HOST COLONIZATION PATTERNS, CUES MEDIATING HOST SELECTION
AND CALIBRATION OF FIELD SURVEYS WITH ESTIMATES OF POPULATION
ABUNDANCE OF LEPTOGLOSSUS OCCIDENTALS IN A SEED ORCHARD

by

Tamara A. Richardson
B.Sc., University of British Columbia, 2007

THESIS SUBMITTED IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
IN
NATURAL RESOURCES AND ENVIRONMENTAL STUDIES
(BIOLOGY)

UNIVERSITY OF NORTHERN BRITISH COLUMBIA
May 2013

© Tamara A. Richardson, 2013

UMI Number: 1525670

All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.

UMI
Dissertation PiiblishMiQ

UMI 1525670
Published by ProQuest LLC 2014. Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code.

ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106-1346

ii

Abstract

The western conifer seed bug, Leptoglossus occidentalis Heidemann (Hemiptera:
Heteroptera: Coreidae) is an important pest in southern interior British Columbia seed
orchards. In spite of its pest status, host colonization patterns and cues mediating host
selection have not been well studied. Additionally standardized population census methods,
which are essential for the development of economic damage thresholds, have not yet been
established. I examined spring immigration patterns of L. occidentalis into a lodgepole pine
seed orchard in 2008 and 2009. In 2008, significant spatial gradients of L. occidentalisinfested trees were found in four of seven surveys, however they were not directionally
consistent. In 2009 a significant spatial gradient was found in a single survey. Spatial
gradients may have been detected more easily in 2008 than 2009, as counts of L. occidentalis
were much larger. I also assessed whether L. occidentalis exhibits clonal preference, if
preferences remained consistent between 2008 and 2009, and examined whether cues
responsible for such preferences could be identified. I found that clone preference remained
consistent between 2008 and 2009 and that predefined clone preference classes had
significantly different levels of the monoterpenes a-pinene, 8-3 carene, and p-cymene.
Additionally, L. occidentalis was found more frequently on clones with cones of greater
diameter and weight. Finally, in 20091 conducted mark-release-recapture studies with
corresponding walk-through surveys for L. occidentalis both in lodgepole pine and Douglasfir seed orchards. I was unable to establish mathematical relationships between population
estimates and walk-through surveys in either location. Data from these surveys indicated that
the walk-through survey were unreliable, at least at low populations. Studies exploiting cues
attractive to L. occidentalis for trap design and additional studies to establish standardized

survey methods are recommended as L. occidentalis continues to cause damage in North
American seed orchards and has spread internationally establishing itself in new areas with
vulnerable and economically important crops.

Table of Contents

Abstract

ii

Table of Contents

iv

List of Tables

vi

List of Figures

viii

Acknowledgement

xi

Chapter One

Introduction
1.1 Background
1.2 Study objectives
1.3 Literature cited

Chapter Two

Host colonization patterns of Leptoglossus occidentalis
in seed orchards after overwintering
2.1 Abstract
2.2 Introduction
2.3 Methods
2.3.1 Study area and species
2.3.2 Survey methods
2.3.3 Trapping and marking methods
2.3.4 Statistical analysis
2.4 Results
2.4.1 Winter and spring temperatures, timing of spring
immigration and the number of L. occidentalis found
during surveys
2.4.2 Spatial gradient of L. occidentalis in lodgepole pine
orchard 307
2.5 Discussion
2.6 Literature cited

Chapter Three

1
1
5
7

Cues mediating host selection of Leptoglossus occidentalis on
lodgepole pine
3.1 Abstract
3.2 Introduction
3.3 Methods
3.3.1 Study area and species
3.3.2 Insect monitoring
3.3.3 Clone ranking and clone preference classification
3.3.4 Host selection cues
3.3.5 Statistical analysis
3.4 Results
v

10
10
10
14
14
14
16
17
19
19

21
28
31

35
35
35
39
39
39
40
40
45
46

3.4.1 Clone ranking and clone preference classification
3.4.2 Host selection cues
3.5 Discussion
3.6 Literature cited

46
51
64
73

Chapter Four

Calibration of walk-through field survey estimates of
Leptoglossus occidentalis (Hemiptera: Heteroptera: Coreidae) with
abundance estimates based on mark-release-recapture techniques
80
4.1 Abstract
80
4.2 Introduction
81
4.3 Methods
84
4.3.1 Study area
84
4.3.2 Trapping and marking methods
84
4.3.3 Population estimation methods
88
4.3.4 Statistical analysis
90
4.4 Results
90
4.4.1 Jolly-Seber population estimates in lodgepole pine
orchard 307
90
4.4.2 Schnabel, Schumacher-Eschemeyer, Jolly-Seber and
whole-tree methods of population estimates in Douglas-fir
block 9
95
4.5 Discussion
98
4.6 Literature cited
104

Chapter Five

General discussion
5.1 Literature cited

108
114

Appendix I

118

Appendix II

121

Appendix III

122

Appendix IV

127

List of tables

Table 2.1 Average monthly maximum, minimum and mean temperatures from December
through May 2007/2008 and 2008/2009.
20
Table 2.2 Extreme monthly maximum and minimum temperatures from December through
May 2007/2008 and 2008/2009.
20
Table 2.3 The number of L. occidentalis caught and marked during 2008 and 2009 surveys.
24
Table 2.4 Direction of statistically significant spatial gradient and the probability of
encountering L. occidentalis infested trees in surveys from 2008 and 2009. Spatial gradient is
measured in metres along the x (323 metres) and >>(117 metres) axes with the point 0,0
representing the southwest comer of the orchard.
26
Table 2.5 Direction of statistically significant spatial gradient and the probability of
encountering L. occidentalis infested trees in 2009 surveys. Spatial gradient is measured in
metres along the x (323 metres) andy (117 metres) axes with the point (0,0) representing the
southwest comer of the orchard.
27
Table 3.1 Clones sampled in 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed
orchard in Vemon B.C., Canada. Each clone that is from the same provenance is from a
different family. Numbers in parenthesis indicate how many ramets of that clone were
planted in a given year.
48
Table 3.2 Results of mixed effects analyses of variance for infrared and physical attribute
measurements from all sampling events in 2008 and 2009 in seed orchard 307 at Kalamalka
Lake Seed orchard in Vemon B.C., Canada.
52
Table 4.1 2008 population estimates of L, occidentalis per 4 ha in lodgepole pine orchard 307
from mark-release-recapture data using the Jolly-Seber method of population estimation at
the Kalamalka Lake Seed orchard in Vemon B.C., Canada. Confidence intervals are within
parentheses.
92
Table 4.2 2009 population estimates of L. occidentalis per 4 ha in lodgepole pine orchard 307
from mark-release-recapture data using the Jolly-Seber method of population estimation at
the Kalamalka Lake Seed orchard in Vemon B.C., Canada. Also included are the number of
L. occidentalis found during walk-through surveys. Confidence intervals are within
parentheses.
93
Table 4.3 Total number of L. occidentalis captured during each sampling interval in 2008,
including single and double recaptures at the Kalamalka Lake Seed orchard in Vemon B.C.,
Canada. Surveys were done on 5 May and 10 May, 2008 but no L. occidentalis were sighted
or captured during this time.
94

Table 4.4 Total number of L. occidentalis captured during each sampling interval in 2009
including single and double recaptures in orchard 307 at the Kalamalka Lake Seed orchard in
Vemon B.C., Canada.
94
Table 4.5 Population estimates and of L. occidentalis per 0.5 ha in Douglas-fir Block 9 using
the Schnabel, Schumacher-Eschemeyer and Jolly-Seber estimation methods at the Kalamalka
Lake Research Station in Vemon B.C., Canada. Also included are the numbers o f L.
occidentalis found during walk-through surveys. Confidence limits are within parentheses.
96
Table 4.6 Population estimates of L. occidentalis per 0.5 ha in Douglas-fir Block 9 using the
whole-tree method at the Kalamalka Lake Research Station in Vemon B.C., Canada. Also
included are the numbers of L. occidentalis found during walk-through surveys. Confidence
limits within parentheses.
97
Table AI. 1 Results of logistic regression models for determining spatial trends in 2008.

118

Table AI.2 Results of logistic regression models for determining spatial trends in 2009.

119

Table AI.3 Chi-square test statistics and AIC values used to select best fit spatial point
process models for 2008 and 2009 surveys.

120

Table AII.1 Results of mixed effects analyses of variance for terpenoids from all sampling
events 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in Vemon B.C.,
Canada.
121
Table AIV.l Comparison of Number of L. occidentalis sighted by each scout in 2008 in
orchard 307 at Kalamalka Seed Orchard Vemon B.C. Canada. Survey dates included are
those where scouts spent an equal amount of time surveying.
127
Table AIV.2 Comparison of Number of L. occidentalis sighted by each scout in 2009 in
orchard 307 at Kalamalka Seed Orchard Vemon B.C. Canada. Survey dates included are
those where scouts spent an equal amount of time surveying.
127

List of Figures

Figure 2.1 a-h) Density plots of L. occidentalis infested trees in orchard 307 at the Kalamalka
Seed Orchard in Vemon B.C., Canada from a) 5 May 2008, b) 8 May 2009, c) 17 May 2008,
d) 15 May 2009, e) 23 May 2008, f) 23 May 2009, g) 28 May 2008, h) 28 May 2009. A
significant spatial gradient in L. occidentalis infested trees was observed in the 23 May 2008
survey. The plot axes are to scale with orchard 307 with the point 0, 0 being the southwest
comer of orchard 307. Note that the scale on the right-hand side of each plot (density of
insect-infested trees per m2) is different for each plot.
22
Figure 2.2 a-f) Density plots of L. occidentalis infested trees in orchard 307 at the Kalamalka
Seed Orchard in Vemon B.C., Canada from a) 4 June 2008, b) 2 June 2009, c) 10 June 2008,
d) 13 June 2009, e) 20 June 2008 and f) 18 June 2009. A significant spatial gradient in L.
occidentalis infested trees was observed in all three June 2008 surveys, and in the 2 June
2009 survey. The plot axes are to scale with orchard 307 with the point 0, 0 being the
southwest comer of orchard 307. Note that the scale on the right-hand side of each plot
(density of insect-infested trees per m2) is different for each plot.
23
Figure 3.1 Correlation between clone rank in 2008 and 2009 in seed orchard 307 at
Kalamalka Lake Seed orchard in Vemon B.C., Canada. Clones with the same rank were
averaged.

