.

stars . M. sexta. Data combined from field and orchard site. In 1985, 60 half plants were sampled per
day. In 1986, 96 half plants were sampled per day until August 25. at which time, 48 half plants were
sampled. .

40
l2§§.§9h21§.§§2§i§§

Laboratory studies determined that u; guinggemagulata
and a; sexta larvae require the same amount of time to
develop. No significant differences were found between the
time required for larvae to reach maturity or for adults to
emerge (Table 2).

Results of the first field cohort study demonstrated
similar findings. While larval development time was shorter
overall for field reared insects compared to laboratory
reared insects (a discrepancy which can probably be ex-
plained by differences in food quality, i.e. diet vs. foli-
age, and/or higher temperatures for field reared insects),
the period required for larvae to reach the prepupal stage
was not significantly different between the two species
(Table 2), requiring approximately 2 - 3 days for each of
the five instars or 64.30 DD i 2.60 (N = 29). In fact,
plots tracing larval development during the field cohort
study, demonstrated that instar peaks occurred on virtually
the same degree day for both species (Figure 4).

The second field cohort study produced no results.

Of the 17 eggs oviposited on August 8 and 10, only 1 larva
could be traced through to the prepupal stage. Of the
others, 2 eggs were parasitized, 2 were nonviable, and 13
larvae disappeared during the first instar. However, larval
development monitored in the latter part of the season from
regular sampling data (monitoring larvae on tagged plants)

indicated that larvae required 3 to 4 days to complete an

41

 

«<2 «we. H no.3 «3.. H 8.3 new. H :.3 83:8szng .2

 

 

 

$2 «3.2 H 3w... «8. H $2 «8. H $9 8x3. .2
38.2 mi 3mm 93 £85
$5.8 muzwcmmzm 628 m2?
H59. 09 82E 2.282.54ng .2563

 

 

.8. u a a .8556 8: ~32 2:8 55 232 .boéooamoe 90:8 Sou
Ska. .2 Ba eBEGQEuawsae .3 .8 2 c5 Quz .928 as .8 Suz Spam 3% met—6 oats Save. .3
new 6839323330.: .6 38:8 as 95 Soc .8 8.5. .SEeo_o>3 28 no.5 Emu—:3 pg 282 .N 93$.

42

100‘

1 M. sexta

80‘

O)
O

L

A
O

N
O
r

A a

 

 

oJ . A . _
f' I V I 1' I Y I I Y I .f I r V—ffi l r Y fi' r r Y I

1300 1400 1500 1600 1700 1800 1900 000

100‘

I M. quinquemaculata

80 J .—

1

60‘

PERCENT OF COHORT

40‘

 

20‘
1

0'1
1300 1400 1500 1600 1700 1800 1900 2000
ACCUMULATED DEGREE DAYS

 

 

 

V I VT fi" V fif"

Figure 4. Seasonal development of M. quinquemaculata and M. sexta field cohorts monitored from eg
hatch to prepupa during 1986 (N = 13 and 15 for M. quinquemaculata and M. sexta respectively). Each
data point represents the percent of total larvae for a particular stage of development on each monitoring
day. Triangles = eggs, closed stars = lst and 2nd instars, open circles = 3rd and 4th instars, squares = 5th

instars. open stars a prepupa.

43
instar compared to the 2 to 3 days required for early season
larvae: although, the number of DD required was very simi-

lar, 67.33 i 5.95 (N = 10).

DISCUSSION

since hornworm larvae are ravenous eaters and a
single larva can consume an entire tomato plant during
development, infestations of less than one larva per plant
should be desirable for growth of either n; sexta or E;
guinguemagulata. Thus, it is reasonable to assume that the
low population levels documented during this research were
conducive to optimum deve10pment in these insects: although,
other hypotheses are also viable.

Degree day accumulation effectively predicted ovipo-
sitional activity and larval development in a. sexta and H;
gningugmggglatg. For both species, initial egg lay occurred
during the first week of July, a difference of only 100 DD
or approximately 4 - 5 calendar days between years. An even
closer relationship was demonstrated for degree day accum-
ulation and peak egg deposition. In both years, peak egg
lay occurred on virtually the same DD. Larvae developed
rapidly and at the same rate in both species. The first
larvae of the season reached maturity by the end of July,
approximately 12 - 14 days (300 - 350 DD) after egg deposi-
tion. Although larvae developing later in the summer re-

quired more days to reach maturity, the apparent discrepancy

44
can be explained by the slower rate of heat accumulation
late in the season. From mid to late July (DD 1519 - 1903)
an average of 24.5 DD were accumulated each calendar day;
however, from mid to late August (DD 2081 - 2378), only 18.0
DD were accumulated each calendar day. Hence, larvae still
required, on the average, approximately 65 DD per instar to
develop.

In the southern states, 5; gninggemagulata and ML
sexta may have as many as four generations per year (Madden
and Chamberlin 1938, 1945). First generation moths emerge
from late May to early June and egg deposition occurs short—
ly thereafter (Madden and Chamberlin 1938, Madden and Cham-
berlin 1945). In Southwestern Michigan, egg deposition by
ML_guinguemagulata moths was common throughout July but was
virtually non-existent after early August. Larval infesta-
tion levels peaked from mid to late July, then decreased
thereafter. These trends were especially evident during
1985. From mid August to the end of the summer, not a
single M1 guinggemaculata larva or egg was obtained from 360
samples of tomato plants, suggesting that larvae developing
from early season egg deposition completed development,
pupated and then entered diapause. While not as apparent, a
similar trend was observed in 1986. Although ML quip;
gngmggulgtg larvae were found through August 28, eleven of
the thirteen total M; guinguemaculatg larvae found from
August 14 to August 28 (out of 384 plants sampled) were

fourth, pre-molt fifth or fifth instar larvae, indicating

45
that egg hatch for these late season larvae occurred in mid
July or early August. Based on these observations, it
appears that M; guinguemagulata has only one generation per
year.

In contrast to the seasonal trends of M; guigguemagu-
lata, ML sexta eggs and early instar larvae were observed
from the beginning of the season until sampling stopped in
September (Figures 2 and 3), indicating that ovipositon was
occurring throughout the season. Unfortunately, it is
difficult to determine how many generations of M; sexta
occur during a year. In 1985, there appeared to be two
distinct ovipositional periods (Figure 2 -- note increase in
eggs levels late in the season), suggesting that this spe-
cies is bivoltine, i.e., eggs deposited during the earliest
period give rise to a first generation of larvae that com-
plete development, pupate and then emerge as adults in early
to mid August. These adults then deposit eggs that develop
into a second larval brood.

However, 1986 ovipositional data supports the hypo-
thesis that genetic variability within the species produced
individuals that emerged and/or entered diapause at dif-
ferent times during the summer (Tauber and Tauber 1981,
Hagen and Lederhouse 1985). Although there were two di-
stinct ovipositional peaks, the period of time between these
was not sufficient to permit larval development, pupation
and subsequent emergence as an adult. Waldbauer (1973)

determined that emergence in gecropig moths was bimodal,

46

i.e., one group of moths emerged in late May and early June
while a second group emerged in late June and early July.
Rabb (1966) has, in fact, demonstrated bimodal emergence in
ML sexta populations in North Carolina. ML sexta adults
pupating during one year emerged at two separate times
during the following season. In addition, the first "wave"
of adults produced larvae that completed development and
emerged simultaneously with the later "wave” of adults,
producing a field population of larvae consisting of a
mixture of first and second generation individuals.

Regardless of the scenario, or combination thereof,
which more appropriately describes the seasonal occurrence
of M; sexta in Michigan, the fact that the females deposit
eggs very late in the summer would appear to be deleterious
to their overall fitness. It seems probable that these
newly hatched larvae would not be able to complete develop-
ment before weather conditions would have an adverse affect
on their growth or on that of the plant. In fact, during
this study, tomato foliage (determinant variety) was vir-
tually nonexistent by mid to late September. While no
explanation is offered for the occurrence of late season
emergence and egg deposition, this phenomenon provides an

interesting area for future investigations.

47

QQHQLQEIQH§

Results of this study demonstrate that egg deposition
and larval development can be realistically predicted by
degree day accumulation. The study also demonstrates that
initial summer emergence of adults and subsequent develop-
ment of larvae is similar for both species.

The only apparent difference between species appears
to be their relative abundance during the season.

3; guigguemaculata is more prevalent early in the season and
becomes relatively rare by mid August, indicating that u;
guinguemagglata populations are univoltine in Southwestern
Michigan. In comparison, M; sexta larvae are persistent
throughout the season and by mid to late August, are the
only species found. The occurrence of eggs and early instar
larvae throughout the summer, suggests that M; sexta popula-
tions are bivoltine, have bimodal emergence or are comprised

of some combination between these two scenarios.

REFERENCES

Baumhover AH (1985) Manduca sexta. In: Handbook of Insect
Rearing. Vol. 2. (Pritam S, Moore RF, eds.). Elsevier
Science Publishers B.V., Amsterdam.

Borth PW, Harrison FP (1984) Within-field distribution
patterns of tobacco and tomato hornworm (Lepidoptera:
Sphingidae) egg and larval populations in southern
Maryland tobacco. Environ. Entomol. 13:859-862.

Cheng HM (1977) Flue-cured tobacco losses caused by the
tomato hornworm, Manduca ggiggggmagglata (Lepidoptera:
Sphingidae), at various infestation levels in Ontario.
Can. Ent. 109:1091-1095.

de Boer G, Hanson PE (1984) Foodplant selection and induc-
tion of feeding preference among host and non-host
plants in larvae of the tobacco hornworm Manduca sexta.
Ent. Exp. appl. 35:177-193.

Gilmore JD (1938) Observations on the hornworms attacking
tobacco in Tennessee and Kentucky. J. Econ. Ent. 31:
706-712.

Hagen RH and Lederhouse RC (1984) Polymodal emergence of the
tiger swallowtail, Eapilig glaucus (Lepidoptera: Papi-
lionidae): source of a false second generation in
central New York State. Ecological Entomology 10:19-28.

Jones GA, Thurston R (1970) Leaf consumption and development
of tobacco hornworm larvae feeding on burley and dark
tobacco. J. Econ. Ent. 63:1938-1941.

Kennedy 66, Romanow LR, Jenkins SF, Sanders DC (1983)
Insects and diseases damaging tomato fruits in the
coastal plain of North Carolina. J. Econ. Entomol. 76:
168-173.

Madden AH, Chamberlin FS (1938) Biological studies of the
tomatoworm on tobacco in Florida, 1936 and 1936. J.
Econ. Entomol. 31:703-706.

Madden AH, Chamberlin FS (1945) Biology of the tobacco
hornworm in the southern cigar-tobacco district. US
Dept. of Agric. Tech. Bull. No. 986. 51 pp.

McFadden MW (1968) Observations on feeding and movement of
tobacco hornworm larvae. J. Econ. Ent. 61:352-356.

48

49

Rabb RL (1966) Diapause in Ezgtgparge sexta (Lepidoptera:
Sphingidae). Ann. Ent. Soc. Am. 59:160-165.

Semtner PJ, Rasnake M, Terrill TR (1980) Effect of host-
plant nutrition on the occurrence of tobacco hornworms
and tobacco flea beetles on different types of tobac-
co. J. Econ. Entomol. 73:221-224.

Tauber CA, Tauber MJ (1981) Insect seasonal cycles: genetics
and evolution. Ann. Rev. Ecol. Syst. 12:281-308.

Waldbauer GP, Sternburg JG (1973) Polymorphic termination of
diapause by Qggrgpig: genetic and geographical aspects.
Biol. Bull. 145:627-241.

MANUSCRIPT II: Nutritional consequences of variable plant
nitrogen: Experiments with plant fertilization and Manduca

Wage.

50

51

ABSTRACT

Tomato foliage treated at higher rates of nitrogen
fertilization and containing higher concentrations of total
plant nitrogen and soluble nitrates permitted tomato horn-
worm larvae to utilize their food more efficiently, grow
faster, suffer lower mortality and develop into more fecund
adults compared to larvae developing on foliage left un-
treated or treated with low levels of fertilizer. Larvae
reared on untreated foliage were able to partially compen-
sate for lower food conversion efficiencies by consuming
food at a higher relative rate; however, larvae reared on
foliage treated at low fertilization levels did not demon-
strate this compensatory response.

Enhanced larval performance was correlated with both
mean percent total plant nitrogen and mean percent soluble
nitrates. Plant water content was high for all treatments;
hence, probably did not influence larval growth. Other
plant components influenced by nitrogen availability and

their subsequent affect on herbivores are discussed.

INTRODUCTION

The relatively recent interest in host plant/insect
interactions has led to the realization that acquiring a
basic understanding of the mechanisms that enable an insect

to utilize its available resources, is essential to the

52
development of theory in insect ecology and evolution. This
realization has, in turn, produced a widespread interest in
the study of nutritional ecology. Much of the recent work
has focused on elucidating the biochemical properties of
host plants influencing larval development and subsequent
adult survival and performance. Results of these studies
have indicated that plant nitrogen is a key component in a
number of important metabolic pathways (see review by
Mattson 1980). Not only does this substance influence the
physiology and nutrition of the insect (Shaw and Little
1972, McNeill and Southwood 1978, Scriber and Slansky 1981,
Hagen et. al. 1984, Wint 1983), but it also affects the
general behavior of the animal (Beck 1974, White 1979, Myers
and Post 1981, van der Meijden et. a1. 1984).

Unfortunately for the herbivore though, growth and
development may be severely limited by inadequate nitrogen
reserves. First, plant nitrogen content varies throughout
the host plant and throughout the growing season (Beck 1972,
Mattson 1980, White 1984). Secondly, a large percentage of
plant nitrogen is unavailable to an herbivore (McNeill and
Southwood 1978, Reese 1978, Mattson 1980). Finally, nutri-
tional value is dependent upon the ratio, pattern and
quantity of amino acids, hence proteins, found in the plant
(Rock 1972, House 1969).

The addition of nitrogen fertilizer to agronomic crOps
is a common cultural practice that, subsequently, alters the

physiology of the plant and its value as a food resource for

53
phytophagous insects. Generally, the nutritional value of
the foliage increases as the availability of soil nitrogen
increases (van Emden and Way 1973, Fox and Macauley 1977,
Auerbach and Strong 1981, McNeill and Prestidge 1982, Al-
Zubaidi and Capinera 1984, Manuwoto and Scriber 1985).
Usually, this increased nutritional value results from
increased concentrations of soluble nitrates (Terman et. al.
1976, McNeill and Southwood 1978, Eppendorfer 1978), pre—
curssors of proteins and amino acids (Murata 1969, Hoff et.
al. 1974, McNeill and Southwood 1978, Rahier 1978), that
provide a ready source of available nitrogen for an insect
feeding on the plant. However, increased nitrogen availa-
bility also stimulates increased cell elongation and divi-
sion, generally producing foliage with a high water content,
a plant component known to have a large influence on insect
fitness (Scriber 1977, 1978, 1979).

Abundant supplies of plant nitrogen do not always
enhance insect performance. Increased nitrogen availability
may initiate the production of certain metabolic compounds
that interfere with food consumption and utilization. Plant
alkaloids, chemical compounds found in many plant families,
are such substances. Not only do these nitrogen containing
compounds generally increase with increased nitrogen availa-
bility (Geissman and Crout 1969, Waller and Nowacki 1978),
but they also often exert toxic affects on certain insects
(Tingey 1984, Sinden et. al. 1986).