47

Figure 3.2 a-d) Canonical scores plots from classical discriminant analysis illustrating the
separation between the two favoured clone preference classes (o = favoured occupied; • =
favoured unoccupied; ▼= non-favoured clones) sampled from seed orchard 307 at
Kalamalka Lake Seed orchard in Vemon B.C., Canada.
50
Figure 3.3 Average maximum cone temperature in "Celsius (± 1 S.E.) found in cones from
thel5-17 July, 2009 sampling period in seed orchard 307 at Kalamalka Lake Seed orchard in
Vemon B.C., Canada. The presence of different lower case letters indicates a significant
difference between clone preference classes.
53
Figure 3.4 a-d) Mean levels of 5-3-carene in parts per million (± 1 S.E.) found in cones from
sampling events in 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in
Vemon B.C., Canada. The presence of different lower case letters indicates significant
difference between clone preference classes.
56
Figure 3.5 a-d) Mean levels of a-pinene in parts per million (± 1 S.E.) found in cones from
sampling events in 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in
Vemon B.C., Canada. The presence of different lower case letters indicates significant
difference between clone preference classes.
57

Figure 3.6 a-d) Mean levels of limonene in parts per million (± 1 S.E.) found in cones from
sampling events in 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in
Vemon B.C., Canada. The presence of different lower case letters indicates a significant
difference between clone preference classes.
58
Figure 3.7 a-d) Mean levels of p-cymene in parts per million (± 1 S.E.) found in cones from
sampling events in 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in
Vemon B.C., Canada. The presence of different lower case letters indicates a significant
difference between clone preference classes.
59
Figure 3.8 a-d) Mean cone length in centimetres (± 1 S.E.) found in cones from all sampling
events 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in Vemon B.C.,
Canada. The presence of different lower case letters indicates a significant difference
between clone preference classes.
61
Figure 3.9 a-d) Mean cone diameter in centimetres (± 1 S.E.) found in cones from all
sampling events in 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in
Vemon B.C., Canada. The presence of different lower case letters indicates a significant
difference between clone preference classes.
62
Figure 3.10 a-b) Mean cone weight in grams (± 1 S.E.) found in cones from 2008 and 2009 in
seed orchard 307 at Kalamalka Lake Seed orchard in Vemon B.C., Canada. The presence of
different lower case letters indicates a significant difference between clone preference
classes.
63
Figure 4.1 Marked L. occidentalis. In a) 2008 insects were coded with Liquitex™
professional acrylic artist water-based colour paint and Pigma Micron™ archival ink pens
and in b) 2009 insects were coded with SRX Metallic Colorsharp™ permanent marker pens
and Pigma Micron™ archival ink pens. Photo a) was taken on 25 June 2008 by Ward Strong
in his lab at the KalamalkaResearch Station, Vemon B.C., Canada.Photo b) was taken on
20 August 2009 in block 8 by Tamara Richardson at theKalamalka Research Station, Vemon
B.C., Canada.
86
Figure 4.2 L.occidentalis on a) lodgepole pine cones and b) Douglas-fir cones. On lodgepole
pine it can be easily overlooked as the colour of the sclerotized portion of the forewing and
dried male cones are very similar. On mature Douglas-fir cones the adult insect can hide
between cone scales rendering it difficult to sight. Photograph a) was taken by Tamara
Richardson on 25 June, 2008 in orchard 307 at the Kalamalka Seed Orchard in Vemon B.C.,
Canada.
103

x

Acknowledgements

I was extremely fortunate and am eternally grateful to have been mentored by several
exceptional people while working on this project. I could not have had a better supervisor
than Dr. Staffan Lindgren whose insight, encouragement and patience made working on this
project a truly enjoyable experience despite moments of frustration. I am enormously
grateful to have been able to work with Dr. Ward Strong who provided invaluable advice, a
fun work environment and much needed support in the design and implementation of this
project. Moreover, this project would not have occurred had he not secured funding from
the Forest Genetics Council Pest Management Advisory Committee. I am very lucky to
have had Dr. Brian Aukema both on my committee and as a statistics professor. In his
classes my fear of statistics and math were replaced by curiosity and a confidence in myself I
would have previously not thought possible. I am also grateful to Dr. Sylvie Desjardins
whose ideas got this project rolling.
I am indebted to Sydney Smith and Nicole Tunbridge for their long hours toiling in
the field with me, their friendship and for making the field seasons fun. I would like to
acknowledge Tracy Zaradnik and Steve Takacs from Simon Fraser University for their
assistance with the host selection work and Darby Strong, Jake King, Jeremy Neilson and
Kari Kirkpatrick for their help with fieldwork. I also want to thank the staff at the Kalamalka
seed orchard, particularly Chris Walsh and Gary Giampa for their insights and input.
I want to acknowledge my PG friends and my family. My experience of Prince
George would not have been as much fun or nearly as interesting without Angelique and
Thibault Grava, Sean Sweeney, Alyssa Shaw, Sonja Foss, Sonja Ostertag and Dave Stinson. I
could not have imagined that I would develop such profound friendships in such a short time;
Finally, I am grateful to my parents for instilling in me a keen interest in animals with
exoskeletons, respect for the natural world and for their love, support and enthusiasm for my
work.

Chapter 1. Introduction

1.1 Background
Conifer seed orchards produce large quantities of seeds of high genetic value to be
used in reforestation in British Columbia with the goal of creating high-quality and
genetically diverse forests (British Columbia Ministry of Forests and Range 1996). The seed
that is produced in these orchards is the result of extensive breeding programs. Tree breeding
is a lengthy process, which includes the identification of so-called ‘plus trees.’ These ‘plus
trees’ are wild parent trees that possess desirable heritable attributes. Seed orchard trees are
the product of these parent trees, and are bred using traditional breeding techniques for
improved growth, wood quality, and pest resistance (British Columbia Ministry of Forests
and Range 1996, Forest Genetics Council of British Columbia 2008).
British Columbia seed orchards currently produce approximately 55% of the seed
used in provincial reforestation efforts. The British Columbia Ministry of Forests and Range
aims to increase this proportion to 75% by 2014 (Forest Genetics Council of BC, 2009). A
major obstacle in achieving the proposed 36.4% increase is the loss of seed in orchards
resulting from damage attributed to conophagous insects. Several insect species threaten
seed production in British Columbia seed orchards. The most economically important species
are the Douglas-fir cone gall midge, Contarinia oregonensis Foote (Diptera: Cecidomyiidae),
and the fir cone-worm, Dioryctria abietivorella Grote (Lepidoptera: Pyralidae), both of
which can destroy entire cone crops (British Columbia Ministry of Forests and Range 2009a,
2009b). For example, $600,000 worth of damage to seed crops was caused in a 2004
outbreak of the latter species (British Columbia Ministry of Forests and Range 2009b).
1

The western conifer seed bug, Leptoglossus occidentalis Heidemann (Hemiptera:
Heteroptera: Coreidae), is also an important cone and seed pest in conifer seed orchards in
southern interior British Columbia (Strong et al. 2001) and is responsible for varying degrees
of seed damage in several conifer species (Hedlin et al. 1981). Although pesticides are used
multiple times per season to combat L. occidentalis in seed orchards, substantial seed loss
still occurs in lodgepole pine (Pinus contorta Douglas ex Loudon var. latifolia Engelm.) and
Douglas-fir (Pseudotsuga menziesii Mirb. Franco); seed loss is less significant in numerous
other conifer species such as spruce (Picea spp.) and larch (Larix spp.) (Hedlin et al. 1981,
Bates et al. 2000, Strong et al. 2001).
The western conifer seed bug, L. occidentalis has the potential to destroy up to 83%
of a seed crop in a season (British Columbia Ministry of Forests and Range 2009c). In a
good year, up to 212 kilogram (kg) of interior lodgepole pine seed, at a value of $6,450 per
kg (Vemon Seed Orchard 2010), can be produced (Forest Genetics Council of BC 2010). The
potential economic loss of lodgepole pine seed to L. occidentalis is estimated at over
$1,000,000 per year.
Leptoglossus occidentalis leave their overwintering sites and arrive in seed orchards
in the spring. L. occidentalis locates its host trees by using specialized infrared receptor
organs situated on the ventral abdomen. Thermal imaging has demonstrated that there is up to
a 15°C difference in temperature between needles and seed cones from spring through fall in
the trees that host L. occidentalis, and seed cones emit substantially greater mid- and longrange infrared radiation than foliage. Both male and female L. occidentalis have been shown
to prefer strong infrared radiation cues to weaker infrared radiation cues, and likely use the
infrared differential between foliage and cones to find the cones upon which they feed
(Takacs et al. 2008a). Once L. occidentalis arrive in seed orchards they exhibit pronounced
2