This research was designed to examine the interactions

54
between host plant physiology/quality and insect perfor-
mance. More specifically, we investigated the affect of
nitrogen fertilization on plant quality and subsequent
larval development and adult survival and fecundity. Field
plots of tomato plants (var. Sunny) were treated with four
levels of nitrogen fertilizer, thus inducing varying degrees
of physiological changes in these plants. Manduca guingug;
maculata larvae, reared from egg hatch to adult emergence on
bouquets of treated foliage, were monitored on a regular

basis to determine relative fitness between treatments.

MATERIALS AND METHODS

o e o

Nitrogen fertilizer was applied directly to the soil
surrounding individual tomato plants at the rate of O, 30,
120 and 240 lbs./acre. Ammonium nitrate (34% nitrogen)
provided the source of nitrogen. To prevent burning the
plants receiving high levels of fertilizer, applications
were split between July 3 and July 14. The first applica-
tion of nitrogen occurred approximately 2 1/2 weeks after

seedlings were transplanted into the field.

QIQEID_§§B§1
Two growth studies were performed utilizing Manduca

guinguemagglata larvae reared from eggs received from the

Oxford Tobacco Research Station in North Carolina. Although

55

this culture had been maintained for at least five years
previous to this study, it was assumed that any inbreeding
of the population would not affect relative differences
between treatments. The first study examined general larval
development and emergence into the adult form. The second
study was designed to examine larval growth in detail by
measuring different parameters and then calculating several
nutritional indices. Originally, both studies were started
at the same time. However, after two failed experiments (a
result of technical errors), the detailed performance study
actually began 24 days later. Altogether, 360 newly hatch-
ed, first instar larvae were randomly placed on foliage
treated with four levels of nitrogen fertilizer (ninety/
treatment). The larvae were initially put into individual
50 m1 plastic containers but were later transferred to 1000
ml plastic cups to allow ample room for continued growth.

Rearing facilities were maintained at ambient air
temperature and relative humidity to simulate field condi-
tions. Temperature remained fairly constant throughout the

study, fluctuating between 24 and 30 degrees C.

er d
Sixty larvae of the initial ninety/treatment were
reared from egg hatch to subsequent emergence as an adult.
Each day, fresh tomato foliage was presented to the larvae

and uneaten plant material and frass were discarded. Larvae

56
entering the prepupal or "wandering stage" were placed into
clean cups and allowed to complete pupation. Six days after
pupation, 9 males and 9 females from each nitrogen treatment
were randomly selected and oven dryed at 60° C. The re-
maining pupae were sexed then left to emerge naturally.
Small sticks (approximately 7 cm high) were placed upright
in each container to provide a surface for newly emerged
moths to crawl up on. To be consistent, pupae were checked
for emergence every day at 10 pm (the moths always emerged
in the late evening). Newly emerged, adult males were
recorded and then discarded. Adult females were starved for

three days, then dissected so that eggs could be counted.

e e e o nce 8 ud

The other thirty larvae from each nitrogen treatment
were used to calculate performance indices (after Waldbauer
1968). These measurements required that five sets of pa-
rameters be determined: a) initial animal weight, b) final
animal weight, c) weight of food provided, d) weight of food
remaining, and e) weight of feces accumulated. Originally
these parameters, measured as grams dry weight, were deter-
mined for first instar larvae. However, because of the
difficulty of obtaining precise measurements on such low
weights (especially dry weight of feces) and high mortality
that occurred during the first two failed experiments in
which first instars were utilized, the nutritional indices

were ultimately calculated for fourth instar larvae.

57

These indices are defined as follows:

1. Relative Consumption Index, (RCI) = F/(T)(A),
calculates the rate of feeding per day relative to

the mean weight of the animal.

2. Relative Growth Rate, (RGR) = G/(T)(A), measures
the daily rate of growth of the animal relative to

its mean body weight.

3. Approximate Digestibility, (AD) = (I - FR)/I
x 100, examines the animal's efficiency for converting

food consumed into food digested.

4. Efficiency of Conversion of Ingested food,

(ECI) = G/I x 100, calculates the overall efficiency
of an animal to utilize an ingested food for its own
biomass. This index is dependent upon the digesti-

bility and nutritional quality of the food.

5. Efficiency of Conversion of Digested food,

G/(I - FR) x 100, measures the overall efficiency of
an animal to convert digested material into actual
body mass. This index is also dependent upon the

digestibility and nutritional quality of the food.

58

where,

= dry weight of food eaten,
8 duration of feeding period,

mean dry weight of animal during feeding period,

0 ? #3 "'1
II

dry weight gain of animal during feeding period,

I dry weight of food ingested, and

FR a weight of frass produced.

Pre-molt fourth and fifth instar larvae were monitored
every three hours between 6 am and 12 midnight (pre-molt
larvae are easy to distinguish since their head capsule
becomes extended from the rest of their body). Newly molted
fourth instar larvae were fed pre-weighed foliage collected
daily from field grown tomato plants. Frass and uneaten
foliage was collected every 24 hours and oven dryed at
60° C. Newly molted fifth instar larvae were immediately
removed from the experiment, frozen, then oven dryed.

Initial dry weight of tomato foliage fed to larvae was
determined by calculating the mean percent dry weight for a
control group of foliage from each treatment each day and
then multiplying the fresh weight of foliage actually pre-
sented to larvae by this number. This mean was obtained by
randomly selecting five freshly picked aliquots of leaves
from each treatment from foliage collected each morning and
then measuring the fresh and dry weights of these aliquots.

The values obtained were also utilized for calculating mean

59
percent water loss for each treatment.

Initial larval weight was determined by randomly
selecting a control group of eight newly molted fourth
instar larvae from each treatment and obtaining their dry
weights. Larval weight during the fourth instar was calcu-
lated by multiplying the initial larval weight by the final
larval weight, then taking the square root of this number
(geometric mean).

All weights utilized in the calculation of the perfor-
mance indices were determined on an analytic balance. The
values obtained were carried to three decimal places.
Performance indices for individual larvae were calculated by
combining all measurements for each parameter, over the
course of the study, into one total measurement, i.e. frass
collected each day from each individual larva was combined

into one total dry weight of frass produced by that larva.

le t o and Chem'c sis

Foliage fed to larvae was collected fresh from the
field grown tomato plants early each morning. Similar
aliquots of leafs or leaflets were picked. Generally, these
aliquots consisted of the first fully expanded leaf from the
terminal bud. Foliage from each treatment was placed in a
separate plastic bag and stored in a cooler. Foliage uti-
lized in the detailed performance study was immediately
weighed and presented to larvae as soon as possible after

being collected. Every three to four days, four aliquots of

60
leaves were randomly selected from each bag of leaves and
oven dryed at 60° C. These samples were used to determine
total nitrogen and soluble nitrates in tomato leaves
throughout the course of the study.

Prior to analysis, all foliage was ground using a 40
mesh screen. To determine total nitrogen, 150 mg of dried
foliage was digested with Rel-catalyst KC-Mc (3.5 g potas—
sium sulphate + 175 mg mercuric oxide) and 9 ml sulfuric
acid for one hour. After cooling for 30 minutes, the solu-
tion was brought up to a volume of 75 ml with distilled
water. Total ppm nitrogen was read using a Technicon Auto
Analyzer II.

To determine soluble nitrates, 20 ml of distilled
water was added to 500 mg of dried foliage and then shaken
for 30 minutes. Total soluble nitrates were assessed using

a nitrate electrode and an expanded scale pH meter.

RESULTS

a ses

During both years of the study, plants which were
treated with high levels of fertilizer produced foliage that
was visibly greener (although not chemically analyzed) than
leaves on plants treated with low levels of fertilizer or
left untreated.

Chemical analyses conducted on foliage collected

during this study demonstrated that mean percent dry weight

61
total plant nitrogen (Table 1) was positively correlated
with nitrogen fertilization treatment during both the gene-
ral and detailed growth studies (r = .95, r = .93 respec-
tively at p = .05). Figure 1 depicts the seasonal trends of
this plant component. Even though initial levels of nitro-
gen were relatively high for all treatments, concentrations
declined quickly and then leveled off throughout the remain-
der of the study.

Mean percent soluble nitrates (Table 1) was found to
be positively correlated with fertilizer treatment during
the general study (r = .98, p = .01) but not during the
detailed study (r = .92, p = .08). Each level of fertiliza-
tion resulted in an initial increase in nitrates that stead-
ily declined during the remainder of the season (Figure 2).
For plants treated at the high levels, this decline spanned
a period of four to five weeks. However, for plants treated
with 30 lbs. nitrogen per acre, decline occurred within 12
days. By the second week of August, nitrate levels for all
treatments were at very low levels.

Percent water content analyzed during the Detailed
Growth Study was high for all treatments (80 - 88%).
Although only slightly higher, foliage treated with 30
lbs. nitrogen per acre had the highest water content and
foliage treated with 240 lbs. nitrogen per acre had the
lowest. Results of a correlation analysis determined that
mean percent water content was negatively correlated with

mean percent plant nitrogen (r = -.95, p < .04). Figure 3

62

 

 

$5 8.8 can
:83 2.3 . ca:
5.3 9.3 <2 8 53:8
33 3.3 o as?
888 one. 5.3 8: ova
885 as. 28.8 as. 8—
:88 n3. 8:: 83. on 8.82
$8.8 :8. G88 :8. c 8328
:33 A} 63.8 2... SN
33 as :83 83. cm:
68.8 T." 55 8.... 8 E882
$3 88 68.8 9.." o as: 58.
c. u 5 e. n 5 §8=8<
5390 8:8: 5:20 .3200 8882 .mm:

 

>§m

 

88:83:: ES: 833:8 Go 5:850:00 05

:o 8:: $8.52 83:5 eons—Baa a: 5 Baas: 203 82? 89C. .634. 05 Co ~88: 8288 :23 mg:
8538 «SEE 3:29. :5 :8ng as: .89 38:8 833 .588 :8 828 8353 :5 «502 A 033.

63

. fiafifizaazgoauacaouazggungussoiécn
«83°88 2382.38.00 683538050: a»§a.fi§5.§fi%§8§ .33:

w><o mum—Gun. §§OO<
oovm comm ooom coma . com: 03; com;

 

 

 

 

 

 

— LP by D — [P b D b h bvllbylr P b n L. - D.- L— l- L—, bl L! b 'O
i
1N E
V H
o
. :—
iv m.
.. 1
a . m
. v
. __|
.© .N..
l
, Ha
. o
. m
.w N
a
.883 To:

 

64

. .88 a: 33% z .2: 8n n as». 5% :5 z .2: 8: a 3.26 88.0 .z a:
on u 83% 5% .z .3. c u 5: 886 .82 2:2. 38:8 guano: 3.5:. 0328 .583 .6 ”BB 1833. N pa»:

 

m><o mmmomo om...<.5<500<
03m 8% 88. com: 82 83 com:
. r . . _to.o
3.0
..m.o

1°.
0
SON .LNBOHBd

 

 

 

.8203 A 52 Too.

6S

 

.08: he: 33%: 2 an: ovm n 3.8m :30 :5 Z an: 03 n 88:6 come—o . Z an: on .I. 88:6 :2“: .2 .ma 0 u
Em U820 Scam oufigotom 3:809 wfibé 363—8 amazom .35" 53%.: :80 3 :89: ES: 88.89 :82 .m gum

w><o mmmwwo Dm._.<._32300<

 

 

comm mnfim omfim mmwm oowm mpmm comm mmmm comm

prhhk_:_._gp,.~_p»._b.:___.L.___.,.____._Iow
“Sm
H V
. N
éwx
. .I
. 3
H V
-8...
H M
. V
53....
. uu
wow

 

66
depicts the levels of this component during the Detailed
Growth Study.

e er ance Stu

Larval survival during this study was approximately
the same for all treatments (70, 67, 62 and 67% for 0, 30,
120 and 240 lbs. nitrogen applied/acre respectively). Most
mortality occurred during the first and second instar and
probably resulted from daily handling of larvae. A G-Test
for independence determined that percent survival was not
associated with fertilization treatment (G = .47, p > .92).

A linear regression on individual observations of male
(Figure 4) and female (Figure 5) pupal dry weights was
highly significant (p < .0002 and .0001 respectively) al-
though, only a small amount of variability within a treat-
ment was explained by nitrogen fertilization. In both
cases, the smallest pupae developed from larvae reared on
foliage treated with 30 lbs. nitrogen/acre and the heaviest
from foliage treated with 240 lbs. nitrogen/acre. Overall,
pupal weights were heavier for females than for males.

A linear regression on percent successful emergence of
adult males from the pupal stage did not produce a slope
significantly different from zero (p = .46). In fact,
emergence of males reared on untreated foliage and foliage
treated with 120 and 240 lbs. nitrogen/acre was very similar
(Figure 6). However, foliar nitrogen, or lack thereof,

appeared to have an affect on percent emergence of males

67

83:8: .8 5:85:98 :0: :8: 8m was .3 386:: 82g :82 :05—
:oon: :80 a o u z .3. u «m .88. v : .mfimmoa he was: 03:: Mo :ommmoao: 53:5 .v pawE

wmo<\om_._n_n_< 2 mm:
3m com 8: om: on o...

F . b p _ r _ p _ : _ p _

O

 

 

V

0.0209
HOO

ABC] SWVHE)

U

'000'
v-‘
1-4

 

r 0' O
N
v—t

5'0.
9M

fifi: '
v-fl
'J.

 

.mé

68

‘ .8393» .«o aonaqaaxo H8 “x2 com .33» E Bowman 83g 532 49,3
5ng some 3 a n z .3. u «m .80. v a gamma? 5 19a ciao.“ .«o .868“on 385 .n 05mm

mm0<\om_.._n_n.< 2 mm.—

 

 

ovm com om: ofi on 3 o
. % _ p P _ p p _ L — r I®.O
fl
mm o w
.04 V
.. W
(gin 8
.fi 0
fig m
f
23 M
fi 9
.31
+0.“

 

69

 

100

4 MALE

90*

 

80‘

 

 

70“

60‘

PERCENT EMERGENCE

 

 

 

 

 

 

 

 

m
0 30 120 240

LBS N/ ACRE

Figure 6. Percent successful emergence of adult males from
larvae reared on tomato foliage treated at four different nitrogen
levels. A linear regression of arcsine transformed data groduced
a $10 that was not significantly different from zero (R = .46,
p < . 9). See text for explanation of statistics.

 

50

7O
reared on foliage treated with 30 lbs. nitrogen/acre. In
contrast, percent successful emergence of adult females from
the pupal stage demonstrated an obvious relationship with
level of fertilization, suggesting that female emergence was
highly dependent upon nitrogen content of foliage fed to
larvae (Figure 7).

Female fecundity was significantly influenced by rate
of fertilization (Figure 8). A linear regression on egg
totals from individual females produced a slope that was
highly significant (p < .009). However, there was a great
deal of variability within a treatment, especially for
females reared on foliage treated with 30 lbs. nitrogen/
acre, and nitrogen treatment only accounted for 17% of the
variability.