clonal preferences, typically settling on trees of moderate height with a moderate cone crop
(Blatt and Borden 1999). After L. occidentalis arrive on a host tree, they begin to feed by
inserting their proboscis into developing cones and secrete enzymes and digest the rich lipid
and protein content of seeds (Bates et al. 2000). Mating commences shortly after the insects
colonize seed orchards and continues throughout the summer. Males produce vibratory
signals that may be involved in mating. These vibratory signals are 20 decibels above the
human hearing threshold and are created by males tapping their abdomen on substrate
(Takacs et al. 2008b). Females oviposit on the undersides o f conifer needles, laying up to 80
eggs in groups of 4-10 (Hedlin et al. 1981). Eggs hatch within approximately two weeks,
and nymphs develop through five instars before reaching maturity.
Leptoglossus occidentalis have several natural enemies in British Columbia, which
target different life-history stages. The parasitoids Gryon pennsylvanicum Ashmead
(Hymenoptera: Platygastridae), Ooencyrtus johnsoni Howard (Hymenoptera: Encyrtidae),
and Anastatus pearsalli Anaupe (Hymenoptera: Eupelmidae) are known to complete their life
cycle in L. occidentalis eggs (Bates and Borden 2004, Maltese et al. 2012). Wasps (Vespidae
spp.) have been observed eating L. occidentalis nymphs and adults throughout the summer,
and Magpies have been observed feeding on L. occidentalis adults as they aggregate on
wooden structures in the fall prior to overwintering (personal observation; W.B. Strong,
personal communication1).
First generation adults die from late July through August overlapping with the
maturing second generation of adults. The second generation of adults form large groups
mediated by aggregation pheromones, and migrate to protected overwintering sites (Blatt and

1Research scientist, Tree Improvement Branch, Ministry of Forest, Lands and Natural
Resource Operations, Kalamalka Forestry Centre, 3401 Reservoir Road, Vemon BC.
3

Borden 1996). Overwintering sites are located inside and outside of seed orchards and
include buildings, shipping containers, and well-protected outdoor areas (Hedlin et al. 1981,
Taylor and Villa 2001). L. occidentalis have also been found overwintering in cavities of
Douglas fir and in bird and rodent nests (Koerber 1963). L. occidentalis is found throughout
North America and has spread throughout Europe and beyond. First detected in Italy in 2001
(Taylor and Villa 2001), it has also been found in Sweden (Lindelow and Bergsten 2012),
Norway (Mjos 2010), Slovakia (Barta 2009), Turkey (Fent and Kment 2011), and Japan
(Ishikawa and Kikuhara 2009).
Previous research has confirmed that feeding by nymph and adult stages of L.
occidentalis significantly reduces seed set in lodgepole pine (Strong et al. 2001, Bates et al.
2002a, Bates et al. 2002b), Douglas-fir (Bates et al. 2000, Bates et al. 2001), ponderosa pine
{Pinus ponderosa Dougl. ex P. & C. Lawson) (Connelly and Schowalter 1991), and western
white pine (P. monticola Dougl. ex D. Don) (Bates et al. 2002b). Feeding by L. occidentalis
also causes increased conelet abortion in western white pine (Bates et al. 2002b). While
nymphs and adults of both sexes can cause significant seed reduction throughout the spring
and summer, more lodgepole pine seeds are damaged by early season adult females than by
any other life stage (Bates et al. 2002a, Strong 2006). Seed consumption by overwintered
females is greatest early in the growing season due to the high energetic cost of reproduction,
which occurs at this time (Bates and Borden 2005, Strong 2006). Using a simple model of
population survival and impact data for each life stage, Bates and Borden (2005) estimated
that at a hypothetical density of one seed bug per tree, a single founding insect and its
offspring could cause a loss of approximately 310 seeds per tree per year
Seed orchards are comprised of multiple copies (ramets) of high quality clones that
are grafted onto rootstock. Ramets of each clone are spatially referenced and randomly
4

distributed throughout the seed orchard. The Kalamalka seed orchard, located 3 km south of
Vemon, British Columbia, is owned and operated by the BC Ministry of Forests, Lands and
Resource Operations. The orchard site spans 32 hectares on a southern slope facing
Kalamalka Lake. The elevation of the site ranges between 450 and 475 metres above sea
level, and is in the Interior Douglas-fir Biogeoclimatic zone (Meidinger and Pojar 1991). The
area is characterized by mild dry winters and hot dry summers. The hot dry summers are
ideal for stimulating cone production, which is critical to seed production (BC Ministry of
Forests and Range 1996). L. occidentalis is a serious pest at this orchard, and management of
the insect is maintained by calendar-sprays of pesticides. Monitoring is difficult and time
consuming as there are no effective traps, and damage estimation methods are unreliable and
therefore difficult to use in pest management plans. Addressing these issues is required to
develop improved management plans for L. occidentalis.
1.2 Study Objectives
In this thesis, I examine spring colonization patterns of L. occidentalis in a lodgepole
pine seed orchard (Chapter 2) and cues influencing host selection of lodgepole pine (Chapter
3), and estimate population abundance based on mark-release-recapture methodology to
assess the accuracy of a visual monitoring system (W.B. Strong1, unpublished data)
employed by seed orchard staff in a lodgepole pine seed orchard and in a Douglas-fir block
(Chapter 4).
In Chapter 2 ,1 look specifically at the timing of insect host colonization in a
lodgepole pine seed orchard, and the pattern of spatial spread of L. occidentalis throughout
the summer to determine if spring invasion of L. occidentalis into a lodgepole pine orchard
exhibits an edge effect on orchard boundaries, and if so, the duration of the effect. This has
strong management implication for early-season control and development of appropriate
5

census methods. In Chapter 3 ,1 establish clone preferences of L. occidentalis, compare
clonal differences in host-selection cues, and examine changes in clone preference from year
to year. L. occidentalis preference for specific clones was first demonstrated by Blatt and
Borden (1999), and identification of specific factors influencing clonal preference may have
significant management implications. For example, such knowledge may aid in the
development of census methods and in the design of traps and orchards. In Chapter 4 ,1
compare operational walk-through field survey estimates of L. occidentalis with abundance
estimates based on mark-release-recapture. A better understanding of the relationship
between survey walk-through estimates, which are currently used for relative population
assessments (W.B. Strong1, unpublished data), and the absolute population abundance
estimates could validate current management practices. In Chapter 5 ,1 synthesize key results
and provide comprehensive management recommendations to seed orchard managers.
I hope that by addressing the research questions listed above, I can provide critical
knowledge for the development of effective pest management strategies for L. occidentalis in
conifer seed orchards in British Columbia. Determining early season immigration patterns
and important host-selection factors, and developing effective quantifiable field monitoring
methods, will increase our understanding of L. occidentalis population processes. This
knowledge is necessary to improve L. occidentalis management practices, and provide pest
management strategies, which will enable seed orchard managers to achieve the increased
seed yield required to meet provincial reforestation goals (Forest Genetics Council of BC,
2009).

6

1.3 Literature Cited

Barta, M. 2009. New facts about distribution and host spectrum of the invasive nearctic
conifer pest, Leptoglossus occidentalis (Heteroptera: Coreidae) in south-western
Slovakia. Forestry Journal. 55:139-144.
Bates, S. L. and J. H. Borden. 2004. Parasitoids of Leptoglossus occidentalis Heidemann
(Heteroptera: Coreidae) in British Columbia. Journal of the Entomological Society of
British Columbia. 101:143-144.
Bates, S.L. and Borden, J.H. 2005. Life table for L. occidentalis Heidemann (Heterpotera:
Coreidae) and prediction of damage in lodgepole pine seed orchards. Agricultural and
Forest Entomology. 7:145-151.
Bates, S.L., Borden, J.H., Kermode, A.R., and Bennett, R.G. 2000. Impact of L. occidentalis
on Douglas-fir seed production. Journal of Economic Entomology. 93:1441-1451.
Bates, S.L., Lait, C.G., Borden, J.H., and Kermode, A.R. 2001. Effect of feeding by the
western conifer seed bug L. occidentalis on the major storage reserves of developing
seeds and on seedling vigor of Douglas fir. Tree Physiology. 21:481-487.
Bates, S.L., Lait, C.G., Borden, J.H., and Kermode, A.R. 2002a. Measuring the impact of L.
occidentalis on seed production in lodgepole pine using an antibody-based assay.
Journal of Economic Entomology. 95:770-777.
Bates, S.L., Strong, W.B., and Borden, J.H. 2002b. Abortion and seed set in lodgepole and
western white pine conelets following feeding by L. occidentalis (Heterpotera:
Coreidae). Environmental Entomology. 31:1023-1029.
Blatt, S.E. and Borden J.H. 1999. Physical characteristics as potential host selection cues for
L. occidentalis (Heteroptera: Coreidae). Environmental Entomology. 28: 246-254.
British Columbia Ministry of Forests and Range. 1996. Kalamalka Forestry Centre Research
and Seed Orchard. Available from
http://www.for.gov.bc.ca/hfd/librarv/documents/bibl5212.pdf [accessed 2009-05-17]
British Columbia Ministry of Forests and Range. 2009a. Cone and Seed Insect Pest Leaflet
No. 2. Douglas fir cone gall midge: Contarinia oregonensis. British Columbia
Ministry of Forests and Range, Tree Improvement Branch, Saanichton, B.C.
Available from http://www.fgcouncil.bc.ca/ptac-area-leaflets.html [accessed 2011-0115]
British Columbia Ministry of Forests and Range. 2009b. Cone and Seed Insect Pest Leaflet
No. 3. Fir cone-worm: Dioryctria abietivorella. British Columbia Ministry of Forests
and Range, Tree Improvement Branch, Saanichton, B.C. Available from
http://www.fgcouncil.bc.ca/ptac-area-leaflets.html [accessed 2011-01-15]
7

British Columbia Ministry of Forests and Range. 2009c. Cone and Seed Insect Pest Leaflet
No. 4. Western conifer seed bug: Leptoglossus occidentalis. British Columbia
Ministry of Forests and Range, Tree Improvement Branch, Saanichton, B.C.
Available from http://www.fgcouncil.bc.ca/ptac-area-leaflets.html [accessed 2011-0115]
Connelly, A. E. and Showalter, T.D. 1991. Seed losses to feeding by L. occidentalis
(Heteroptera: Coreidae) during second year cone development in western white pine.
Journal of Economic Entomology. 84:215-217.
Fent, M., and Kment, P. 2011. First record of the invasive western conifer seed bug L.
occidentalis (Heteroptera: Coreidae) in Turkey. North-Western Journal of Zoology.
7:72-80.
Forest Genetics Council of British Columbia Annual Report 2007/2008. 2008. Available
from http://www.fgcouncil.bc.ca/FGCAnnRpt-0708-Web.pdf faccessed 2009-0311].
Forest Genetics Council of British Columbia Annual Report 2009/2010. 2010. Available
from www.fgcouncil.bc.ca/FGC-AnnReport-0910-Web.pdf [accessed 2011-01-16].
Forest Genetics Council of British Columbia Strategic Plan 2009. Available from
http://www.fgcouncil.bc.ca/StratPlan0914-Lavout-Web-22Dec09.pdfraccessed 2011 01-15].
Hedlin, A.F., Yates III, H.O., Tovar, D.C., Ebel, B.H., Koerber, T.W., and Merkel, E.P.
1981. Cone and Seed Insects of North American Conifers. Canadian Forestry
Service, USDA and Forest Service and Secretaria de Agriculture y Recursos
Hidraulicos, Mexico.
Ishikawa, T., and Kikuhara, Y. 2009. L. occidentalis Heidemann (Hemiptera: Coreidae), a
presumable recent invader to Japan. Japanese Journal of Entomology. 12:115-116.
Koerber, T. W. 1963.. occidentalis (Hemiptera, Coreidae), a newly discovered pest of
coniferous seed. Annals of the Entomological Society of America. 56:229-234.
Lindelow, A., and Bergsten, J. 2012. The invasive western conifer seed bug, L. occidentalis
(Heteroptera: Coreidae), established in Sweden. Entomologisk Tidskrift. 133:55-58.
Maltese, M., Caleca, V., Guerrieri, E., and Strong, W. B. 2012. Parasitoids of L. occidentalis
Heidemann (Heteroptera: Coreidae) recovered in western North America and first
record of its egg parasitoid Gryon pennsylvanicum (Ashmead) (Hymenoptera:
Platygastridae) in California. The Pan-Pacific Entomologist. 88:347-355.
Meidinger, D. and Pojar, J. 1991. Ecosystems of British Columbia. Special Report Series 6.
British Columbia Ministry of Forests, Victoria. British Columbia, Canada.
8

Mjos, A. T., Nielsen, T. R., and 0degaard, F. 2010. The western conifer seed bug
(Leptoglossus occidentalis Heidemann, 1910)(Hemiptera: Coreidae) found in SW
Norway. Norwegian Journal of Entomology. 57:20-22.
Strong, W.B. 2006. Seasonal changes in seed reduction in lodgepole pine cones caused by
feeding of L. occidentalis (Hemiptera: Coreidae). The Canadian Entomologist.
138:888-896.
Strong, W.B., Bates, S.L., and Stoehr, M.U. 2001. Feeding by L. occidentalis (Hemiptera:
Coreidae) reduces seed set in lodgepole pine (Pinaceae). The Canadian Entomologist.
133:857-865.
Takacs, S., Bottomley, H., Andreller, I., Zaradnik, T., Schwarz, J., Bennett, R., Strong W.
and Gries. G. 2008a. Infrared radiation from hot cones on cool conifers attracts seedfeeding insects. Proceedings of the Royal Society B-Biological Sciences. 276:649655.
Takacs, S., Hardin, K., Gries, G., Strong, W., and Bennett, R. 2008b. Vibratory
communication signal produced by male western conifer seed bugs (Hemiptera:
Coreidae). The Canadian Entomologist. 140:174-183.
Taylor, S. J., Tescari, G., and Villa, M. 2001. A Nearctic pest of Pinaceae accidentally
introduced into Europe: L. occidentalis (Heteroptera: Coreidae) in northern Italy.
Entomological News. 112:101-103.
Vemon Seed Orchard. 2010. Interior Lodgepole Pine Seedlot Bulletin 2010/2011. Available
from http://www.vsoc.ca/pineprices.htm [accessed 2011-01-15]

9

Chapter 2: Host colonization patterns of Leptoglossus occidentalis into seed orchards
after overwintering

2.1 Abstract
The temporal variability and host colonization pattern of a pest population can
influence population estimates, and hence management practices. The movement of
Leptoglossus occidentalis into seed orchards from overwintering sites is an important stage
of colonization as seeds in immature cones are highly susceptible to damage. I investigated
the timing of spring colonization of L. occidentalis in a lodgepole pine orchard in 2008 and
2009 by analyzing the spatial patterns using both logistic regression and spatial point process
models. L. occidentalis was first detected in the seed orchard on 17 May 2008 and 8 May
2009. In 2008, multiple significant, but not directionally consistent, spatial gradients of L.
occidentalis-infested trees were detected on four of seven sampling dates, whereas in 2009 a
spatial gradient was found only on one sampling date. Spatial gradients may have been
easier to detect in 2008 due to a much higher L. occidentalis population than in 2009.
Further study of temporal and spatial trends in early spring, along with a study of the
behaviour of L. occidentalis at varying population densities, is required to improve sampling
methods and our understanding of orchard colonization dynamics of this insect into conifer
seed orchards.

2.2 Introduction
Phytophagous insects and their host plants are seldom distributed randomly across a
homogeneous landscape (Stanton 1983). Agricultural landscapes, including seed orchards,
typically feature high-density, monoculture plantings (Pedigo and Rice 2009), frequently
10

attracting high pest densities that can result in serious economic impacts on crop systems
(Stanton 1983). Understanding the spatial and temporal variability of pest populations and
their relationship to host phenology can help to improve the development of effective pest
population sampling methods and management practices (Stanton 1983). For example, the
Eastern spruce budworm, Choristoneura fiimiferana Clemens (Lepidoptera: Tortricidae), has
somewhat predictable outbreak cycles in eastern Canada (Tothill 1922, Blais 1958,
Greenbank 1963, Royama 1984), and this information has been used in the development and
deployment of management plans (Miller and Kettela 1976, Kettela 1995).
The spatial and temporal variability of population abundance is poorly described for
many species of pest insects, including the western conifer seed bug, Leptoglossus
occidentalis (Hemiptera: Coreidae). Knowledge of an insect’s spatial pattern on a landscape
can have direct management implications particularly in reference to sampling methods. For
example, edges are typically omitted when sampling for pest insects unless it has been
established that dispersal into the system is initiated at an edge, such as for armyworms
(Lepidoptera: Noctuidae), grasshoppers (Orthoptera: Acrididae) (Pedigo and Rice 2009), and
the Douglas-fir twig beetle, Pityophthorus orarius Bright (Coleoptera: Curculionidae)
(Preisler et al. 1997). Conspicuous edges of host-plant patches may acquire a
disproportionate percentage of the insect population and thus incur a higher level of damage
than non-edge (interior) portions of the patch (Stanton 1983, Hunter 2002). Conifer seed
orchards tend to have well-defined and conspicuous edges where cone crops may potentially
accrue heavy damage (Pedigo and Rice 2009). Preliminary observations have indicated that,
in the spring, L. occidentalis is first detected along the edge of the orchard, resulting in a

11

distinct edge effect, which appears to diminish with time as the invasion progresses
(W.B.Strong1, personal communication).
Temporally, the movement of pest insects into monoculture crop systems from
overwintering sites is a particularly important stage of colonization, as young plants can be
highly susceptible to insect damage (Kovaleski et al. 1999, St. Toepfer et al. 1999). Seeds of
lodgepole pine (Pinus contorta Dougl. ex Loud. var. latifolia Engelm.) and white pine {P.
monticola Dougl. ex D. Don) in seed orchards are particularly vulnerable to damage from
feeding L. occidentalis in early spring during the second year of cone development (Connelly
and Schowalter 1991, Bates et al. 2002b, Strong 2006). During early cone development in
conifers, such as ponderosa pine (Krugman and Loerbe 1969), the coat of developing seeds
has not yet hardened, leaving them more vulnerable to damage by overwintered reproductive
females. It is estimated that each female L. occidentalis consumes approximately 5.1 seeds
per day throughout May (recalculated from Strong 2006); consumption decreases to
approximately 1-1.25 seeds per day in June and to approximately 0.25 seeds per day through
August (Strong 2006).
Temperature can also affect the timing and rate of dispersal. Colder winter
temperatures are associated with delayed insect development and increased mortality in
overwintering generations of several species (Jones and Sullivan 1981, Nuvonen et al. 1999,
Pedigo and Rice 2009, Musolin 2012). The relationship between temperature and
development of L. occidentalis is currently unknown; however, it has been observed in other
species of Hemiptera that extreme cold events increase overwintering mortality and/or delay
the arrival of insects on host plants (Jones and Sullivan 1981, Musolin 2012). Establishing
1 Research scientist, Tree Improvement Branch, Ministry of Forest, Lands and Natural
Resource Operations, Kalamalka Forestry Centre, 3401 Reservoir Road, Vemon BC.