Host General Growth parameters significantly
influenced by nitrogen fertilization level were also corre-
lated with mean percent dry weight plant nitrogen and mean
percent soluble nitrates (Table 2--mean percent plant nitro-
gen and soluble nitrates calculated from foliage collected
while each separate larval cohort was developing--see Table

1).

d Fe 0 a c tu
Of the initial thirty larvae per treatment started in
this experiment, survival was highest for larvae developing
on foliage treated with 120 and 240 lbs. nitrogen/acre (97

and 100% respectively--Figure 9). In fact, only 26 of the

71

 

100

FEMALE

 

90‘

 

80‘

 

70‘

60‘

PERCENT EMERGENCE

 

 

 

 

 

 

 

 

 

 

 

50
O 30 120 240

LBS N/ACFlE

Figure 7. Percent successful emergence of adult females from
larvae reared on tomato foliage treated at four different nitrogen
levels. A linear regression of arcsine transformed data produced
a 310 significantly different from zero (R2 = ..97 p < ..O2) See
text or explanation of statistics.

72

.95..“ .3
uncommon 82g 582 496— 8ng 53 8 2 u z .80. u m .S. u «a fiasco“ £35m mo sump—we H854 .w 85mm

mmo< awn. hum—Jan? 2 mm.—
ovm oom o3 omfi on o... o

— bf L! P — b b P _ P L by l—l LP b P .F r n P — L p r p o

 

EWVWSA Had $993

 

73

”89:88 3.2.65 8 303:8 83359

 

 

an. a: G8 68¢ 28c 63 ”3.52
9.. 3. a. 8. 8. 3. 0328 e
33 93 38¢ A88.V $3 23 5352 :33
8.- 3 S. R. 8. 3.. £303 in s
36.3 535$ 035". «32 oaaam ca:
3.5 e 035m Lac? 853.5
to as: 23.8% e

 

.85u5=w_m ..o ~26. 332083 885% 5 ”.3852 .couSBBu
.8 806508 «beacon Bofi €385: .«o 32 98. .355 5320 .2050 05 Eat 33:28 30888
one—BEBE RES 55 mag? 0338 93 5ng Ema Emma: be E82 :88 .8 couSpcoU .N 033.

74

 

100

 

90‘

80l

 

7O ‘

7. LARVAL SURVIVAL

60‘

 

 

 

 

 

 

 

 

 

 

 

50
O 30 120 240

LBS N/ ACRE

Figure 9. Percent survival of larvae reared from egg hatch to
premolt fifth instar on foliage treated at four rates of nitrogen
fertilization. A linear regression of arcsine transformed data
determined that survival was associated with fertilization level
(R2 = .95, p = .02). N = 30 larvae per treaunent.

75
30 larvae reared on untreated foliage and only 25 reared on
foliage treated with 30 lbs. nitrogen per acre even reached
fourth instar. A linear regression of arcsine transformed
data determined that larval survival from egg hatch through
fourth instar was highly associated with fertilization
treatment.

The number of days required for larvae to reach newly
molted fourth instar (Figure 10) was slightly shorter for
larvae reared on foliage treated at the highest fertiliza-
tion rate (7.33 i .10); approximately a half day quicker
than individuals reared on foliage treated with 120 lbs.
N/acre and untreated foliage (7.90 i .15 and 7.76 i .15 for
0 and 120 lbs. nitrogen/acre respectively). Most inte-
resting though, larvae reared on foliage treated with 30
lbs. N/acre required approximately a day longer to reach
this stage (8.89 i .27).

An analysis of variance, utilizing the general linear
method procedure (SAS, 1985), was performed on all nutri-
tional indices. Since AD, ECD and ECI calculate relative
percentages, these particular indices were converted to
their arcsine transformations prior to analysis. In addi-
tion, AD, ECD and ECI were analyzed using a weighted regres-
sion (Steel and Torrie 1980) to correct for heterogeneity of
variance.

Results (Table 3) demonstrated significant differences
between treatments for many of the nutritional indices.

Only final larval weight and relative foliage consumption

76

 

 

 

 

 

MEAN DAYS TO FOURTH

 

 

 

 

 

 

 

 

 

 

 

O 30 120 240
LBS N APPLIED/ ACRE

Figure 10. Means and standard errors for days to premolt fourth
instar. F = 14.04, p < .0001. N = 26, 25, 30 and 30 for 0, 30,
120 and 240 lbs. nitrogen applied/acre respectively.

 

77

.88 5 5% 558 .o 833 n 25...

 

 

.é no
onmESoa ..8 89:8 9 to»: 88:9. we as.» 32303 .53 38.828 0588 ..o 8:52 330..

Auve

6:: . mm; #03 A<ZE

No. V a find 3.82:.
:8. v e

m 8 Rd 58

88. v a :8 mom

88. v a 8.: ..6m

88. v a 2: .8m

8. v a an .3

m 383$ 865

Doggorom

 

 

.889: Ho 88888 e5 83888. ..o gage .8 «850:. Ea
«inane 8m .Go. a a a 852.85% .9. .1. «.5 8385mm 8o 3:35:me 5858 5ng .0 £26—

.38 55 3.38 0938 cc v28.— 85 .8065 855.88g :35 co 3:85, .80 «was... .m 038.

78
were not significantly different at p = .05, although dif-
ferences in relative consumption were significant at p =
.07.

Student-Newman-Keuls mean separation of these results
(Table 4) showed that larvae reared on foliage treated with
120 and 240 lbs. nitrogen/acre converted both digested and
ingested food more efficiently (ECD and ECI) than larvae
reared on the other two foliar treatments. Larvae devel-
oping on these other treatments utilized digested plant
material with the same approximate efficiency; however,
larvae developing on untreated foliage were more efficient
at converting ingested food than larvae developing on
foliage treated with 30 lbs. nitrogen/acre. Tomato plants
with lower fertilization levels produced larvae that had the
highest digestion efficiencies (AD). Overall, larvae that
were more efficient at converting digested plant material
into actual body weight demonstrated lower digestion effi-
ciencies (Figure 11).

Larvae developing on foliage treated at the highest
rate of fertilization grew at a faster relative rate (RGR)
(Table 4) and reached premolt fifth instar in fewer days
than all other larvae (Figure 12). Larvae reared on foliage
treated with 30 lbs. nitrogen/acre grew at the slowest
relative rate and required the longest amount of time to
reach premolt fifth instar. The difference in mean develop-
ment time between larvae on these two treatments was equal

to one day. In contrast, larvae reared on untreated foliage

  

:1::.;:...§.._2“ "h

79

panacea—causes"?

so... 95 238 :38 canon Esofi 839, .3532 8.8 85883 0388 :o 3585a 3352......

2.6.2.8. essences ..e as as as .8 .o a. a 9a 8 .3 .2 u z.

 

 

so: 83 and a: $3 a:

o 2.”... a 8.8 a 2.3. a 3.. a S... a n3 ova
a; $3 God a: 88 63

a 2.9 u an: a $9. a 8+ a who a as o2
$6 and no.0 63 63 G8

a 8.3 a 3.: a 2.: a 3,. a Gd :3 8
ass 83 88 63 $3 $3

a was a 8.2 a 2.: a as a who a as o
:9. ..6m ..8m mom mom Ewe 5B8?

Ems 55E .5sz and

mo cousin—

 

8823 so 838.? as 823 so sausage 3. 8232
Ba area: now .928 no. u a a 898% 358558 8: 08 8:3 088 05 55 282 .8298 5ng
329.858 55§o§8 83539.3 gogvggaoggufig €038.

80

88:63“? 2 .8: com u was... some use 2 .me o2 n 8380 383 .z .2: cm n «.3080
some .2 an: o mam 385 .888: Mo confines .8 E8 3m .obexpouada compass an: 03 use o2 dm 6 “8 mm 88 cm
.2 .2 n Z do: u E80508 :83?th Mcofiaom .coumoma usafimxeomx‘ 88 BM 5933 coufioboo .325 .: 85E

ZO_._.mm0_D wk<$=x0mni<

 

om cm on co co ow om.
F t _ _ r Li p L F . _ p t r t L _ p . t P L _ . p _ p t L FL
.0
CO H
0 2:
i H
M. o pom
w « l
04. Q s
« a .om
«a 8 a *«a 00 . a
O
a.
o. 3...... .0. a .2. a
«Mao o. . W
a «a... ’.0 Von
a o .
aco.w H
a a a. ..oo
8. .
on ”on

 

81

 

14.

 

131 I

12;

 

11‘ T

 

 

10‘.

MEAN DAYS TO FIFTH

 

 

 

 

 

 

 

 

 

 

 

O 30 120 240
LBS N APPLIED/ ACRE

Figure 12. Means and standard errors for days to remolt fifth
instar. F = 9.52, p < .0001. N = 15, 18 20 and 2 for 0, 30,
120 and 240 lbs. nitrogen applied/acre respectively.

 

82
and foliage treated with 120 lbs. nitrogen acre grew at the
same relative rates, although larvae on untreated foliage
required approximately one third of a day longer to reach
fifth instar.
Even though differences in relative consumption (RCI)
were not significant at p = .05, larvae reared on foliage
left untreated or treated with a low level of nitrogen
fertilizer consumed slightly more plant material relative to
their body weight (Table 4). This was especially apparent
for those larvae developing on untreated foliage. E
Table 5 lists Pearson's correlation coefficients and 1
corresponding level of significance for all growth param-
eters calculated from data collected during the Detailed
Growth Study. Most growth parameters significantly in-
fluenced by level of fertilization were also correlated with

total plant nitrogen, soluble nitrates or plant water.

DISCUSSION

Results of this study are mostly in agreement with the

 

literature. Increased rates of fertilization were corre-
lated with increased concentrations of total nitrogen and
soluble nitrates, which subsequently enhanced larval growth
and development and adult fitness. Larvae reared from egg
hatch on foliage treated at the two highest rates of fer-
tilization utilized their food more efficiently and devel-

oped into heavier pupae and more fecund females. In

83

.859, :85 808 358—8 acorn—0.80 =<

 

$3 83 $3 Goo $88 83 $3 80:80

3.. me. am. no: 3.. 3.- me. has?

A: .V 8o. $5 8.3 83 Gog 33 8.32

am. mm: on... ow. 5. mm. mm: 0338

83 83 :3 £3 E: 83 83 5352

8. 3.. ow: 3. ma. 8. 8.. .58 38.

3533 505 2 8m Um mom mom

.95 5» 550m 8o
c5355

 

.8065 .8 85:83 8 is 00m 60505550 .3 _0>0_ 0.30888 0085:28 5 58.32
392050 .8 806508 0.8058 397. e353 .8 32 n8. Scam 5380 3:88 05 :58 33328 88g
00558.8 its .23 885: 0328 v5 535: 5503 he 4:088 has? .588 50:. 8o 5530.50 .n 038.

84
comparison, larvae reared on untreated foliage or foliage
treated at a very low fertilization rate converted food less
efficiently and developed into smaller pupae.

Differences in insect performance were especially
apparent for larvae developing on foliage treated with 30
lbs. nitrogen/acre and those developing on foliage treated
with 240 lbs. nitrogen/acre. In both the detailed and
general growth experiments, larvae reared on foliage treated
at the high fertilization rate showed the greatest overall
fitness while those reared on foliage treated at the low
level were the least fit. Larvae reared on foliage treated
with 30 lbs. of nitrogen required, on the average, two
additional days to grow from egg hatch to pre-molt fifth
instar, a response probably resulting from their low rate of
foliage consumption and utilization.

Initially, it was assumed that untreated foliage would
contain the lowest levels of total nitrogen and soluble
nitrates: hence, produce the slowest developing and least
efficient of all larvae. Surprisingly though, larvae devel-
oping on foliage treated at 30 lbs. nitrogen/acre, consis-
tently had lower performance values than larvae reared on
untreated foliage, even though this foliage contained in-
creased concentrations of total plant nitrogen and soluble
nitrates (Table 1; Figures 2 and 3). During the general
growth study, calculated values were lower for percent
successful emergence of adult males, and pupal dry weights

of both males and females (Figures 7, 8 and 9). In

85
addition, performance indices calculated during the detailed
growth study, showed that these larvae were less efficient
at utilizing ingested food and, as already mentioned, re-
quired a longer amount of time to reach both premolt fourth
and fifth instar (Figures 10 and 12).

The discrepancy in performance between these two
groups of larvae may partially be explained by the higher
relative consumption rate of larvae feeding on untreated
foliage (Table 4). During each day of the detailed
consumption study, these larvae consumed an average of .6
grams of foliage more per day. This observation suggests
that, even though larvae reared on untreated foliage did not
utilize their food efficiently (compared to larvae devel-
oping on foliage treated at the higher rates), they par-
tially compensated for this lack of efficiency by consuming
more food per gram of body weight per day. In fact, it
appears that this compensatory response enabled larvae
developing on untreated foliage to grow at the same relative
rate as larvae developing on foliage treated with 120 lbs.
nitrogen/acre. Similar types of compensatory behaviors have
been demonstrated in other insects (Slansky and Feeny 1977,
Slansky 1982). Scriber and Slansky (1981) indicated that a
number of larval species adjust their feeding rates in
response to decreasing consumption efficiencies, an adapta-
tion apparently demonstrated during this study.

It is uncertain why larvae developing on foliage

treated at 30 lbs. nitrogen/acre did not increase

 

 

86
consumption to compensate for low food utilization efficien-
cies. However, this observation, in itself, suggests that
foliage treated with 30 lbs. nitrogen/acre probably con-
tained certain repellent compounds or lacked stimulatory
ones. One possible explanation for the observed response
might be that very slight increases in plant nitrogen pro-
duced an imbalance in the chemical composition of the plant
by decreasing the substances which elicit feeding behaviors
or permit growth.

Nitrogen fertilization directly influences herbivore
nutrition by increasing concentrations of amino acids, hence
proteins, found in the plant. However, due to its interac-
tion with other plant metabolites, availability of this
element can indirectly influence herbivore performance. Two
plant factors affected by nitrogen availability and having a
large affect on insect nutrition are plant water and plant
alkaloids.

Earlier studies have demonstrated a positive corre-
lation between nitrogen concentration and plant water con-
tent. However, results of this study do not support these
earlier findings. During the two weeks that leaf water
content was determined, the level of this plant component
was highest in untreated foliage and foliage treated with 30
lbs. nitrogen per acre. However, since percent leaf water
content was high for all treatments (80 - 88%), it seems
unlikely that water supplies were ever inadequate or at a

level that inhibited larval growth. Scriber (1979) indi—

87
cated that leaf water contents of 75 - 90% would not limit
herbivore growth. It is interesting though, that larval
fitness, was highest on foliage containing the lowest per-
centage of plant water. One hypothesis is that lower water
levels may have concentrated available nutritional sub-
stances (which, due to higher fertilization rates, were
probably higher to begin with) and hence, enhanced larval
performance. Conversely, higher percentages of plant water
in foliage left untreated or treated with low levels of
fertilizer may have diluted the nutritional substances in
the plant (which, due to lower fertilization rates, were
probably lower to begin with) hence, inhibited larval growth
(White 1984, Mattson and Haack 1987).

Lygopegsicon gsculegtum, the plant species utilized in
this study, contains the steroidal glycoalkaloid «stoma-
tine. This alkaloid, composed of an aglycone and a glyco-
side, contains nitrogen as a structural component; thus,
will be affected by nitrogen availability. Generally, plant
alkaloids increase as plant nitrogen increases. However,
because this particular alkaloid has a high C:N ratio,
concentrations of this substance may actually decrease with
increasing availability of plant nitrogen. When plant
nitrogen is abundant, carbohydrates and other carbon con-
taining compounds are utilized to synthesize nitrogen accep-
tors (Selvendran and Selvendran 1973) thus, depleting the
reservoir of carbon utilized to produce glycoalkaloids.