12

the timing and patterns of arrival of L. occidentalis into seed orchards in the spring may
facilitate early intervention, and lead to more effective management, potentially reducing
both pesticide use and L. occidentalis damage. For example, if invading individuals stay at
the edge of the orchard before diffusing inwards, targeting pesticide sprays such
(Sevin®XLR is used frequently) on the orchard edge early in the growing season may reduce
subsequent invasions and lower the infestation levels in the orchard especially if the control
substance has a lasting residual effect. Trap crops have been used to prevent the spread of
several pest and/or alien species such as the screw worm, Cochliomyia hominivorax Coquerel
(Diptera: Calliphoridae) (Marsula and Wissel 1994) and the gypsy moth, Lymantria dispar
Linnaeus (Lepidoptera: Lymantriidae) (Leonard and Sharov 1995). The application of
pesticides to trap crops has also been used extensively to protect against human disease
vectors such as mosquitoes (Diptera: Culicidae) and biting midges (Diptera:
Ceratopogonidae) (Lindquist and Mcduffie 1945, Kelly et al. 1997, Royal 2004).
In this study, I considered the timing of spring colonization of I. occidentalis into a
lodgepole pine orchard in 2008 and 2009, and examined temperature records from 2007/2008
and 2008/2009 to see if the arrival dates and the numbers of L. occidentalis could be
explained by differences in winter/spring weather conditions. I used generalized linear
models and spatial point process models (Badderly and Turner 2005) to describe and quantify
the spatial gradient of L. occidentalis infested trees in a lodgepole seed orchard. I used these
observations to determine whether infestations in the orchard occurred on a spatial gradient
in both years, and to look for concentration of infestation on the edges of orchards. Finally, I
assessed if the spatial pattern observed in 2008 was the same as that observed in 2009.

13

2.3 Methods
2.3.1 Study Area and Species
This study was conducted in lodgepole pine seed orchard 307 at the BC Ministry of
Forests, Lands, and Natural Resource Operations Kalamalka Seed Orchards, Vernon, B.C.
(50.24° N , 119.28° W, 489 m above sea level) during the spring and summer of 2008 and
2009. The orchard is rectangular and consists of 1794 tree locations on 3.8 ha, with 1556
trees from 66 different clones remaining as of spring 2009. Orchard 307 spans 117 metres in
length in the north to south direction, and 323 metres in length in the west to east direction.
Scions are grafted onto rootstock of unknown provenance. Each clone was represented by 1
to 43 ramets. The oldest trees were planted in 1985 and the youngest in 2002. Each tree is
spatially referenced with an x, y coordinate corresponding to row and within-row position.
Row-spacing is 7 metres and within-row tree-spacing is 3.5 metres. On average, trees were
approximately 6 metres in height with an average crown width of 4 metres. The orchard is
planted in a modified randomized design with the caveat that a tree must be at least two rows
and four positions away from other ramets of the same clone.

2.3.2 Survey Methods
Orchard surveys were conducted during the summers of 2008 and 2009 to determine
when L. occidentalis arrived in the orchard and its spatial distribution, specifically
quantifying edge effects. In both years, the surveys began after daytime temperatures
exceeded 12 °C. This criterion for a start date was based on unpublished field observations
that L. occidentalis are inactive below 12 °C (W.B. Strong1, personal communication).
Surveys commenced on 5 May 2008 and 8 May 2009, and were conducted eight times in
2008 and ten times in 2009. To compare trends from 2008 and 2009,1paired the data sets by
14

date. Furthermore, as I was primarily interested in describing spatial patterns associated with
host colonization in the orchard in the spring, I only analysed the data from the first seven
corresponding surveys for each year, from 5 May to 20 June in 2008, and 8 May to 18 June
in 2009.
In 2009, the first five rows and the first five trees in all other rows were surveyed in
neighbouring orchards to estimate L. occidentalis levels in those orchards. The orchard
orientation relative to orchard 307, the orchard number, and tree species in the neighbouring
orchards are: north, orchard 305 Engelmann spruce (Picea engelmanni Parry); northeast, 333
western larch (Larix occidentalis Nuttal); east, open field; southeast, 332 western larch; and
south, 341 and 304 Engelmann spruce.
Orchard rows were assigned numbers and scouts randomly selected their starting row
for each monitoring event using a random number table. During the surveys conducted
between 5 May and 17 May 2008 every other tree on every third row was surveyed on foot.
From 23 May 2008 onwards every tree in the orchards was surveyed using Girette
Chipmunk™ orchard lifts (Allied Farm Equipment). This produced a more accurate
assessment of insect numbers because scouts were able to access the entire canopy of the
tree. When surveying on foot scouts were able to access only the bottom two metres of trees.
Scouts drove orchard lifts up and down rows at a speed of approximately 0.2 km per hour.
Each tree was scanned for L. occidentalis from the bottom of the crown to the top of the
crown, with particular attention paid to the cones.
When an insect was sighted, the monitor stopped and captured it for marking (see
below). Clone number, time o f capture, and spatial location were recorded in the field. In

15

2008, insects were brought to the labs for marking and released the following morning.
Insects were not returned to the same tree where they were found the previous day, however
they were all placed upon trees where insects were captured. In 2009, a modified coding
system enabled easy marking in the field so insects were released to cones on the tree where
they were found immediately after marking. The method was changed in 2009 to reduce
insect handling and to increase efficiency in the field.

2.3.3 Trapping and Marking Methods
Live insects were collected using pheromone traps modified by W.B. Strong1
(Unitrap™, Contech Enterprises Inc., Delta, BC). The top portion of the trap was removed
and the entrance and inside of the trap were painted with Insect-a-Slip™, a fluoropolymer
resin (PTFE-30) (BioQuip, Rancho Dominguez, CA). After drying, the polymer forms a
slippery surface that inhibits arthropods from gaining traction on the treated area. When
located, usually on a cone, L. occidentalis were captured by placing the modified trap
underneath the insect and tapping the top of the cone. When disturbed, L. occidentalis drop
before taking flight, causing them to fall into the funnel entrance o f the trap where they
couldn’t gain enough traction to escape. Once the insect entered the trap, a square piece of
black insect screening was used to cover the entrance and prevent escape. Duringsurveys,
insects were marked to ensure individuals were not counted more them once.
In the 2008 mark-recapture studies, I used a marking method developed by W.B.
Strong1 and S. Desjardins2 in their 2006 and 2007 field studies (unpublished data). This

2 Professor Mathematics, Irving K. Barber School of Arts and Sciences, University of British
Columbia Okanagan, Kelowna BC
16

method involved capturing insects and bringing them to the lab where they were individually
marked by painting the pronotum with Liquitex™ (Liquitex Artist Materials, Piscataway,
New Jersey) professional acrylic artist water-based colour paint. After allowing sufficient
time for the paint to dry, an alphanumeric code was inscribed on the paint-mark with a Pigma
Micron™ (Sakura of America, Hayward, California) archival ink pen. The combination of
colour and alphanumeric code was used to identify the orchard in which the insect was
caught and the capture/release date, as well as the individual insect. In 2009,1 developed a
method for marking L. occidentalis in the field. I replaced the use of Liquitex™ paint with
quick drying SRX Metallic Colorsharp™ permanent marker pens (MEGA Brands Inc.,
Montreal, Quebec). I assigned a unique number to individual insects, written on the
pronotum using a Pigma Micron™ pen, and I used a system of coloured dots, applied with
SRX Metallic Colorsharp™ permanent marker pens on the forewings, to code the date of
marking. Both marking methods were tested on caged individuals in the laboratory.
Observations were recorded for both marked and unmarked individuals; no differences in
behaviour, fecundity, and longevity were found (W.B. Strong1, B. Lalonde3(unpublished
data) and personal observation).

2.3.4 Statistical analyses
For each data set, I fit a logistic regression model with a binary response variable
(presence/absence) to model the spatial trend of infestation in the x (west to east) andy (north
to south) directions (Crawley 2007). Because plots of trees harbouring insects appeared to be
clustered in the centre of the orchard in our exploratory data analysis, I also included a
3 Professor of Entomology, Irving K. Barber School of Arts and Sciences, University of
British Columbia Okanagan, Kelowna BC
17

polynomial logistic regression to model the spatial trend in x using the equation y=flo+ft2X
+P2X ■ I also plotted the back-transformed regression equation to determine the probability
of finding a tree with an insect along a particular directional gradient whenever a model
exhibited a significant directional trend and the best model for each survey date was selected
using the lowest Akaike’s Information Criteria (AIC) value. The AIC value is a measure of
model fit: the smaller the AIC value associated with a model the better the fit (Crawley
2007). I also used spatial point process models to model the number of insects per metre
squared over the entire orchard, and compared each with the null model assuming no spatial
trend (Badderly and Turner 2005). AIC values and Chi-square tests were used to select the
best fitting model for each survey date. These models were used because they generate
density plots that enabled easy visualisation of the spatial pattern. I visually examined
models with corresponding dates from 2008 and 2009 to assess if the same trends were
observed from one year to the next. Statistical analyses were done using R statistical
software (R Development Core Team 2011). The R package spatstat (Badderly and Turner
2005) was used for spatial point process models and density plots.

18

2.4 Results
2.4.1 Winter and spring temperatures, timing o f spring immigration and the number ofh.
occidentalis found during surveys
Weather data were obtained from the Environment Canada website (Environment
Canada 2012). Average monthly winter temperatures were lower in December (2008) and
February 2009, than in December and February of the previous year. There were slightly
lower average monthly minimum winter temperatures in December (2008), February and
March of 2009 than in the corresponding months in 2008. Average monthly spring
temperatures were similar in both years, but an unusually low minimum temperature event (17.8°C) occurred in March 2009, compared with a miniumum temperature of -4.8°C in
March of 2008 (Environment Canada 2012) (Table 2.1 and 2.2).

19

Mean maximum
temperature (°C)

Mean minimum
temperature (°C)

Mean average
temperature (°C)

Month

2008

2009

2008

2009

2008

2009

December

-0.4

-4.2

-6.1

I— *

I
o
'o

Table 2.1 Monthly mean maximum, minimum, and average temperatures from December
through May 2007/2008 and 2008/2009.

-3.3

-7.6

January

2.0

-1.7

-10.1

-7.5

-6.0

-4.6

February

4.7

1.4

-4.0

-7.9

0.3

-3.3

March

9.0

5.8

-1.2

-4.5

3.9

0.6

April

12.3

14.2

0.5

1.2

6.4

7.7

May

21.1

21.3

7.3

6.1

14.2

13.7

Table 2.2 Extreme monthly maximum and minimum temperatures from December through
May 2007/2008 and 2008/2009.
Extreme minimum
temperature (°C)

Extreme maximum
temperature (°C)
Month

2008

2009

2008

2009

December

n/a

5.0

n/a

-27.8

January

6.4

5.6

-17.7

-21.0

February

11.0

6.7

-12.6

-16.8

March

14.0

14.0

-4.8

-17.8

April

20.7

n/a

-6.0

n/a

May

30.1

32.9

0.8

1.4

20

2.4.2 Spatial gradient o f L. occidentalis in lodgepole pine orchard 307

Spatial point process models can be used to examine trends and “hotspots” of insect
abundance per unit area through time (see Figures 2.1 and 2.2). In 2008, no insects were
found during the first survey. In the second survey, a single insect was captured on the
southern edge of the orchard. Throughout the remainder of May, insects appeared to spread
along the southern edge, and by 28 May, they were distributed throughout the orchard. The
population increased from the end of May through 20 June as L. occidentalis spread
throughout the orchard with hotspots observed near the eastern edge of the orchard. In 2009,
L. occidentalis were found in the centre of the orchard during the first survey. Throughout
May, hot spots were distributed throughout the orchard. At the beginning of June, L.
occidentalis were most abundant on the eastern side of the orchard, and by the final survey
on 18 June they were concentrated in the centre of the orchard.

21

FI I
0

50

a) 5 May 2008

100

150

100

150

200

250

300

360

300

350

300

350

b) 8 May 2009

200

250

300

3S0

50

c) 17 May 2008

100

150

200

250

d) 15 May 2009

8-

0

50

150

200

250

300

350

50

e) 23 May 2008

100

150

200

250

f) 23 May 2009

s
s
o

0

SO

100

150

g) 28 May 2008

200

250

300

350

0

50

100

150

200

h) 28 May 2009

Figure 2.1 a-h) Density plots of L. occidentalis infested trees in orchard 307 at the Kalamalka
Seed Orchard in Vernon B.C., Canada from a) 5 May 2008, b) 8 May 2009, c) 17 May 2008,
d) 15 May 2009, e) 23 May 2008, f) 23 May 2009, g) 28 May 2008, h) 28 May 2009. A
significant spatial gradient in L. occidentalis infested trees was observed in the 23 May 2008
survey. The plot axes are to scale with orchard 307 with the point 0,0 being the southwest
comer of orchard 307. Note that the scale on the right-hand side of each plot (density of
insect-infested trees per m2) is different for each plot.

22

0

SO

100

ISO

200

250

300

350

0

a) 4 June 2008

50

KM

150

200

250

300

0

50

5C

100

150

200

e) 20 June 2008

150

20C

280

300

350

£2
100

t

1

r“

150

200

250

300

J50

d) 13 June 2009

c) 10 June 2008

0

100

b) 2 June 2009

I
kki

0

56

250

300

350

0

50

I

100

150

200

250

300

350

f) 18 June 2009

Figure 2.2 a-f) Density plots of L. occidentalis infested trees in orchard 307 at the Kalamalka
Seed Orchard in Vernon B.C., Canada from a) 4 June 2008, b) 2 June 2009, c) 10 June 2008,
d) 13 June 2009, e) 20 June 2008 and f) 18 June 2009. A significant spatial gradient in L.
occidentalis infested trees was observed in all three June 2008 surveys, and in the 2 June
2009 survey. The plot axes are to scale with orchard 307 with the point 0, 0 being the
southwest comer of orchard 307. Note that the scale on the right-hand side of each plot
(density of insect-infested trees per m2) is different for each plot.

23

Table 2.3 The number of L. occidentalis caught and marked during 2008 and 2009 surveys.
Survey date
2008

Number of L.
occidentalis caught and
marked 2008

Survey date
2009

Number of L.
occidentalis caught and
marked 2009

5 May

0

8 May

5

17 May

1

15 May

4

23 May

9

23 May

11

28 May

74

28 May

21

4 June

52

2 June

7

10 June

132

13 June

22

20 June

325

18 June

38

24

I used logistic regression to quantify the trends depicted in Figure 2.1 and 2.2. A
significant spatial gradient was detected in four of the surveys in 2008, and two surveys in
2009 (Table 2.4 and 2.5). L. occidentalis did not appear to invade seed orchards in a
consistent manner. As the population increased, areas of high concentration moved around
within the orchard. In 2008, a gradient of L. occidentalis infested trees appeared in the north
to south direction within two weeks of sighting the first L. occidentalis in the field that year.
In the following weeks however, the gradients fluctuated from west to east and, by the end of
the study, most of the insects were concentrated near the centre of the orchard. Overall, the
probability of detection, obtained from the plotting the back-transformed regression equation,
remained low, with a jump to 40% near the end of June. In 2009, there was only one
statistically significant gradient, which appeared on 2 June and occurred from west to east.
The probability of finding a tree with an insect along that particular directional gradient was
25%. The statistical analysis can be found in Appendix I.

25

Table 2.4. Direction of statistically significant spatial gradient and the probability of
encountering L. occidentalis infested trees in the 2008 surveys. Spatial gradient is measured
in metres along the x (323 metres) andy (117 metres) axes with the point (0,0) representing
the southwest comer of the orchard.

No
Gradient

Gradient

Survey
date

/’(detection)

Typea

Min.
Where
/’(detection)1’ (m)c

5 May

0%

n/a

n/a

n/a

n/a

n/a

17 May

0.06%

n/a

n/a

n/a

n/a

n/a

23 May

n/a

Linear,
south to
north

0%

117

8%

0

28 May

3.8%

n/a

n/a

n/a

n/a

n/a

4 June

n/a

Linear,
west to
east

3%

0

10%

323

10 June

n/a

Quadratic 3%
(y=*-*2),
west to
east

0

17%

197

20 June

n/a

Quadratic 5%
(y=x-x2),
west to
east

0

40%

211

Max.
Where
/’(detection)1’ (m)c

a ‘Type’ refers to the type of regression equation used to model the spatial gradient and the
direction of the spatial gradient, e.g. south to north, west to east.
b ‘P(detection)’ means the probability of finding L. occidentalis in a tree.
c ‘Where’ indicates the location in the orchard where the maximum or minimum probability
of finding L. occidentalis in a tree occurs; the southern edge of the orchard is at 0m, the
northern at 117m. The western edge of the orchard is at 0m and the eastern at 323m, e.g.
linear, north-south gradient at 117m means that the highest density of infested trees is found
at the northern edge of the orchard.
26

Table 2.5. Direction of statistically significant spatial gradient and the probability of
encountering L. occidentalis infested trees in 2009 surveys. Spatial gradient is measured in
metres along the x (323 metres) andy (117 metres) axes with the point (0,0) representing the
southwest comer of the orchard.

No
Gradient

Gradient

Survey
date

/ ’(detection)

Type3

Min.
Where
P(detection)b (m)°

Max.
Where
/ ’(detection)1’ (m)c

8 May

0.2%

n/a

n/a

n/a

n/a

n/a

15 May

0.7%

n/a

n/a

n/a

n/a

n/a

23 May

0.6%

n/a

n/a

n/a

n/a

n/a

28 May

1%

n/a

n/a

n/a

n/a

n/a

2 June

n/a

Linear,
west to
east

0%

0

25%

323

13 June

1%

n/a

n/a

n/a

n/a

n/a

18 June

2%

n/a

n/a

n/a

n/a

n/a

a ‘Type’ refers to the type of regression equation used to model the spatial gradient and the
direction of the spatial gradient, e.g. south to north, west to east.
b ‘P(detection)’ means the probability of finding L. occidentalis in a tree.
c ‘Where’ indicates the location in the orchard where the maximum or minimum probability
of finding L. occidentalis in a tree occurs; the southern edge of the orchard is at 0m, the
northern at 117m. The western edge of the orchard is at 0m and the eastern at 323m, e.g.
linear, north-south gradient at 117m means that the highest density of infested trees is found
at the northern edge of the orchard.
27

2.5 Discussion
Edges are typically omitted when sampling for pest insects in agricultural systems
unless it has been established that they immigrate into the system at an edge, e.g.,
armyworms (Lepidoptera: Noctuidae) and grasshoppers tend to first appear in crops on the
edges of a field (Orthoptera: Acrididae)(Pedigo and Rice 2009). Despite the prediction that
an edge effect may be present (W.B. Strong1, personal communication), my observation did
not show that L. occidentalis consistently colonized orchard 307 starting at the southern edge
in 2008 and 2009. Although significant spatial gradient of L. occcidentalis was observed on
the southern edge of orchard 307 in the 23 May survey 2008 season, none was detected in
corresponding survey in 2009. However, these results are inconclusive since low population
density in 2009 may have affected our ability to detect presence of L. occidentalis on the
edge of the orchard early in the season.
While my results are inconclusive, it is premature to conclude that there are no edge
effects present in the invasion patterns of L. occcidentalis. Transient early season edge effects
of pest insects have been observed, for example, in two species of thrips ( Thysanura spp.) in
high bush blueberries Vaccinum corymbosum Linnaeus in New Jersey, where plants located
on edges closer to forested areas had more insects in the early part of the growing season
(Rodriguez-Saona et al. 2010).
In addition, even though L. occidentalis are strong fliers, it is possible that the
direction of prevailing winds in spring may play a role in determining where L. occidentalis
invade seed orchards following emergence from overwintering sites. The Douglas-fir beetle,
Pityophthorus orarius Bright (Coleoptera: Curculionidae), provides a possible example of