Results of experiments conducted on figlangm glycoalkaloids

88
demonstrated that an increase in plant nitrogen resulted in
a depletion of the carbohydrate pool and a subsequent de-
crease in glycoalkaloid synthesis (Nowacki et. al. 1975,
1976).
The affect of glycoalkaloids on u; e l t are
not well known. Yamamoto and Fraenkel (1960b) hypothesized

that plant glycosides, commonly found in the host plants of

this insect, stimulate feeding; although, this theory re-

mains to be proven. However, even if these compounds ini- l
tiate feeding behaviors, it is reasonable to assume that at f
high concentrations they exert a toxic affect on the her- L

bivore. Plant alkaloids are well known for their toxic and
repellent effects on certian insects (Harley and
Thorsteinson 1967, Nowacki et. al. 1976), even those that
are specialists on hosts containing these compounds (Tingey
1984, Sinden et. al. 1986). Based on the preceeding discus-
sion, it seems unlikely that plant glycoalkaloids increased
to a level, in any of the treatments, that would have been
toxic to u; guinguemaculatg larvae. Although not chemically
analyzed, it is likely that alkaloid concentrations
decreased as plant nitrogen increased. In fact, the respon-
ses observed in larvae reared on foliage treated with 30
lbs. of nitrogen per acre may have partially been due to
slight increases in the level of alkaloids in these plants,
i.e., small increases in nitrogen availability may have been
shunted to metabolic pathways producing alkaloids and away

from those synthesizing proteins.

89

The ecological implications of this research are
important. Insects with a higher overall fitness will in-
crease their chances of making a genetic contribution to the
next generation. Higher survival rates during both the
larval and the pupal stage, will increase the number of
individuals that emerge as adults and hence, the number
available to mate. In addition, heavier pupal weights will
produce males and females that have a longer life span and a
greater reproductive potential.

The economical considerations are also significant.
The application of nitrogen fertilizer is a common cultural
practice in agricultural systems. Since higher levels of
fertilization can result in insects with faster growth rates
and increased fitness, the potential for population out-
breaks will also increase. The crop damage caused by these
outbreaks could far outweigh the economical advantages of
increased yields and may result in considerable losses to

the grower.

conclusions

M; guingggmagulgtg fitness is significantly enhanced
by the application of high levels of nitrogen fertilizer to
its host plant. Increased nitrogen availability directly
influences both larval and adult performance. Conversion
efficiencies, growth rates, pupal weights, fecundity and
survival are significantly correlated with rate of

fertilization. Nitrogen availability also indirectly

90
affects insect fitness. Since larvae develop faster on
foliage treated with high levels of fertilizer, exposure
time on the plant is shorter and therefore, the probability

of attack by natural enemies will decrease.

REFERENCES

Al-Zubaidi FS, Capinera JL (1984) Utilization of food and
nitrogen by the beet armyworm, Spgdgptgra exiggg
(Hubner) (Lepidoptera: Noctuidae), in relation to food
type and dietary nitrogen levels. Env. Ent. 13:1604-
1698.

Auerbach MJ, Strong DR (1981) Nutritional ecology of
Heliconig herbivores: experiments with plant fer-
tilization and alternative hosts. Ecological Mono-
graphs 51:63-83.

Baumhover AH (1985) Manduca sexta. In: Handbook of Insect
Rearing. Vol. 2. (Pritam S, Moore RF, eds.). Elsevier
Science Publishers B.V., Amsterdam.

Beck SD (1972) Nutrition, adaptation and environment.
In:Insect and Mite Nutrition. North-Holland Publishing
Co., Amsterdam.

Beck SD (1974) Theoretical aspects of host plant specificity
in insects. In: Proc. Summer Institute on Biological
Control of Plant Insects and Diseases. (Maxwell FG,
Harris FA, eds.). Univ. Press Miss., Jackson, Miss.

Eppendorfer WH (1978) Effects of N-fertilisation on Amino
Acid Composition and Nutritive value of spinach, kale,
cauliflower and potatoes. J. Sci. Fd. Agric. 29:305-
311.

Fox LR, MacCauley BJ (1976) Insect grazing on Eucalyptus in
response to variation in leaf tannins and nitrogen.
Oecologia. 29:145-162.

Geissman TA, Crout DHG (1969) Organic Chemistry of Secondary
Plant Metabolism. Freeman, Cooper and Co., San Fran-
sico, California.

Gilmore JD (1938) Observations on the hornworms attacking
tobacco in Tennessee and Kentucky. J. Econ. Ent.
31:706-712.

Hagen KS, Dadd RH, Reese J (1984) The food of insects. In:
Ecological Entomology. (Huffaker CB, Rabb RL, eds.).
John Wiley and Sons, Inc.

Harley KLS, Thorsteinson AJ (1967) The influence of plant
chemicals on the feeding behavior, development and
survival of the two-stripped grasshopper, Melanoptgs
biyittatgg (Say), Acrididae: Orthoptera. Can. J. Zool.
45:305-319.

91

92

Hoff JE, Wilcox GE, Jones CM (1974) The effect of nitrate
and ammonium nitrogen on the free amino acid composi-
tion of tomato plants and tomato fruit. J. Am. Soc.
Hort. Sci. 99:27-30.

House HL (1969) Effects of different proportions of nu-
trients on insects. Ent. exp. & appl. 12:651-669.

Madden AH, Chamberlin FS (1945) Biology of the tobacco
hornworm in the southern cigar-tobacco district. US
Dept. of Agric. Tech. Bull. No. 986.

Manuwoto S, Scriber JM (1985) Consumption and utilization of
experimentally altered corn by southern armyworm:iron,
nitrogen and cyclic hydroxamates. Journal of Chemical
Ecology. 11:1469-1483.

Mattson WJ (1980) Herbivory in relation to plant nitrogen
content. Ann. Rev. Ecol. Syst. 11:119-161.

Mattson WJ, Haack RA (1987) The role of drought in outbreaks
of plant-eating insects. BioScience. 37:110-118.

McNeill S, Prestidge RA (1982) Plant nutritional strategies
and insect herbivore community dynamics. Proc. Fifth
Int. Symp. Insect-Plant Relationships. PUDOC,
Wageningen.

McNeill S, Southwood TRE (1978) The role of nitrogen in the
development of insect/plant relationships. In:Biochem-
ical aspects of plant and animal coevolution. (Har-
borne JB, ed.). Academic Press. London.

Murata Y (1969) Physiological responses to nitrogen in
plants. In: Physiological Aspects of Crop Yield
(Eastin JD, ed.) Am. Soc. Agron. Madison, Wis.

Myers JH, Post BJ (1981) Plant nitrogen and fluctuations of
insect populations: a test with the cinnabar moth -
tansy ragwort system. Oecologia 48:151-156

 

Nowacki E, Jurzysta M, Gorski P (1975) Effect of availa-
bility of nitrogen on alkaloid synthesis in Solana;
ggag. Bull. Acad. Pol. Sc. Ser. Sc. Biol. 23:219-225.

Nowacki E, Jurzysta M, Gorski P, Nowacka K, Waller GR (1976)
effect of nitrogen nutrition on alkaloid metabolism in
plants. Biochem. Physiol. Pflanzen. 169:231-240.

Rahier M (1978) Performance of Myzus persigge and production
of its host plant, Brassica raga, related to plant
mineral nutrition. Ent. exp. & appl. 24:421-424.

93

Reese JC (1978) Chronic effects of plant allelochemics on
insect nutritional physiology. Ent. exp. & appl.
24:425-431.

Rock GD (1972) Optimal proportions of dietary amino acids.
In:Insect and Mite Nutrition. (Rodriquez JG, ed.).
North-Holland Publishing Co., Amsterdam.

SAS User's Guide: Statistics (1985) SAS Institute Inc. Cary,
NC

Scriber JM (1977) Limiting effects of low leaf-water content
on the nitrogen utilization, energy budget, and larval

growth of fixalgnhgra segrenia (Lepidoptera: Satur-
niidae). Oecologia. 28:269-287.

Scriber JM (1978) The effects of larval feeding specializa-
tion and plant growth form on the consumption and
utilization of plant biomass and nitrogen: an ecolo-
gical consideration. Ent. exp. & appl. 24:494-510.

Scriber JM (1979) Post-ingestive utilization of plant bio-
mass and nitrogen by lepidoptera: legume feeding by
the southern armyworm. NY Entomol. Soc. 87:141-53.

Scriber JM, Slansky F (1981) The nutritional ecology of
immature insects. Ann. Rev. Entomol. 26:183-211

Selvendran RR, Selvendran S (1973) Chemical changes in young

teaplant (Camellia sinensis L.) tissues following
application of fertilizer nitrogen. Ann. Bot. 453-461.

Shaw GG, Little CHA (1972) Effect of high urea fertilization
of balsam fir trees on spruce budworm development. In:
Insect and Mite Nutrition. (Rodriquez JG, ed.). North-
Holland Publishing Co., Amsterdam.

Sinden SL, Sanford LL (1986) Leptine glycoalkaloids and
resistance to the Colorado potato beetle (Coleoptera:

Chrysomelidae) in Sglanum gnagggn_g. Env. Ent. 15:
1057-1062.

Slansky F (1982) Toward a nutritional ecology of insects.
Proc. Fifth Int. Symp. Insect-Plant Relationship.
PUDOC, Wageningen.

Slansky F, Feeny P (1977) Stabilization of nitrogen ac-
cumulation by larvae of the cabbage butterfly on wild
and cultivated food plants. Ecol. Monogr. 47:209-228.

Solkal RR, Rohlf FJ (1981) Biometry: The principles and
practice of statistics in biological research. Free-
man. San Fransico.

94

Steel RGD, Torrie JH (1980) Principles and Procedures of
Statistics. A Biometrical Approach. McGraw-Hill.

Terman GL, Noggle JC, Hunt CM (1976) Nitrate-N and total N
concentration relationships in several plant species
Agromony J. 68:556-560.

Tingey W (1984) Glycoalkaloids as pest resistance factors.
Am. Potato J. 61:157-167.

van Emden RH, Way (1973) Host plants in the population
dynamics of insects. Insect/Plant Relationships. Sym-
posia of the Royal Entomological Soc. of London. No.
6. Blackwell. Oxford.

van der Meijden E, van Bemmelen M, Kooi R, Post BJ (1984)
Nutritional quality and chemical defence in the rag-
wort-cinnabar moth interaction. J. Animal Ecol. 53:
443-453

Waldbauer GP (1962) The growth and reproduction of maxillec-
tomized tobacco hornworms feeding on normally rejected
nonsolanaceous plants. Ent. exp. & app. 5:147-158.

Waldbauer GP (1964) The consumption, digestion and utiliza-
tion of solanaceous and non-solanaceous plants by
larvae of the tobacco hornworm, Ezgtgpargg sexta
(Johan.) (Lepidoptera:Sphingidae). Ent. exp. & appl.
7:251-269.

Waldbauer GP (1968) The consumption and utilization of food
by insects. In: Advances in Insect Physiology. 5:229-
288.

Waldbauer GP, Fraenkel GS (1961) Feeding on normally rejec-
ted plants by maxillectomized larvae of the tobacco

hornworm, Ergtgpazgg sexta (Lepidoptera, Sphingidae).
Ann. Ent. Soc. Amer. 54:477-485.

Waller Nowacki (1978) Alkaloid Biology and Metabolism in
Plants. Pleunum Press, NY and London.

White TCR (1979) An index to measure weather-induced stress
of trees associated with outbreaks of psyllids in
Australia. Ecology. 50:905-909.

White TCR (1984) The abundance of invertebrate herbivores in
relation to the availability of nitrogen in stressed
food plants. Oecologia. 65:90-105.

95

Wint GRW (1983) The effect of foliar nutrients upon the
growth and feeding of a lepidopteran larva. In:Nitro-
gen as an Ecological Factor. (Lee JA, McNeill S, Rori-
son HI, eds.). Sym. British Ecological Soc. Blackwell.
London.

Yamamoto RT, Fraenkel G (1960b) Assay of the principal
gustatory stimulant for the tobacco hornworm, Proto-
paggg sexta, from solanaceous plants.

Yamamoto RT, Jenkins RY, McClusky RK (1969) Factors deter-
mining the selection of plants for oviposition by the

tobacco hornworm Manduca sexta. Ent. exp. & appl.
12:504-508.

 

 

MANUSCRIPT III.

Conspecific host selection and subsequent
larval performance in specialist phytophages: the influence

of plant quality on herbivore behavior and physiology
(Lepidoptera:Sphingidae).

96

97

ABSTRACT

Mgndggg_§ppL, specialist herbivores, appear to dis-
criminate among individual tomato plants treated at dif-
ferent rates of nitrogen fertilization. Host preference,
measured as the percent of total eggs deposited and the
percent of plants with eggs, was higher on plants treated at
higher fertilization rates and containing higher concentra-
tions of soluble nitrates and total nitrogen.

Plants most preferred by moths also conferred the
highest fitness on developing larvae. Larvae reared on
foliage from preferred plants grew faster and developed into
larger, more fecund adults. Plant components affected by
nitrogen availabilty and thought to influence these be-

havioral responses are discussed.

INTRODUCTION

In many insect species, newly hatched larvae are
virtually immobile (Beck 1974, Renwick 1983). For these
individuals, survival and growth will be, to a large extent,
dependent upon the quality of the plant that the adult
female ”selects" (Singer 1986, Dethier 1947, Rausher 1979,
Mackay 1985). If the female oviposits on an inferior food
source, larval performance will be directly affected. Egg
deposition on plant species containing toxic substances or

lacking minerals and nutrients essential for growth can kill

98
the larva outright (Beck 1965, Rhaodes and Cates 1976,
Bernays 1982) or, more commonly, can inhibit the digestion
and assimilation of food consumed (Rhoades and Cates 1976,
Mitchell and Harrison 1985, Slansky and Feeny 1977, Reese
1978, 1983 Robinson 1979, Barbosa et. al. 1982). Although
some species are able to regulate consumption, or otherwise
compensate, in response to variations in plant quality
(Slansky and Feeny 1977, Slansky 1982), larval development
on a nutritionally inadequate food source will generally
produce smaller, less fecund, less fit adults (McNeill and
Southwood 1978, Scriber and Slansky 1981, Slansky 1982,
McNeill and Prestidge 1982, Prestidge and McNeill 1982a,
1982b).

Since host selection by an adult female can affect the
success of her offspring and the subsequent success of the
adults produced, it is reasonable to assume that host plant
quality will influence the evolution of host acceptance
behavior and ovipositional response. Through the process of
natural selection, insects should have evolved the ability
to distinguish between ”suitable” and ”unsuitable” host
plants. Presumably, those individuals that have this capa-
bility will have an evolutionary advantage over those in-
sects that lack this mechanism.

Historically, research has focused on elucidating the
factors that enable an insect to discriminate among
different plant species. These investigations have led to

the formulation of hypotheses regarding the ecology of host

99
acceptance behavior. However, there is no reason to assume
that discrimination occurs only at the interspecific level.
In fact, since host plant quality may influence the evolu-
tion of ovipositional response, it seems likely that host
searching females will exhibit intra-, as well as, inter-
specific discriminatory behavior.

Several recent studies have demonstrated that certain
insects have the physiological and behavioral capabilities
to distinquish among individual plants within a species
(Mitchell 1977, Ives 1978, Wolfson 1978, 1980, Myers and
Post 1981, Myers 1985, Rausher and Papaj 1983, Papaj and
Rausher 1987, McKay 1985). Results of these earlier studies
have shown that host searching females cue into plant size
(Ives 1978, Karban and Courtney 1987), density (McKay and
Singer 1982, McKay 1985), moisture (Shorey 1964, Sparks
1973, Wolfson 1980) and chemistry (Mitchell 1977, Wolfson
1978, 1980, Myers and Post 1981, Myers 1985) when "choosing"
among conspecific plants.