28

this as wind direction may influence the spatial patterns of attacks on Douglas-fir in seed
orchards (Priesler et al. 1997). Additional research should therefore include monitoring wind
prevalence to assess the potential effects of prevailing winds.
My observations suggest that population density, spatial distribution, and behaviour
may be linked in L. occidentalis. The differences in spatial patterns detected in the population
of L. occidentalis in 2008 and 2009 may in part be due to the very different population
densities observed between years. Dispersion and population density are often interrelated in
several insects species and deviations from randomness in spatial patterns can be extremely
difficult to detect at low population densities (Taylor et al. 1978). Since the design of a
suitable sampling protocol for an insect pest requires information on the density and the
spatial patterns the insect occupies within its host crop system (Pedigo and Rice 2009),
further study of temporal and spatial trends in early spring, along with an examination of
potential differences in behaviour by L. occidentalis at different population densities, is
required to help improve sampling methods.
The influence of temperature on the timing of overwintering emergence has been
well documented for many insect and mite species across several taxa (Gangavalli and
Alinazee 1985, Bergh and Judd 1993, Zeiss et al. 1996). Although there is currently no
degree-day model for the development and phenology of L. occidentalis, the developmental
thresholds of other coreid and lygaeid species are reported to be between 8-12 °C (Champlain
and Butler 1967, Steinbauer 1997). Field observations also indicate that L. occidentalis are
not active below 12° C (W. B. Strong1, personal communication). The slightly cooler average
temperature observed in April 2008 may therefore have influenced the phenology of L.
occidentalis, potentially resulting in later emergence from overwintering sites in 2008 than in

29

2009. Another possible explanation for the dearth of L. occidentalis observed in the earliest
spring surveys could also be the fact that the first few surveys in 2008 were conducted on
foot. During walk-through surveys, scouts scanned only the bottom half of most trees as the
upper crown was not completely visible from the ground. Trees were, on average, six or
more metres tall and the top five whorls of a lodgepole pine tree tend to bear greater numbers
of seed cones (O’Reilly and Owens 1988). Additionally, L. occidentalis may have been
present in higher numbers than indicated through the surveys, as they are fairly cryptic on
their host (Hedlin et al. 1981, personal observation).
Although I am not currently able to recommend targeting the southern edge of
orchard 307 for early season sprays, the presence of a spatial gradient in spring 2008 and
previous field observations (W.B. Strong1, personal communication) indicate that this may
warrant further investigation. Additional surveys are required to determine if the spatial
trends detected in 2008 are observed in other years. If a consistent edge effect is detected,
trials of early season edge sprays should be conducted to test if spring edge sprays have the
potential to reduce seed damage as well as the economic and environmental cost of spraying
the entire orchard.

30

2.6 Literature Cited

Badderly, A., and Turner, R. 2005. Spatstat: An R package for analyzing spatial point
patterns. Journal of Statistical Software. 12: 1-42.
Bates, S.L., Borden, J.H., Kermode, A.R., and Bennett, R.G. 2000. Impact of L. occidentalis
(Hemiptera: Coreidae) on Douglas-fir seed production. Journal of Economic
Entomology. 93: 1444-1451.
Bates, S.L., Strong, W.B., and Borden, J.H. 2002b. Abortion and seed set in lodgepole and
western white pine conelets following feeding by L. occidentalis (Heteroptera:
Coreidae). Environmental Entomology. 31: 1023-1029.
Bergh, J. C., and Judd, G.J.R. 1993. Degree-day model for predicting emergence of pear rust
mite (Acari: Eriophyidae) deutogynes from overwintering sites. Environmental
Entomology. 22: 1325-1332.
Blais, J. R. 1958. Effects of defoliation by spruce budworm on radial growth at breast height
of balsam fir and white spruce. Forestry Chronicle. 34: 39—47.
Champlain, R.A., and Butler, G.D. 1967. Temperature effects on development of the egg and
nymphal stages oiLygus hesperus (Hemiptera: Miridae). Annals of the
Entomological Society of America. 60: 519-521.
Connelly, A.E., and Showalter, T.D. 1991. Seed losses to feeding by L. occidentalis
(Heteroptera: Coreidae) during second year cone development in western white pine.
Journal of Economic Entomology. 84: 215-217.
Crawley, M.J. 2007. The R Book. John Wiley and Sons, West Sussex, England.
Environment Canada.(n.d.). National Climate Data and Information Archive. Accessed
online 3 December 2012, from
http://www.climate.weatheroffice.gc.ca/climateData/hourlvdata e.html?timeframe= 1
&Prov=BC&StationID=46987&hlvRange:=2007-11-3012012-1202&Year=2012&Month=l 2&Dav=2
Ferguson, A.W., Klukowski, Z., Walczak, B., Clark, S.J., Mugglestone, M.A., Perry, J.N.,
and Williams, I.H. 2003. Spatial distribution of pest insects in oilseed rape:
implications for integrated pest management. Agriculture, Ecosystems &
Environment. 95: 509-521.
Gangavalli, R.R., and Alinazee, M.T. 1985. Temperature requirements for development of
the obliquebanded leafroller, Choristoneura rosaceana (Lepidoptera: Tortricidae).
Environmental Entomology. 14: 17-19.

31

Greenbank, D.O. 1963. The development of the outbreak. Memoirs of the Entomological
Society o f Canada, 95, pp. 19-23.
Hedlin, A.F., Yates III, H.O., Tovar, D.C., Ebel, B.H., Koerber, T.W., and Merkel, E.P.
1981. Cone and seed insects of North American conifers. Canadian Forestry Service,
US Department of Agriculture, Forest Service and Secretaria de Agricultura y
Recursos Hidraulicos, Mexico.
Hunter, M.D. 2002. Landscape structure, habitat fragmentation, and the ecology of insects.
Agricultural and Forest Entomology. 4: 159-166.

Jones, W. A., and Sullivan, M. J. 1981. Overwintering habitats, spring emergence patterns,
and winter mortality of some South Carolina Hemiptera. Environmental Entomology.
10: 409-414.
Kettela, E. G. 1995. Insect control in New Brunswick, 1974-1989. Pages 655-665 in J. A.
Armstrong and W. G. H. Ives, editors. Forest insect pests in Canada. Natural
Resources Canada, Canadian Forest Service, Ottawa, Canada.
Kovaleski, A., Sugayama, R.L., and Malavasi, A. 1999. Movement of Anastrepha fraterculus
from native breeding sites into apple orchards in Southern Brazil. Entomologia
Experimentalis et Applicata. 91: 457-463.
Kelly, D.W., Mustafa, Z., and Dye, C. 1997. Differential application of lambda cyhalothrin to
control the sandfly Lutzomyia longipalpis. Medical Veterinary Entomology. 11: 13—
24.
Krebs, C.J. 1999. Ecological Methodology. Addison Wesley Longman, Menlo Park, CA.
Krugman, S.L., and Koemer, T.W. 1969. The effect of cone feeding by L. occidentalis on
ponderosa pine seed cone development. Forest Science. 15: 104-111.
Leonard, D.S., and Sharov, A.A. 1994. Slow the spread project update: developing a process
for evaluation. Pages 82-85 in Proceedings of the USDA Interagency Gypsy Moth
Research Forum. US Forest Service General Technical Report NE-213.
Lindquist, A.W., and McDuffie, W.C. 1945. DDT residual spray tests in Panama. Journal of
Economic Entomology. 38: 608.
Marsula, R., and Wissel, C. 1994. Insect control by a spatial barrier. Ecological Modeling.
75/76: 203-211.
Mellanby, K. 1939. Low temperature and insect activity. Proceedings of the Royal Society of
London, Series B, Biological Sciences. 127: 473-487

32

Miller, C.A., and Kettela, E.G. 1975. Aerial control operations against the spruce budworm
in New Brunswick, 1952-1973. Pages 94-112 in M. L. Prebble, editor. Aerial control
o f forest insects in Canada. Information Canada, Ottawa, Canada.
Morris, R.F. 1955. The development of sampling techniques for forest insect defoliators with
particular reference to the spruce budworm. Canadian Journal of Zoology. 33: 225294.
Musolin, D. L. 2012. Surviving winter: diapause syndrome in the southern green stink bug
Nezara viridula in the laboratory, in the field, and under climate change conditions.
Physiological Entomology. 37: 309-322
Neuvonen S., Niemela, P., and Virtanen, T. 1999. Climatic change and insect outbreaks in
boreal forests: the role of winter temperatures. Ecological Bulletins. 63-67.
Owens, J.N. 2006. The reproductive biology of lodgepole pine. Forest Genetics Council of
British Columbia. Victoria, BC. ISBN 0-7726-5342-9
O'Reilly C., and Owens, J. N. 1988. Reproductive growth and development in seven
provenances of lodgepole pine. Canadian Journal of Forest Research. 18: 43-53.
Pedigo, L.P., and Rice, M.E. 2009. Entomology and Pest Management. Prentice-Hall, Upper
Saddle River, NJ.
Phillips, W.M. 2008. Effects of leaf age on feeding ‘preference’ and egg laying in the
chrysomelid beetle, Haltica lythri. Physiological Entomology. 1: 223-226.
Preisler, H.K., Rappaport, N.G., and Wood, D.L. 1997. Regression methods for spatially
correlated data: an example using beetle attacks in a seed orchard. Forest Science. 43:
71-77.
R Development Core Team. 2011. R: A language and environment for statistical computing.
R Foundation for Statistical Computing, Vienna, Austria. ISBN3-900051-07-0, URL
http ://www .R-proj ect. org.
Royal, A. 2004. A new tool for the control of mosquitoes, biting midges, and flies. Wing
Beats. 15: 18-19,22.
Royama, T. 1984. Population dynamics of the spruce budworm Choristoneura fumiferana.
Ecological Monographs. 54: 429-462.
Rodriguez-Saona, C. R., Polavarapu, S., Barry, J. D., Polk, D., Jomsten, R., Oudemans, P.
V., and Liburd, O. E. 2010. Color preference, seasonality, spatial distribution and
species composition of thrips (Thysanoptera: Thripidae) in northern highbush
blueberries. Crop Protection. 29: 1331-1340.