The present research was initiated to determine if
field populations of a specialist phytophage, in this case
Manduca §RR~ (Ml sexta and L W) . would die-
criminate among individual plants of the same host species.
In particular, the study focused on the chemical differences
among these plants (measured as percent total nitrogen and
soluble nitrates). A second objective of the study tested
the hypothesis that adult females prefer to oviposit

on those plants most suitable for larval growth and

100
survival.

These objectives were investigated by treating small
field plots of tomato plants with four different levels of
nitrogen fertilizer, thereby inducing physiological changes
in these plants. Host preference was determined by docu-
menting the distribution of eggs on treated plants from
field populations of Manduca Egg; Relative fitness was
examined by rearing laboratory cohorts of H1 guiggggmagglgga
larvae on bouquets of treated foliage and calculating vari-

ous parameters of fitness.

MATERIALS AND METHODS

W

This study was part of a sustainable agriculture
project at the Kellogg Biological Station in Hickory
Corners, Michigan. Investigations into the affects of
intercropped systems on the abundance and distribution of
insects continues to be an important aspect of this long
term project. In accordance with these goals, two separate
sites, a tomato monoculture, and a tomato/apple tree inter-
cropping, were included in this particular study. The
intercropped site consisted of a determinant cultivar of
tomato, var. Sunny, planted between tree rows in a dwarf
apple orchard (designated orchard site). The tomato mono-
culture consisted of the same cultivar of tomato planted in

an adjacent field (designated field site). In 1986, the

101
field site was moved from its original location to a similar
nearby area. The orchard site remained at the same location
throughout both years of the study (Figure 1).

The study sites were designed such that groups of
treated tomato plants occurred in equal proportions and grew
in close proximity to one another. Both sites were ini-
tially divided into multiple plots of tomato plants. Each
plot consisted of double rows of plants approximately 50 m
long. Plots in the field site were spaced approximately 5 m
apart while those in the orchard site were spaced 7 m apart
and interplanted with apple trees (Figure 1). Plots at both
sites were subdivided into three blocks. Each block was
subsequently divided into four small, rectangular areas
containing eight tomato plants. One of four levels of
nitrogen fertilizer was randomly applied to plants within a
rectangle until all four rectangles within a block had been
treated with a fertilizer application (Figure 2). A total
of 24 blocks were set up during the first year of the study
and 21 during the second year. An area approximately 1 m
between rectangles and 2 m between blocks was left un-
treated (Figure 2). These untreated areas, were designed to
prevent or decrease the movement of nitrogen through the

soil between adjacent treatments and blocks.

Wm

Fertilizer was applied directly to the soil surroun-

ding individual plants. In 1985, urea (45% nitrogen) was

dammed mo nous—«axe 858..“ how $5502 use €82 com .353 9689

880.52 now :88 .82. 03% .53“. 88982 Soc 033 Scam 8:88.88 “no: 050388 8m 593 :wmmoc 20E A oSwE

camp 25 0.3....

 

oooooooooooooooooooo
00000000000000000000
ooooooooooooooo

 

 

 

 

 

 

au0h<30h

Quatt<

 

 

Asian .0 «0:355 3

02w Dump—0.0 mmmr 03w EOE

8855.... H83 .55

103

ultra... a -1

.583 8 555388 855.. Be «.8602 “.5 .5552 8m .853 0589 8535 53.0.53 888
gun—m .855 0589 385 53.58 8.86 823 £83 035 .«o 55:98“ 5:888 58% 5E .m 95mm

 

 

8:85»:
5ng 58 95

 

 

 

.8803 025 95
33:55 83 035m

 

 

 

 

 

 

 

 

 

 

8:95am: coach... .o 2:09: 393..

104
applied at the rate of 0, 15, 60, and 120 lbs. nitrogen/
acre. Applications were split between June 29 and July 9 to
prevent burning the plants receiving high levels of fer-
tilizer. The first fertilizer application occurred ap-
proximately 3 and 1/2 weeks after tomato plants were trans-
planted into the field. Analyses of egg frequencies from
the first year of the study revealed that plants treated
with 0 and 15 lbs. nitrogen and those treated with 60 and
120 lbs. nitrogen elicited similar responses from the female
moths (see Results). Because of this similarity, fertiliza-
tion rates were changed to 0, 30, 120 and 240 lbs nitro-
gen/acre for experiments conducted during the second year of
the study. These particular levels were chosen for their
anticipated effect on ovipositional responses. The highest
application, 240 lbs. nitrogen, was hypothesized to be a
level at which physiological changes in the plant might
actually become deleterious or repellent to the moths.
Thirty pounds nitrogen was proposed to be the amount at
which differences might first be detected between low and
high nitrogen levels (=threshold level). To retain conti-
nuity between seasons, two rates of application (0 and 120
lbs. nitrogen applied/acre) were replicated between years.
Ammonium nitrate (34% nitrogen) provided the source of
nitrogen in 1986. As in 1985, total fertilization levels
were divided between two separate dates, July 3 and July 14,
with the first application occurring approximately 2 weeks

after transplantation of tomato plants into the field.

 

 

105
W
Since u; guingugmagglata is the more prevalent horn-

worm species during the early part of the season (see Manu-
script I), the period during which moth preference was ana-
lyzed, and eggs are virtually identical for both insects, no
attempts were made to distinguish between a; sexta and H;
W.

Tomato plants (half plants) were randomly sampled l
twice weekly during both field seasons. In addition, in
1986 plants were sampled every other day from July 13 to

July 21. Actual plants sampled were chosen by utilizing a “

 

random numbers table. On each sampling day in 1985, 60 half
plants, divided among all plots, were visually scanned for
eggs. In 1986, 96 half plants (24 per N treatment) were
sampled each day until August 21. However, beginning on
August 25, only 48 half plants (12 per N treatment) were
sampled per day (in 1986, plants in the orchard site became
heavily infected with geptgria lygopersici and in late
August, sampling was terminated at this site due to lack of

foliage.

 

Tomato plant stems and flowers and upper and lower
leaflets, were visually scanned for eggs. When an egg was
found, its location was recorded and the host plant was
tagged. To eliminate counting the same eggs more than once,
tagged plants were not sampled on succeeding days. Documen-
tation of plant height and diameter and general plant qual-

ity (vigor, color, etc.) were also recorded for each plant

106

sampled.

WW

Larval performance was assessed in 1986. For detailed
description of Materials and Methods of larval fitness
study, see Manuscript II. n; guingggmagulgtg was chosen as
the organism for the laboratory studies because a) it is the
most abundant species found in the field during July hence, I
the majority of eggs sampled during the host preference

study will be deposited by u; guingggmggglgtg moths, and
b) few studies examine the biology and ecology of this

fi- \“ . .

 

animal while a large amount of literature has accumulated
with respect to the biology and physiology of H; sexta.

Foliage fed to larvae was collected fresh from the
field grown tomato plants early each morning. Similar
aliquots of leafs were picked from each treatment. General-
ly, these aliqouts consisted of the first fully expanded
leaf from the terminal bud.

Two cohorts of u. gninguemagulata larvae were reared

from eggs received from the Oxford Tobacco Research Station

 

in North Carolina. Initially, larval populations were to be
generated from field captured moths. However, after at-
tempts to collect moths from natural populations proved
futile, it was decided that laboratory populations would be
used instead.

The first larval cohort was utilized to determine

percent larval survival, pupal dry weights, percent

107

successful adult emergence and female fecundity. Devel-
opment of this group of larvae coincided with larval devel-
opment of early field populations. The second larval cohort
was utilized to determine percent larval survival and days
required to reach newly molted fifth instar. Because of two
failed experiments (a result of technical errors), this
second larval cohort was begun on foliage three weeks later
than the first larval cohort.

To determine overall insect fitness, performance
potential (based on Laughlin 1965), was calculated from the
following equation utilizing values obtained from each

larval cohort experiment:

Performance Potential =

(% Larval Survival)(% Successful Female Emergence)
(MWNM)

Mean Larval Period

 

Although this index is not meant to reflect an exact measure
of fitness, it does provide a realistic estimate of an
individual's potential for conferring genes on to the next

generation.

Wis

Tomato foliage utilized for chemical analysis was
collected in 1986. Every three to four days, four similar
aliquots were randomly selected from bags of leaves col-
lected for larval feeding studies. These aliquots were

combined into one foliage sample per treatment per each

 

108
sampling day and then oven dryed at 60° C.

Foliage samples were analyzed for total nitrogen and
soluble nitrates. Prior to analysis, all foliage was ground
using a 40 mesh screen. To determine total nitrogen, 150 mg
of dried foliage was digested with Rel-catalyst KC-M3 (3.5 g
potassium sulphate + 175 mg mercuric oxide) and 9 ml sul-
furic acid for one hour. After cooling for 30 minutes, the
solution was brought up to a volume of 75 ml with distilled
water. Total ppm nitrogen was read using a Technicon Auto
Analyzer II.

To determine soluble nitrates, 20 mls of distilled
water was added to 500 mg of dried foliage an then shaken
for 30 minutes. Total soluble nitrates was read using a

nitrate electrode and an expanded scale pH meter.

RESULES

Foliage Analyses
During both years of the preference study, plants

treated with high levels of fertilizer produced foliage that
was visibly greener (although not chemically analyzed) than
leaves on plants treated with low levels of fertilizer or
left untreated.

Chemical analyses conducted on foliage collected in
1986 demonstrated that mean percent dry weight total plant
nitrogen was positively correlated with nitrogen fertili-

zation treatment (r = .93 at p = .05), i.e., plants

109

receiving higher levels of nitrogen fertilization contained
higher concentrations of total plant nitrogen. Figure 3
depicts the seasonal trends of this plant component. Even
though initial concentrations were relatively high for all
treatments, decline occurred quickly and concentrations
leveled off throughout the remainder of the study. Table 1
lists means and standard errors of percent total plant
nitrogen during each separate aspect of the separate study.

Mean percent soluble nitrates (Figure 4) was also
found to be positively correlated with fertilizer treatment
(r = .93 at p = .05). Each application of nitrogen resulted
in an intial increase in percent soluble nitrates that
steadily declined during the remainder of the season. For
plants treated at the high fertilization levels, this
decline spanned a period of four to five weeks. However,
for tomato plants treated with 30 lbs. nitrogen/acre, ni-
trate decline occurred within 12 days. By the second week
of August, nitrate levels for all treatments were very low.
Table 1 lists means and standard error of percent soluble

nitrates during each separate aspect of the study.

HQ§£_EI§£§I§DE§

Since host size can influence the number of eggs
oviposited on a particular plant (Ives 1978, Karban and
Courtney 1985), the amount of foliage available at each
nitrogen treatment was determined prior to examining moth

preference for a particular treatment. Mean plant volume

Tm

 

110

 
 

.28 be Bias 2 .3. as u as... 5% v5 z .8: 3 u 8.28 383 .z a... on
u 8.8%. sumo .z .8: o a 88m 385 .32 ”5:6 380:8 ems—om .5ng Emmet to Sop—om co 358 1.538 .n 0.5mm

m><o mmmomo owk<5§300<
oowm comm ooom com; ooofi oovfi OONH

 

 

 

 

_ s F . — b . . — . p L — p P s b L b s _ p p . bio
a
em %
. H
O
s a
. m.
e 0
. l.—
V
. ql.
.o N
I—
. H
. 0
9
a 2.—
N

 

 

 

lll

 

 

 

 

53 3mm ova
18$ «3.» c2
53 3.3 <2 <2 on .838
$3 3.2 o as?
885 a8. 58 as. $3 8... 9."
885 MB. :85 as. :35 an 9:
:85 «3. ass «.3. 885 new on g2
$85 80. $85 So. A83 85. o 0328
$85 8% 885 2.... and 8s 9a
as: «.2 “83 a? 5.: $8 a:
685 8.8 3.3 as $23 as an 5822
58 8.” 68¢ 88 A85 8s c :33 Boy

sue a.» 5 @u 5 Bessie.

= :28 E28 88.35 88$ 5852 .84

52.5

 

“8:838 .53 3835 ..o coup—.588
05 .588 mation—8935 sous—95055385338332 $93 05380938888 account—6
388.38.320.33»? 83.52.5339 45:83:32.8 gnaw—53882 A 3nd.

112

 

.8. 3: Bias 2 .2: SN u as” 5% :5 z .2: as u 3:23 88.9 .z .2:
on I 83% some .2 .8: o a as... @820 . 32 misc 38:8 ous—om 6.23.? 0328 seep—on co «Bob ficomeom .v Baum

 

853 $53 0m._.<5<500<
03m comm ooom 82 com: 83 com:
_.....L_..r.FL_»F. .tL.L.....rILL.LLL...._IO.O
:6
..NS

T ‘1 I ffi if

‘3?
0
con lNEOHSd

‘1
o

 

Ijfijfrj—rfir
LO
0
O

(D
O

 

 

.8203. 1 52

113

(plant volume =vxr2h = volume of a cylinder) was calculated
from observed plant measurements of height and diameter for
each treatment on each day that plants were sampled. The
figures obtained provided a close approximation of the
amount of plant material available as an ovipositional
substrate. Results of an ANOVA performed on mean plant
volume for each sampling date determined that there were no
significant differences between plant volume and nitrogen
application at p = .05 from the time that sampling first
began each year until July 23 (Julian date 204, DD 1620) in
1985 and July 19 (Julian date 200, DD 1552) in 1986 (Figure
5 represents these analyses-note similarities among areas
under the curve for all treatments during a particular year
from the time that sampling began at the start of the season
up to the point designated by the reference bars). These
results suggest that, during this period of time, any dif-
ferences in numbers of eggs found can be attributed to
factors other than plant size (note however, that plant
volume does not take into account leaf surface area, a
measurement which may be quite different for plants treated
with no, or low, amounts of fertilizer compared to those
treated with higher levels of fertilizer, although personal
observations suggested that this was not the case).

The frequency of hornworm eggs in the field is most
appropriately described as a negative binomial distribution
(Borth and Marrison 1984), i.e., a distribution heavily

skewed towards the larger classes. This study demonstrated

CUBIC METERS

CUBlC METERS

114

      
   
     
  

 
 

I tut-240 Ibo when
/ /
0.5 / I
I I
°-‘ , ’ rnt-tzo to. when
0 3 I I
. I I
I I
0.2 - Int-30 Ibo Ween
/
0.! // /
/ .
0.0 tut-o lbl N/ocro
no too too zoo 21° 2'20 230 240 260
JULIAN DATE
tut-120 Mn N/ocre
0.5
0.:
. . mtiso Ibo N/ocre
0.3
I
0.2 ' ’ Int-IS Ita- when
0.! [I
I
0.0

I
mr-o lb. N/oen
too too 200 210 220 230 2‘0 260

JULIAN DATE

WWWMQWASWEDFWTI—E
WWW-Ema: MR.

Figure 5. Plot of mean plant volume at each fertilization level. Y-axis represents mean plant volume.

X-axis represents Julian date. Z-axis represents nitrogen treatment applied. See Results for further
explanation of plots and calculation of plant volume.

 

 

115
a similar frequency distribution. Observed data fit to the
negative binomial were not significantly different from
expected values (Chi-square = 2.531; .25 < p < .5, and 3.68;
.1 < p < .24 in 1985 and 1986 respectively). The majority
of plants sampled received no eggs, a small proportion
received one egg and a very low percentage received more
than one egg (Table 2).

Moth preference was analyzed from July 2 to 23 (DD
1460-1620) in 1985 and July 7 to 19 (DD 1255-1552) in 1986.
In both years, this "window of time" coincided with maximum
egg deposition and treated blocks comprised of plants with
similar volume (Figure 6: Figure 5--horizontal reference
bars depict same day in each figure). Egg levels were low
during both years of the study, but especially in 1986
(Table 2; Figure 6). The low infestation levels became
especially apparent when data were separated by sampling
date, plot location and level of nitrogen applied. When
divided into these separate components, only 8% of the total
cells had egg frequencies higher than 5 in 1985, and only 6%
in 1986. Although negative binomial data are commonly ana-
lyzed by a contingency table, a Chi-square test for homo-
geneity between sites could not be performed since expected
cell frequencies were lower than 5 (Little and Hill 1978,
Steel and Torrie 1980, Solkal and Rohlf 1981).

To increase cell frequencies, statistical analyses
were performed on data that had been combined across all

sampling dates even though a more statistically appropriate

116

Table 2. Distribution of eggs on sampled plants. Data separated by treatment and year. The top
row of numbers, within each treatment, represents the frequency of occurrence and the bottom row
represents the percent of total.

 

NUMBER OF PLANTS SAMPLED BEARING
LBS. N APPLIED

 

 

 

 

 

YEAR PER ACRE 0 EGGS l EGG 2 EGGS 3 EGGS 4 EGGS
1985 O 72‘ 16 3 1 0
20” 4 <1 <1 0
15 74 13 2 0 0
21 4 <1 0 0
60 65 23 8 0 0
18 6 2 0 0
120 52 21 6 0 l
15 6 2 0 <1
Total Plants Sampled 263 73 19 1 1
Percent Sampled 74% 20% 5% <1% <1%
1986 0 125 16 3 0 0
22 3 <1 0 0
30 126 16 1 . l 0
22 3 <1 <1 0
120 118 21 2 2 1
20 4 <1 <1 <1
240 118 23 2 1 0
21 4 <1 <1 0
Total Plants Sampled 487 76 8 4 1
Percent Sampled 84% 13% 1% <1% <1%
‘Number of plants sampled

”Percent of total plants sampled

117

-101 ' 1985

1
0.8‘ .
0.6‘
0.4‘

J

0.2‘

HALF PLANT EGG MEANS

 

 

003 . , . , . , . . . r , , . ,
1140 1340 1540 1740 1940 2140 2340 2540

 

10: 1986
00‘
031
04‘

0.2‘

l

 

 

HALF PLANT EGG MEANS

 

0-0.” ' l ' 1 ' 1 ' l ' l ' l ' I
1140 1340 1540 1740 1940 2140 2340 2540

ACCUMULATED DEGREE DAYS

figure 6. Seasonal hornworm egg densities. Mean eggs calculated per half plant. Reference bars denote period
of maximum egg deposition. Data combined from field and orchard site. N = 60 and 96 half plants sampled per
day in 1985 and 1986 respectively.

118
analysis would have been to examine moth preference on a
daily basis. Once data had been pooled across dates, a
Chi-square test for homogeneity (Solkal and Rohlf 1981)
determined that data could be combined across site location,
i.e. ovipositional responses were not significantly in-
fluenced by site location (Chi-square = 1.50; p > .65, and
3.03; p > .40 in 1985 and 1986 respectively).

Based on the observation that only a small proportion
of plants sampled had more than one egg, cells of the table
were divided into level of nitrogen applied and plant
status, i.e., plants with or without eggs. Data from both
years of the study were analyzed separately.

In 1985, a G-test for independence (Solkal and Rohlf
1981) determined that the percent of plants with eggs was
associated with the amount of nitrogen applied (Figure 7).
However, in 1986 the results of this test did not show a
significant association between the number of plants with
eggs and fertilization rate (Figure 8) although, results
demonstrated that higher nitrogen levels had slightly more
plants with eggs.

Since data was collapsed into only two classes (plants
with or without eggs) prior to analysis, a test for indepen-
dence does not take into consideration those plants bearing
multiple eggs. This is unfortunate since the majority of
plants having more than one egg were also those receiving
the higher levels of nitrogen fertilizer, especially during

the first year of the study (Table 2). However, because a

119

 

40
1985

 

 

30‘

 

 

PERCENT
N
1 ‘3

p—s
O

L

 

 

 

 

 

 

 

 

 

 

 

O 15 60 120
LBS N/ACRE

Figure 7. Percent tomato plants with eggs at each nitrogen level.
Data combined from field and orchard site--l985 (heterogeneity
Chi-square = 1.50, p > .65). A G-test analysis for independence
determined that the proportion of plants with eggs is associated
with the level of nitrogen applied (G = 10.168, p < .025). See
text for explanation of statistics.

120

 

1A
O

1 986

CO
CD

 

 

PERCENT
2‘3

 

p—s
O

L

 

 

 

 

 

 

 

 

 

 

 

O 30 120 240
LBS N/ACRE

Figure 8. Percent tomato plants with eggs at each nitrogen level.
Data combined from field and orchard site--1986 (heterogeneity
Chi-square = 3.03, p > .40). A G-test analysis for independence
determined that the proportion of plants with eggs is independent
of the level of nitrogen applied (G = 5.02, p < .20). See text for
explanation of statistics.

121
large number of total plants was sampled during each year's
"window of time” (367 in 1985 and 576 in 1986, see Table 2),
it was possible to calculate the percent of total eggs found
during both seasons of the study and obtain a very stable
approximation of moth response to each nitrogen level (Table
3). To correct for unequal sample size in 1985, the percent
of total eggs was calculated by relating the actual number
of eggs to a relative sample size of 92 (Table 3). Results
of this procedure demonstrated that, in both seasons, ap-
proximately 40% of the total eggs were found on plants
treated with low levels of nitrogen fertilizer while 60%
were found on plants treated with higher levels. Lending
support to this analysis is the observation that similar re-
sponses were seen during both years of the study for the
replicated treatments (0 and 120 lbs. nitrogen applied/
acre). A plot of these percentages and the corresponding
quadratic regression line is shown in Figure 9.

In 1986, a correlation analysis utilizing values
obtained from foliage analyses and preference data collected
from July 7 to July 19, determined that the percent of
plants with eggs and percent total eggs were not correlated
with mean percent dry weight nitrogen or mean percent sol-
uble nitrates. However, moths did demonstrate higher re-
sponses towards plants containing higher percentages of

these chemical components (Figures 10 and 11).

 

122

swam .852 2.83 see 3858...

 

 

when: 88 moom 1.8.:
mm 8 v3 ova
mm mm 3.: cm:
2 <2 <2 G 3: on
on mm <3 o 33
32 pal: £2 88 45.8
.2 Due mo R 8 cm:
Ln vsm me am we 8
.3 8.: no D on n :
.8 93 Na mm 8 o 32
.2.—.2. so meow ogm meow saga mmoa. Mme m<m>

szumma MNBZDZ madam MmmEDZ mg Omar—ma:
Ea mg; :25an #5030 Z .mmd

52.882 88 2a 32 s Essa 85.: 59 8R 9a 5 u 2 282.53 8 82893 as .5.
8m .88» 53 wet—6 958 «mm» 39 06 ..o .582. a 9 3:95... 83 ~96. coups? :08 on 958 ammo

..o been... 8F .33» 8:98.05 05 ..o .aoa :08 met—6 3368 ammo 583.com ..o .8255 38. .m 058.“.

123

 

 

doumusm Mo 35535 .6.“ axe 8m .83 33 u was“ .53 32 u «£85 .8. u a
.m n. u «m 39 .563th 35828 a 9 E Sun .358 ammo 39 9 88588 5ng some .3 32.358 888m .a onE

mm0<\ow3n_n_< 2 mm;
ovm oom om: oma cm 0% o

3993 .LNEOHHd

 

 

 

20‘

14_¥A+

p—s
a:

- . A A

ZPLMT8WITHEGGS
p.-
Q

12‘

 

10‘

r.
p-

(b)

.78
.21

124

of?

 

7. TOTAL EGGS
N N on ea
o 0‘ G en

y—u
U‘
.1 L A A

A A. A

 

I T Y Y

 

p
O
L A

Figure 10. Ovipositional response to treated plants. Correlation of percent nitrogen dry weight
with total number of eggs deposited (a) and proportion of plants with eggs (b) at each treatment.
Data collected during 1986 only. Total number of eggs deposited = 108. N = 576 total half
plants sampled. Closed stars = 0 lbs. N, open circles = 30 lbs. N, closed circles = 120 lbs. N

(a)

V ‘1' U

4

WANZMTROGEQ

and open stars = 240 lbs. N applied per acre.

125

 

 

 

 

207
. r= .80
‘ p= .19
m i
16:
m i
14‘.
j t
N J
12‘ O
1 (b)
104! T l 1 r r j I l w l
0.0 0.1 0.2 0.3 0.4 0.5
351
. r= .77
«p' .22 Q
30?
1 $3!
$25:
:2 2
'5 L *
"20J O
x 1
l
15‘.
3 (a)
to‘rfi..,..-,.......,....
0.0 0.1 0.2 0.3 0.4 0.5

MEAN 7. MTRATES

Figure l 1. Ovipositional response to treated plants. Correlation of percent soluble nitrates with
total number of eggs deposited (a) and proportion of plants with eggs (b) at each treatment. Data
collected during 1986 only. Total number of eggs deposited = 108. N = 576 total half plants
sampled. Closed stars = 0 lbs. N, open circles = 30 lbs. N, closed circles = 120 lbs. N and open
stars = 240 lbs. N applied per acre.

126
Lamllitneaa

Larval performance was significantly enhanced by
nitrogen fertilization. Results of both Cohort I and Cohort
II growth studies determined that larvae grew faster and
developed into more fit adults when reared on foliage
treated at higher rates of fertilization. Table 4 lists the
means (where applicable) and standard errors of the means
for all measured parameters of fitness for each foliage
treatment.

Regression analyses of Cohort I determined that male
and female pupal dry weights, successful adult female emer-
gence and female fecundity were significantly influenced by
rate of nitrogen fertilization (Table 5). Similar regres-
sion analyses of Cohort II demonstrated significant associa-
tions between level of fertilization and number of days
required for larvae to reach newly molted fifth instar, as
well as the percent survival of these larvae (Table 5).

Only survival of Cohort I larvae and successful adult male
emergence were not linearly correlated with nitrogen treat-
ment, although emergence of males reared on foliage treated
with 30 lbs. N per acre was affected by dietary nitrogen--
Table 4). Most fitness parameters significantly influenced
by nitrogen fertilization were also correlated with mean
percent dry weight plant nitrogen and mean percent soluble
nitrates (Table 6--mean percent plant nitrogen and soluble
nitrates calculated from foliage collected while each sepa-

rate larval cohort was developing--see Table 1). See

 

127

 

 

 

 

 

 

 

$3 $2: $3 33
o9 8 mod ow; : 3 on NmA w: ova
$3 $3.: $3 8.
8 No v0.9 9.3— mm ow SA a: 02
at ESQ 98 a3
a.» S 8.: ooéw 2. mm :4 mo; On
8»; $3: $3 33
t. on 3.: 09mm 2. ms 5: 84 o
8T5 8W5 Signs 875 8W5 8T5 ans ans.
= c200 35:00 cams .EE bassoon 255m 032 295m 032 3:9? 2 .mmq
fiztsm :33 mo 9 $30 oEEom oocowbEm Emma?
Emmoousm 2*. be 395

 

.vum~nna nonmamuuuu cowouumn
uo Ho>o~ June on noonumu ~u>u0~ we uuouoluucn newsman! Han new Aauenuo uvunvcnuu «v «can! .e manna

128

”mung 38.825 838 cc 33:28 803095
885030 3.385 8 35:28 5858.

 

88. vn a. .5889 :8
88. v n no. a 89.00
.3 ..I H :38
385,5 :23
88. X. 2. Sagan cg
8. X. a. flex
.3 .... on:
.auooaumbfim :52
88. vn 3. 33¢
88. va 3: on:
.9833 to if
a «M

 

A8. u m a .58an .8 a .38 88am

.933 a ..o .083qu v8 83> mm 838 336:2 $38 865m 155 ..o ”835” 5186.32 883 . m 2.3.

 

129

 

as... :3 3583:
28538 .8885 58 82283..

 

 

8.5 8.5 8.5 :83 93 38.5 28.5 8.5 8:52
8. 2.- mm. 8..- 3. 8. 8. 3. 0328 a.
$3 8.5 A85 38.5 98.5 888.5 83 8.5 5852
8. 8.- 8. 8.- S. R 3. S. .883
E 8
= :28 32.8 :aaSBm .5568 22E ca: 223 $2
835m 85 R. 2880““ 8 £3 L583 3coEoEm
be :85 E§§m a.

88. u m a

.58958 6: u .25 859.358 .3 .26— 838083 83553.. 5 ”89:32 dong—9:8 mo 838508 Magnum :55 838:: ..o 32
n8. 355.888 885 no covena— 82888 55 8888 0328 98 5mg! .53 .583 ha .522. :88 Ho Sun—Boo .o 038.“.

130

Manuscript II for detailed analysis of the affect of nitro-
gen fertilization on u; guinguemggulgta larval perfor-

mance .

W

Host preference for conspecific plants was signifi-
cantly correlated with larval performance (Table 7). Pupal
dry weights, female fecundity, adult female emergence and
percent larval survival of Cohort II were positively as-
sociated with moth response, measured as percent of plants

with eggs and percent of total eggs deposited at each level

 

of fertilization. Larval periods through pre-fifth instar
were negatively associated with moth response. Only percent
larval survival of Cohort I was not correlated with moth
preference.

Performance Potential (overall larval performance),
based on survival values from Cohort I, increased as the
percentage of plants with eggs increased, although this
relationship was not significantly correlated (Figure 12a).
However, Performance Potential based on survival values from

Cohort II, did demonstrate a significant correlation with

 

moth response (Figure 12b).

131

 

 

 

 

 

 

as 8.5 88 :8 88.5 :88 88 85 88m as»,
8. 8. 8.. q... x. 8. a. 8. as: e
63 8.5 :3 88.5 88.5 28.5 8.5 8.5 88m 88 e
a. s. 8. s. 9.. in a. p.
a 8980 :5on ”BE bassoon 088$ 082 ”Beam 282
Erram RES 8 9 £5 oaeom Emu? oocomoosm
to 395 88883 e

 

.mo. I m u. unnumu
Imam. you I .u.—& ouauwumuumo no agoa uuwnuuaoluv unsung-.0»!— dw swan-5: .uugumumuuoou oomunmouuou
n.60uu-vm ’0... anon-5c no u’ou non. .ouvnumu Auk—n.— numi ounouuuoun amnion no acmuaaouuoo . h van—nu.

132

 

 

 

 

100‘ 1‘3:
1' - .96
90‘ p = .04
80‘ O
70‘
l
60‘
50: *
‘ O
40‘ based on Cohort II survival
J (b)
2 30 l V l v I v | ' I ' T
10 12 14 16 18 20
1004
‘ r - .38
u. 90‘ p " -10
i .04
1
70‘ it
60‘
J
50‘ '
40‘ *
‘ 0 based on Cohort I survival
30*, (a) . , , , . , . , . r
10 12 14 16 18 20

ZPLANTSWITHEGGS

Figure 12. Correlation between percent plants with eggs and larval performance potential. Percent
plants with eggs based on 144 plants sampled per treatment. See text for definition of larval
performance potential. Closed stars = 0 lbs. N , Open circles = 30 lbs. N, closed circles = 120 lbs.
N and open stars = 240 lbs. N applied per acre.

.44

 

133

DISCUSSION

HQ§£_£I§£§IQDQ§

Results of this research, suggest that Manduca moths
show a greater preference for plants that have been treated
with higher levels of nitrogen fertilizer. The discrimi-
natory behavior exhibited by the moths was especially ap-
parent in 1985. The proportion of plants bearing at least
one egg showed a strong positive association with nitrogen
treatment. In addition, sixty-six percent of the total eggs

deposited were on plants treated at the two higher fer-

 

tilization levels. Unfortunately, in 1985 foliage was not
collected for chemical analysis. Hence, while we assume
that increased rates of fertilization increased concentra-
tions of total nitrogen and soluble nitrates, no conclusions
can be drawn concerning the level of moth response to these
plant components during this particular year.

A similar response to fertilization treatment was
demonstrated in 1986, but of a lower magnitude. Although
not significantly correlated, the proportion of plants with

eggs and the percent of total eggs were higher on plants

 

treated at higher fertilization levels and on plants con-
taining higher concentrations of total nitrogen and soluble
nitrates. Since the percent of total eggs deposited was
very similar in both years of the study, i.e., approximately
40% on plants receiving no or low levels of fertilizer

versus 60% on plants receiving higher levels (Table 3), it

134
is likely that these responses would have been significant
had the total number of eggs deposited, with respect to
plants sampled, been as high in 1986 as it was in 1985
(Table 3).

Although fertilization levels were applied in discrete
increments, moth response to nitrogen treatment did not
appear to increase incrementally. In both years, the per-
cent of total eggs oviposited on plants treated with "low"
levels of nitrogen fertilizer was very similar. Likewise,
the percent of total eggs oviposited on plants treated with
”high” levels of nitrogen fertilizer was very similar
(Figure 9: Table 3). Although, this response, at least in
1986, is partially due to the similar concentrations of
total plant nitrogen in foliage treated with 120 and 240
lbs. N/acre, these observations also suggest that treated
plants only elicited two general levels of response from the
moths, i.e., at some ”threshold level" physiological changes
in the plant, induced by nitrogen fertilization, elicit a
higher ovipositional response in the moths; however, below
this level, no increased response is observed.

Two results of the study lend additional support to
this hypothesis. First, plants treated with 240 lbs. nitro-
gen/acre and containing the highest concentrations of total
nitrogen and soluble nitrates elicited a slightly lower
response in moths than plants treated with 120 lbs. nitro-
gen/acre (Table 3; Figure 9). Although it is impossible to

determine if this response is real (since only one data

135

point is available for examination), it seems reasonable to
assume that moth response to plant quality will reach some
peak and then level off or even decrease thereafter but will
not continue to increase incrementally. Secondly, female
moths exhibited the lowest response towards plants treated
at the low rates of nitrogen fertilization (15 or 30 lbs.
nitrogen per acre, depending on the year) not towards un-

treated plants. This response occurred even though low

v-.rv

applications of fertilizer increased levels of both total
nitrogen and soluble nitrates in these plants (Figure 3;

Figure 4). ,

 

Moth preferences demonstrated within this study were
generally as predicted although, plants treated with 30 lbs.
nitrogen/acre elicited some unexpected responses (discussed
in greater detail at a later point). Insect response to
nitrogenous compounds has been demonstrated by Howitt and
Connor (1965) and Myers and Post (1981) (see also review by
Thorsteinson 1960). Unfortunately, since nitrogen interacts
with a number of plant components, it is difficult to ascer-
tain which particular plant cues nanggga gpp. utilize to

discriminate among individuals within their host species.

 

It is conceivable that nitrogen fertilization affects other
aspects of plant physiology influencing ovipositional re-
sponses besides the concentrations of total nitrogen and
soluble nitrates.

td-Tomatine, a secondary compound containing nitrogen,

is a steroidal glycoalkaloid common to Lyggpgrsiggg

136
(tomato). This particular alkaloid is composed of an agly-
cone and a glycoside (a sugar) and has a high carbon to
nitrogen ratio; hence, the availability of either of these
minerals will impact synthesis of the alkaloid and subse-
quently, concentrations found in the plant.

Juvik et. al. (1982) determined that mean K-tomatine
concentrations in Lyggpgrsiggn eggplantum, range from 3.3
mg/g to 16.0 mg/g depending on the particular cultivar
analyzed. Although our initial reaction would be to assume
that as plant nitrogen content increases, .a-tomatine con-
centrations will also increase, it is probable that in-
creased nitrogen reserves result in decreased glycoalkaloid
production. With increased levels of plant nitrogen, carbo-
hydrates and other carbon containing componds are utilized
to synthesize nitrogen acceptors, usually aspartic and
glutamic acid (Selvendran and Selvendran 1973) thus depele-
ting the amount of carbon available for glycoalkaloid pro-
duction. Results of experiments conducted on solanum glyco-
alkaloids demonstrated that an increase in plant nitrogen
did, in fact, result in a depletion of the carbohydrate pool
and a subsequent decrease in glycoalkaloid synthesis
(Nowacki et. al. 1975, 1976).

since glycosidic compounds are thought to stimulate
feeding in Manduca larvae (Yamamoto 1960b), concentrations
of a-tomatine may also influence host acceptance behavior
and egg deposition in the female moths. However, based on

the preceeding discussion, it seems reasonable to assume

137
that glycoalkaloid concentrations either did not induce dis-
criminatory behavior in nanggga at the intraspecific level
or exerted only a small influence, i.e., the slight decrease
in response to plants treated with 240 lbs. nitrogen/acre
may have been the result of decreased alkaloid production at
this high level of nitrogen.

Leaf water content is another plant component that
elicits ovipositional responses in phytophagous insects
(Shorey 1964, Sparks 1973), and that is also influenced by
nitrogen availablilty. The covariance between this com-
ponent and plant nitrogen is well documented (Slansky and
Feeny 1977, Radin and Parker 1979, Scriber and Slansky 1981,
Mattson 1980, Manuwoto and Scriber 1985, Mattson and Haack
1987). Increased levels of plant nitrogen induce lush,
succulent foliar growth resulting in plants with a high
concentration of leaf water. In contrast, nitrogen defi-
cient plants are chlorotic, stunted, non-succulent and have
tough cell walls. Because these two plant components co-
vary, it is difficult to determine which variable actually
induces discriminatory behavior in insects. Wolfson (1980)
implicated plant water content as the factor controlling
ovipositional preferences in Pieris ragga. However, in all
cases where preferences were significant, nitrogen, as well
as water concentrations, were higher for the preferred
plants, and in all cases where preferences were not ex-
hibited, nitrogen, as well as water levels, were lower.

Sparks (1973) has already determined that leaf

138
moisture elicits ovipositional responses in H1 sexta.
Hence, we can hypothesize that higher responses towards
plants treated with higher levels of fertilizer may have
partially been a result of increased water concentrations
and the slight decline in response towards plants treated
with 240 lbs. nitrogen/acre a result of the dilution of
certain stimulatory compounds caused by increased levels of
water. Unfortunately, water content was not analyzed during
the host preference study. However, foliage analyzed ap-
proximately five weeks later, did not support this hypothe-
sis. Percent water content determined during the larval
performance study was negatively correlated with rate of
fertilization.

Determining the plant components inducing these be-
havioral responses is further complicated by the interaction
of water and nitrogen on chlorophyll production and subse—
quent photosynthesis. This interaction may affect host
selection in two ways. First, a deficiency of either nitro-
gen or hydrogen atoms will influence production of chloro-
phyll, the pigment associated with plant color and hence,
plant "greenness". Since recent studies have implicated
plant color as a cue utilized by females during host search
and acceptance (Miller and Harris 1985, Myers 1985), it is
possible that nanducg moths demonstrated higher responses
towards plants treated at higher levels of fertilization
because of the slightly darker color of these plants.

Secondly, photosynthetic activity influences the

139
production and availability of plant sugars. Generally,
decreased photosynthesis, resulting from decreased avail-
ability of nitrogen and water, decreases the concentrations
of plant sugars (Devlin and Witham 1983). The fact that
these compounds are attractive to insects is well
documented. With regards to this study, it is reasonable to
hypothesize that plants treated at low levels of fertiliza-
tion or left untreated had inadequate sugar supplies; hence,
elicited a lower response from host searching females than

plants treated at higher levels of fertilization.

figs; Ezgfexpnce VS, Lazyal Egrfgzmgngg

A number of researchers have examined the affect of
female host choice on subsequent larval fitness. In some
cases, results indicated that certain insect species prefer
those plants that produce the most viable offspring. An
excellent study by Via (1986) demonstrated a positive gene-
tic covariance between the oviposition preference of Ligig:
myzia satiyae and the relative pupal weight of each female's
offspring. Studies demonstrating population level associa-
tions between host preference and larval performance include
those by Chew (1975), Myers and Post (1981), Holdren and
Ehrlich (1982), Rausher and Papaj (1983), Williams (1983),
Leather (1985), Grossmueller and Lederhouse (1985) and Papaj
and Rausher (1987).

Equally well documented though, are results of studies

that did not find a correlation between moth response and

140

larval fitness. Courtney (1981) reported that Anthggnaris
gaxgamings preferred to oviposit on plants that caused the
highest larval mortality. Other studies demonstrating these
seemingly maladaptive host acceptance behaviors (although,
see Singer 1984) include those by Rausher (1979), Renwick
and Radke (1982), Mackay (1982, 1985 and Karban and Courtney
(1987).

This study indicates that Manduca not only exhibits E
conspecific discriminatory behavior, but that "preferred"

plants are also the most suitable for larval growth and

 

development. Ovipositional responses were significantly 7
correlated with most individual measurements of insect fit- i
ness (Table 7). In addition, performance potential, an
overall measure of insect fitness, was higher on the most
preferred plants (Figure 12a,b).

At the beginning of the study, we assumed that un-
treated plants would be the least preferred and most nutri-
tionally inadequate. However, plants treated with 30 lbs.
nitrogen/acre) elicited the lowest response from female E
moths and provided the least amount of nutritional support

for developing larvae. Larvae reared on foliage from these

 

plants required more days to reach fifth instar, had the
lowest pupal weights and also the lowest overall fitness.
In fact, 50% of the male larvae died during the pupal stage
compared to a 20-25% mortality rate for larvae reared on the
other three treatments. These results suggest that small

increases in nitrogen availability may adversely affect

141
herbivores, probably by creating an imbalance among those
plant factors eliciting positive responses and/or providing
adequate nutrition and those factors inhibiting oviposition,
feeding and/or development. Since the actual behavioral or
physiological mechanisms involved in these responses are not
known, it would be worthwhile to investigate the chemical
differences between untreated foliage and foliage treated
with 30 lbs. nitrogen/acre and subsequently, the affect of T”
these differences on insect performance and host selection.

Overall larval performance appeared to be affected by

 

nitrogen fertilization in a manner similar to that demon- ;
strated by ovipositing moths, i.e., tomato plants left i
untreated or treated at lower fertilization rates induce one

level of response while plants treated at higher fertiliza-

tion rates elicit a greater level of response. Although

certain measured parameters of fitness in this study were

higher or lower for particular treatments, larval perfor—

mance appears to show a similar "two level" response.

Larvae reared on foliage treated with high levels of fer- F‘
tilizer show a high survival rate and develop into heavier

pupae and more fecund females. Conversely, larvae reared on

 

untreated foliage or foliage treated with 30 lbs. nitrogen/
acre show a lower survival rate and develop into smaller
pupae and less fecund adults.

Since both host seeking female moths and developing
larvae exhibit similar responses to plant nitrogen content,

it appears that similar plant components influence both the

142
behavior and the fitness of these phytophages, thus implying
a genetic basis for the correlations demonstrated in this
study. Females that "select” and then deposit eggs on
plants supporting the highest larval fitness will increase
their chances of making a genetic contribution to the next
generation by increasing the probability that their off-
spring will complete devlopment and emerge as viable, fecund
adults. In addition, the genetic ”program" that enabled
these females to distinguish between ”suitable" and "un-
suitable" hosts will be conferred onto their offspring, thus

producing a new generation of individuals with similar

 

behavioral mechanisms for host discrimination.

However, a second explanation for the responses docu-
mented during this study is equally viable. It is possible
that adult females become "conditioned", thru prior ex-
perience, to select plants comprised of physical and chemi-
cal components similar to those contained in their larval
host plants. This phenomenon, commonly reffered to as L

Hopkins Host Selection Principle (Hopkins 1917) has recently

 

been demonstrated by Jaenike (1983). In fact, the affect of
conditioning on Manduca has already been demonstrated. I
Yamamoto and Fraenkel (1960a) determined that M; sexta moths

preferred to oviposit on the host plant species fed on

during the larval stage. With regards to this study, an

abundance of nitrogen rich host plants, either other tomato

plant varieties or other host species, occurring near the

study sites may have preadapted Manduca moths to oviposit on

143
those plants containing higher concentrations of nitrogen.
Unfortunately, the responses demonstrated in this

study do not provide the information needed to eliminate one
or the other of these two hypotheses. While the results are
interesting, further experimentation will be required to
test these theories. However, regardless of the complexity
or simplicity of this system and of the mechanisms involved,
it is inherently interesting that Manggga females prefe-
rentially oviposit on plants producing individuals with the
highest overall fitness. Both the ecological and economical
implications of these findings are important. Hopefully,
future studies will investigate, in detail, the factors in
both the plant and the insect, that elicit these responses
and the physiological and behavioral basis of these interac-

tions.

0 c o

Manduca spp., specialist herbivores, appear to dis-
criminate among conspecific plants. Plants treated at
higher rates of nitrogen fertilization and containing higher
concentrations of total nitrogen and soluble nitrates eli-
cited the highest responses from female moths. Unfortu-
nately, because of the complexity of interactions that occur
among plant minerals, water, secondary compounds and meta-
bolism, it is difficult to delineate the particular factors
that induce discriminatory behavior in these moths.

Furthermore, plants most preferred by ovipositing

144
females also produced individuals with the greatest overall
fitness. Larvae reared on foliage from "preferred" plants

grew faster and developed into larger, more fecund adults.

 

 

REFERENCES

Barbosa P, Saunders JA, Waldvogel M (1982) Plant mediated
variation in herbivore suitability and parasitoid
fitness. Proc. Fifth Int. Symp. Insect-Plant Relation-
ships. PUDOC, Wageningen.

Beck SD (1965) Resistance of plants to insects. Ann. Rev.
Entomol. 10:207-232.

Beck SD (1974) Theoretical aspects of host plant specifi-
city in insects. In: Proc. Summer Institute on
Biological Control of Plant Insects and Diseases.
(Maxwell FG, Harris FA, eds.). Univ. Press Miss.,
Jackson, Miss. E

Bernays EA (1982) The insect on the plant-a closer look.
Proc. Fifth Int. Symp. Insect-Plant Relationships,
PUDOC, Wageningen.

 

Borth PW, Harrison FR (1984) Within-field distribution 7
patterns of tobacco and tomato hornworm (Lepidop-
tera: Sphingidae) egg and larval populations in
southern Maryland tobacco. 13:859-862.

Chew FS, Robbins RK (1984) Egg-laying in butterflies.
In: The Biology of Butterflies (Vane-Wright RI,
Ackery PR, eds.). Academic Press.

Chew FS (1975) Coevolution of Pierid butterflies and
their cruciferous foodplants. I. The relative quality
of available resources. Oecologia. 20:117-127.

Courtney SP (1981) Coevolution of Pierid butterflies and
their cruciferous foodplants. III. Anthgghgzig
cardamings (L.) survival, development and oviposition
on different host plants. Oecologia. 51:91-96.

Dethier VG (1947) Chemical Insect Attractants and Repel-
lents. Blakiston. Philadelphia.

 

Grossmueller DW, Lederhouse RC (1985) Oviposition site
selection: an aid to rapid growth and development in
the tiger swallowtail butterfly, Papilig glaugu_.
Oecologia. 66:68-73.

Gulmon SL, Shu CC (1981) The effects of light and nitrogen
on photosynthesis, leaf characteristics, and dry
mater allocation in the Chaparral shrub, Qiplacus

guzagtiagngL 0cologia. 49:207-212.

145

146

Gunn A, Gatehouse AG (1985) Effects of the availability of
food and water on reproduction in the African army-

worm, Spgggntera gxgmpta. Physiological Entomology.
10:53-63.

Holdren CE, Ehrlich PR (1982) Ecological determinants of
food plant choice in the Checkerspot butterfly

Enpnydryas editna in Colorado. Oecologia. 52:417-
423.

Hough JA, Pimentel D (1978) Influence of host foliage on
development, survival and fecundity of the Gypsy
Moth. Env. Ent. 7:97-102.

House HL (1965) Effects of low levels of the nutrient con-
tent of a food and of nutrient imbalance on the feed-
ing and the nutrition of a phytophagous larva, Qelerio
ggphgpiag (Linnaeus) (Lepidoptera: Sphingidae). Can.
Ent. 97:62-68.

Howwitt AJ, Connor LJ (1965) The response of Bnaggletis
pgmgnglla (Walsh) adults and other insects to trap
boards baited with protein hydrolysate bait. Proc.
Entomol. Soc. Ontario. 95:134-136.

Ives PM (1978) How discriminating are cabbage butterflies?
Australian J. of Ecology. 3:261-276.

Jaenike J (1983) Induction of host preference in ngsgpnila
mglangga§t_r. Oecologia. 53:320-325.

Juvik JA, Stevens MA, Rick CM (1982) Survey of the genus
Lyggpgrsiggn for variability in (K-tomatine content.
Hort. Sci. 17:764-766.

Karban R, Courtney S (1987) Intraspecific host plant

choice:lack of consequences for Streptantggs tgrtggsus

(Cruciferae) and Eggnlgg hyantis (Lepidoptera:Pieri-
dae). Oikos. 48:243-248.

Laughlin R (1965) Capacity for increase: a useful popula-
tion statistic. J. Animal Ecology. 34:77-91.

Leather SR (1985) Oviposition preferences in relation to
larval growth rates and survival in the pine beauty

moth, Banglis flamm a. Ecol. Ent. 10:218-217.

Lindroth RL, Batzli G0, Avildsen SI (1986) Lgspgdgza
phenolics and Bengtgmgn alkaloids: Effects on diges-
tion efficiencies and growth of voles. Journal of
Chemical Ecology. 12:713-728.

147

Mackay DA (1985) Conspecific host discrimination by oviposi-
ting Egphygryas gditha butterflies:its nature and its
consequences for offspring survivorship. Res. Popul.
Ecol. 27:87-98.

Mackay DA, Singer MC (1982) The basis of an apparent prefe-
rence for isolated host plants by ovipositing Eupty;
gala libye butterflies. Ecol. Ent. 7:299-303.

Manuwoto S, Scriber JM (1985) Consumption and utilization
of experimentally altered corn by southern armyworm:
iron, nitrogen and cyclic hydroxamates. J. of Chem.
ECO. 11:1469-1483.

Mattson WJ (1980) Herbivory in relation to plant nitrogen
content. Ann. Rev. Ecol. Syst. 11:119-161.

Mattson WJ, Haack RA (1987) The role of drought in out-
breaks of plant-eating insects. BioScience. 37:110-
118.

McNeill S, Prestidge RA (1982) Plant nutritional strategies
and insect herbivore community dynamics. Proc. Fifth
Int. Symp. Insect-Plant Relationships. PUDOC, Wagen-
ingen.

McNeill S, Southwood TRE (1978) The role of nitrogen in the
development of insect/plant relationships. In:Biochem-
ical aspects of plant and animal coevolution. (Har-
borne JB, ed.). Academic Press. London.

Miller JR, Harris MO (1985) Viewing behavior-modifying
chemicals in the context of behavior: lessons form
the onion fly. In:Chemical Ecology of Insects. (Bell

WJ, Carde RT, eds.). Sinauer Associates, Sunderland,
MA.

Mitchell BK, Harrison GD (1985) Effects of figlangm glyco-
alkaloids on chemosensilla inthe colorado potato
beetle. A mechanism of feeding deterrence? J. Chem.
ECO. 11:73-83.

Mitchell ND (1977) Differential host selection by Eigrig
brassigag (The large white butterfly) on Braggiga

glgragga subsp. glgzagga (The wild cabbage). Ent.
exp. & appl. 22:208-219.

Myers JH, Post BJ (1981) Plant nitrogen and fluctuations of
insect populations:a test with the cinnabarmoth-tansy
ragwort system. Oecologia. 48:151-156.

 

 

148

Myers JH (1985) Effect of physiological condition of the
host plant on the ovipositional choice of the cabbage
white butterfly, Eigrig zapa_. Jour. of Animal Eco.
54:193-204.

Natr L (1972) Influence of mineral nutrients on photsyn-
thesis of higher plants. Photosynthetica. 6:80-99.

Nowacki M, Jurzysta M, Gorski P (1975) Effect of availa-
bility of nitrogen on alkaloid synthesis in Solana;
gage. Bull. Acad. Pol. Sc. Ser. Sc. Biol. 23:219-225.

Nowacki E, Jurzqsta M, Gorski P, Nowacka D, Waller GR (1976)
Effect of nitrogen nutrition on alkaloid metabolism in
plants. Biochem. Physiol. Pflanzen. 169:231-240.

Papaj DR, Rausher MD (1987) Components of conspecific host
discrimination behavior in the butterfly Battus agile; 1
nor. Ecology. 68:245-253.

Prestidge RA, McNeill S (1982) Instar duration, adult con- L
sumption, oviposition and nitrogen utilization ef-
ficiencies of leafhoppers feeding on different quality
food. Ecological Entomol. 7:91-101.

 

Prestidge RA, McNeill S (1983) The role of nitrogen in the
ecology of grassland Auchenorryncha. In: Nitrogen as
an Ecological Factor. (Lee JA, McNeill S, Rorrison HI,
eds.). Symp. British Ecological Soc. Blackwell Scien-
tific Publications. London.

Radin JW, Parker LL (1979) Water relations of cotton plants
under nitrogen deficiency. I. Dependence upon leaf
structure. Plant Physiol. 64:495-498.

Rausher MD, Papaj DR (1983) Demographic consequences of
discrimination among conspecific host plants by fiattgs
philgng; butterflies. Ecology. 64: 1402-1410.

 

Rausher MD (1979) Larval habitat suitability and oviposi-
tion preference in three related butterflies. Eco-
logy. 60:503-511.

Reese JC (1978) Chronic effects of plant allelochemics on
insect nutritional physiology. Ent. exp. & appl.
24:425-431.

Reese JC (1983) Nutrient-allelochemical interactions in
hostplant resistance. In: Plant Resistance to Insects
(Hedin PA, ed.). ACS Symposium Series, No. 208. Ameri-
can Chemical Society. Washington, DC.

149

Renwick JAA (1983) Non-preference mechanisms: plant charac-
teristics influencing insect behavior. In:Plant Resis-
tance to Insects. (Hedin PA, ed.). ACS Symposium
Series, No. 208. American Chemical Society, Washing-
ton, DC.

Renwick JAA, Radke CD (1982) Ovipositional choice and larval

survival of the Spruce Budworm, gaggiaaaaaara {Quiz
faraaa (Lepidoptera: Tortricidae). Env. Ent. 11:503-
505.

Rhoades DF, Cates RG (1976) Toward a general theory of plant
antiherbivore chemistry. Rec. Adv. Phytochem. 10:168-
213.

Robinson T (1979) The evolutionary ecology of alkaloids.
In:Herbivores, Their Interactions with Secondary
Metabolites. (Rosenthal GA, Janzen DH, Applebaum, SW,
eds.). Academic Press. New York.

Selvendran RR, Selvendran S (1973) Chemical changes in young

tea plant (Camallia ainanaia L.) tissues following ap-
plication of fertilizer nitrogen. Ann. Bot. 453-461.

Scriber JM (1978) The effects of larval feeding speciali-
zation and plant growth form on the consumption and
utilization of plant biomass and nitrogen: an ecolo-
gical consideration. Ent. exp. & appl. 24:494-510.

Scriber JM, Slansky F (1981) The nutritional ecology of
immature insects. Ann. Rev. Entomol. 26:183-211.

Shimishi D (1970) The effect of nitrogen supply on some
indices of plant-water relations of beans (Enaaaglaa
yalgaria L.) New Phytol. 69:418-424.

Shorey HH (1964) The biology of Triangplaaia n1 (Lepidop-
terazNoctuidae). III. Response to the oviposition
substrate. 57:165-170.

Singer MC (1986) The definition and measurement of ovipo-
sition preference in plant-feeding insects. In:
Insect-Plant Interactions. (Miller JR and Miller TA,
eds.). Springer-Verlag. New York.

Singer MC (1984) Butterfly-hostplant relationships: host
quality, adult choice and larval success. In: The
Biology of Butterflies. (Vane-Wright RI, Ackery PR,
eds.). Academic Press.

Slansky F (1982) Toward a nutritional ecology of insects.
Proc. Fifth Int. Symp. Insect-Plant Relationship.
PUDOC, Wageningen.

 

 

150

Slansky F, Feeny P (1977) Stabilization of nitrogen accu-
mulation by larvae of the cabbage butterfly on wild
and cultivated food plants. Ecol. Monogr. 47:209-228.

Slansky F, Scriber JM (1985) Food consumption and utiliz-
ation. In: Comprehensive Insect Physiology Biochemi-
stry and Pharmacology (Kerkut GA, Gilbert LI, eds.).
Pergamon Press.

Solkal RR, Rohlf FJ (1981) Biometry: The principles and
practice of statistics in biological research. Free-
man. San Fransico.

Sparks MR (1973) Physical and Chemical stimuli affecting
oviposition preference of Mandaga aaxta (Lepidop-
tera: Sphingidae). Ann. Ent. Soc. Am. 66:571-73.

Taylor MFJ (1984) The dependence of development and fecun-

dity of Saaaa malgipligalia on early larval nitrogen
intake. Insect Physiol. 30:779-785.

Thorsteinson AJ (1960) Host selection in phytophagous in-
sects. Ann. Rev. Ent. 5:193-218.

Tichenor LH, Seigler DS (1980) Electroantennogram and ovipo-
sition responses of Mangaga aaxga to volatile com-
ponents of tobacco and tomato. J. Insect Physiol.
26:309-314. ‘

Treshow M (1970) Environment and Plant Response. McGraw-
Hill.

Via S (1986) Genetic covariance between oviposition prefe-
rence and larval performance in an insect herbivore.
Evolution. 40:778-785.

Williams KS (1983) The coevolution of Eaphydryaa ghalgaggaa
butterflies and their larval host plants. III. Ovipo-

sition behavior and host plant quality. Oecologia.
56:336-340.

Wolfson JL (1978) Environmentally induced variation in
fixaaaiga nigra (Koch): affects on generalists and
specialist herbivores. Ph.D. dissertation. State
Univ. of New York. Stony Brook.

Wolfson JL (1980) Oviposition response of Biazia raga; to
environmentally induced variation in Brassica nig_a.
Ent. exp. & appl. 27:223-232.

Yamamoto RT, Fraenkel GS (1960a) The specificity of tobacco
hornworm, Bxgtgparga se t , to solanaceous plants.
Ann. Ent. Soc. Amer. 53:503-507.

151

Yamamoto RT, Fraenkel GS (1960c) The physiological basis
for the selection of plants for egg-laying in the
tobacco hornworm, Rrgtgpazga aaxga (Johan.) XI Int.
Kongr. ent. Wien. 3:127-133.

 

 

POSTCRIPT

As this research project progressed, it became ap-
parent that few investigators had examined conspecific host
selection. This is unfortunate. Based on the findings pre-
sented in this thesis, it seems obvious that intraspecific
discriminatory behaviors are as important and have as many
implications as interspecific ones. Results demonstrated
that nitrogen fertilization of field grown tomato plants
induced physiological changes in the plants which, subse-
quently, influenced conspecific host discrimination and
larval performance in Haggaga_app. Plants treated at higher
fertilization rates elicited a greater response in gravid
females and produced larvae with a greater overall fitness.
Results further indicated that plants most preferred by
females also conferred the highest fitness on larvae reared
on these plants. Future studies should focus on describing
these behaviors and elucidating the components involved.
This study, alone, generated a number of interesting ques-
tions.

The observation that plants treated at low levels of
nitrogen fertilizer elicited the lowest response in moths
and supported the lowest larval performance is of particular
interest. It appears that small amounts of fertilizer
induced a biochemical reaction in the plant that repelled

152

153
the moths and inhibited feeding and digestion in the larva.
Unfortunately, because of the number of plant components
that are affected by nitrogen availability and the complex-
ity of their interactions, it is impossible to conclude
which factors, or combination thereof, elicited these re-
sponses.

The finding that female moths prefer to oviposit on
those plants conferring the highest fitness on their off-
spring presents a whole area of research that remains to be
described. Observations on individuals within a family,
instead of the population level approach utilized in this
study, is a more appropriate method for determining the
physiological basis of these responses. Experiments with
full sibs could provide basic information regarding the
genetic components that induce these behaviors. Simple
experiments could determine, and possibly rule out, the
existence of host plant conditioning fairly quickly. If the
behavioral response has a genetic basis, then similar, but
more detailed backcrossing studies could be utilized to ex-
amine the gene complexes involved.

The ecological implications of these findings are
important. Moths selecting plants with a higher nutritional
content will produce offspring that complete development
earlier, suffer lower mortality and develop into more fecund
adults. As a result the female will increase her genetic
contribution to the next generation. For ecologists and

evolutionary biologists interested in examining these

154

plant/herbivore interactions, this system provides a method
for investigating the components of nautral selection and
other related processes in phytophagous insects.

The economical considerations are also significant.
The development of cultural programs that consider the
affect of nitrogen fertilization on insect behavior and
performance may enable the grower to balance the economical
advantages of increased yield with the potential disad-
vantages of increased pest populations. Furthermore, re-
searchers investigating plant resistance may be able to
develop a program that confers resistance on crop plants by
manipulating the nutritional quality of those plants. The
results of this study suggest that low levels of nitrotgen
fertilizer may actually repel insects. Hence, it would be
interesting to determine if other insect populations demon-
strate similar responses. At the very least, a system that
examines insect abundance in relation to host quality could
provide important information on the plant factors that

induce resistance in certain crop plants.