33

Srivastava, N., Campbell, R.W., Torgersen, T.R., and Beckwith, R.C. 1984. Sampling the
western spruce budworm: Fourth instars, pupae and egg masses. Forest Science. 30:
883-892.
Stanton, M.L. 1983. Spatial patterns in the plant community and their effect on insect search.
Pages 125-157 in S. Ahmad, editor. Herbivourous insects: host-seeking behaviour
and mechanisms. Academic Press, Orlando, FL.
Steinbauer, M.J. 1997. Seasonal phenology and developmental biology of Amorbus
obscuricornis (Westwood) and Gelonus tasmanicus Le Guillou (Hemiptera :
Coreidae). Australian Journal of Zoology. 45: 49-63.
Stilwell, A.R., Wright, R.J., Hunt, T.E., and Blankenship, E.E. 2010. Degree-day
requirements for alfalfa weevil (Coleoptera: Curculionidae) development in eastern
Nebraska. Environmental Entomology. 39: 202-209.
Strong, W.B. 2006. Seasonal changes in seed reduction in lodgepole pine cones caused by
feeding of L. occidentalis (Hemiptera:Coreidae). The Canadian Entomologist. 138:
888-896.
St. Toepfer, G.H., and Dorn, S. 1999. Spring colonisation of orchards by Anthonomus
pomorum from adjacent forest borders. Entomologia Experimentalis et Applicata. 93:
131-139.
Taylor, L.R. 1963. Analysis of the effect of temperature on insects in flight. Journal of
Animal Ecology. 32: 99-117.
Taylor, L.R. 1978. The density-dependence of spatial behaviour and the rarity of
randomness. The Journal of Animal Ecology. 47: 383-406
Tothill, J.D. 1922. Notes on the outbreaks of spruce budworm, forest tent caterpillar, and
larch sawfly in New Brunswick. Proceedings of the Acadian Entomological Society.
8: 172-182.
Torgersen, T.R., Colbert, J.J., and Hosman, K.P. 1993. Patterns of occurrence and new
sampling implications for instar IV western spruce budworm (Lepidoptera:
Tortrieidae). Forest Science. 39: 573-593.
Zeiss, M.R., Koehler, K.J., and Pedigo, L.P. 1996. Degree-day requirements for development
of the bean leaf beetle (Coleoptera: Chrysomelidae) under two rearing regimes.
Journal of Economic Entomology. 89: 111-118.

34

Chapter 3. Cues mediating host selection of Leptoglossus occidentalis on lodgepole pine

3.1 Abstract
The seed predator Leptoglossus occidentalis Heidemann, can cause significant
damage in conifer seed orchards. Host selection is not completely understood, but earlier
research in seed orchards has shown preference for certain clones of lodgepole pine, Pinus
contorta Douglas ex Louden var. latifolia, indicating that insects may respond to chemical or
physical cues in host selection. I tested whether L. occidentalis shows clonal preference in a
consistent manner between 2008 and 2009, and examined whether host cues responsible for
such preference could be identified. Surveys were conducted in a lodgepole pine seed
orchard in British Columbia. I established three classes of clone preference based on
presence or absence of L. occidentalis. Infrared radiation (IR) emitted from cones, cone
terpenes, cone size, cone count, cone weight, and seed counts were measured. I established
that clone preference remained consistent between 2008 and 2009, and that clone preference
classes had significantly different levels of the monoterpenes a-pinene, 8-3 carene, and pcymene. Leptoglossus occidentalis was found more frequently on clones with cones of
greater diameter and weight. However, level of IR radiation, which has recently been shown
to be significant in host orientation, did not differ between clone preference classes.

3.2 Introduction
Phytophagus insects often respond to phenotypic variation in their host plants. This
variation can affect the susceptibility or suitability of an individual plant for insect
herbivores, potentially having dramatically different effects on their fitness and that of their
offspring (Fritz and Simmons 1992, Lara and Tanzini 1997, Cronin et al. 2001). If there is a
35

positive relationship between oviposition preference and offspring fitness, herbivores should
select the most beneficial plant phenotypes (Thompson 1988, Joshi and Thompson 1995,
Mopper 1996). However, insects do not always choose the most beneficial host, and some
host-plant genotypes may produce compounds that act as deterrents to insects but have no
deleterious effects on adult or offspring fitness (Courtney and Kibota 1990). Insects may
also switch between alternative hosts. For example, an insect may prefer a particular host,
accidentally land on another suitable host, and then seek out the second host as well as the
initially preferred host (Courtney and Kibota 1990). A possible constraint limiting herbivore
adaptation to a specific host plant is phenotypic fluctuations due to host responses to
environmental factors, which may result in insects favouring a more generalized diet
(Thompson 1988, Joshi and Thompson 1995, Mopper 1996).
Specific heritable plant characteristics eliciting responses from insects include visual
cues such as host plant architecture (Grevstad and Klepetka 1992), foliage and/or flower
colour (Roques 1987), size of food resources (Rappaport and Roques 1991), and olfactory
and gustatory cues provided by host volatiles (Miller and Borden 1990). Insects can use
these cues to differentiate between host plants of different species or to assess differences in
host quality within a single host-plant species (Courteny and Kibota 1990).
Visual cues, for example size, colour, shape, and reflectance of plants play an
important role in host selection and identification for some insects (Judd and Borden 1991,
Turgeon et al. 1994). Insect response to these cues can be general. For example, the
butterfly Pieris virginiensis W.H. Edwards (Lepidoptera: Pieridae) is attracted to basic
perpendicular forms (Cappuccino and Kareiva 1985), and the weevil Hylobius warreni Wood
(Coleoptera: Curculionidae) responds to tree silhouettes (Machial et al. 2012). Alternatively,
insect responses can be highly specific. For example, the butterfly Battus philenor Linnaeus
36

(Lepidoptera: Papilionidae) can differentiate between conspecific hosts of high and low
quality by using leaf bud size as a cue (Papaj and Rausher 1987). Infrared cues have also
been shown to play a role in host selection and reproduction. Females of the pyrophilic beetle
Melanophila acuminata DeGeer (Coleoptera: Buprestidae) use infrared cues to detect and
oviposit into fresh fire-killed wood enabling their larvae to bypass defence mechanisms of
live trees (Evans 1966, Evans 2010). Infrared is also used as a host navigation cue for the
hematophilic bug Rhodnius proxilus Stal (Hemiptera: Reduviidae) (Schmitz et al. 2000).
Takacs et al. (2009) recently discovered that the western conifer seed bug, Leptoglossus
occidentalis Heidemann (Hemiptera: Coreidae), orients to experimental infrared cues with
the aid of infrared receptors located ventrally on the abdomen. But it remains unclear whether
these insects can use infrared cues to differentiate between conspecific hosts.
The olfactory cues used by insects may elicit responses at long- or short-range.
Attractive volatile chemicals in air currents can stimulate receptive insects to fly upwind
towards the source, or host plant search behaviour could be random with chemical attractants
having a more influential role in host acceptance following initial host landing or selection
(Finch 1978, Turgeon et al. 1994). Monoterpenes are a group of volatile chemicals involved
in pheromone and defensive compound production, feeding and oviposition site selection,
and host location and acceptance in numerous phytophagous insect species. Insect responses
to host monoterpenes can be highly variable (Turgeon et al. 1994), and a single terpenoid can
produce a different response in different insect species. For example, 8-3-carene in carrots
acts as a defensive compound against Trioza apicalis Foerster (Homoptera: Psylloidea)
(Nehlin et al. 1994), whereas it acts as an attractive compound for Hylobius abietis Linnaeus
(Coleoptera: Curculionidae) (Visser 1986). Considerable variation in both constitutive and
induced defensive monoterpene compounds has been found in different family groups of
37

lodgepole pine, and several of these compounds are associated with tree defense against
attacks from the mountain pine beetle Dendroctonus ponderosae Hopkins (Coleoptera:
Curculionidae) (Yanchuk et al. 2008, Clark et al. 2010, Ott et al. 2011).
Clonal differences in the susceptibility and/or resistance to damage from insects in
seed orchard environments has been recorded in several conifers including slash pine,
lodgepole pine, Douglas-fir, and Sitka spruce (DeBarr et al. 1972, Blatt and Borden 1999,
Bains et al. 2009). Blatt and Borden (1999) found that L. occidentalis exhibited pronounced
clonal preferences in lodgepole pine (Pinus contorta Dougl. var. latifolia Engelm.)
(Pinaceae) and Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) (Pinaceae) seed orchards,
and that preference was relatively consistent from one year to the next. They also found that
L. occidentalis preferred trees of intermediate height with moderate cone crops, and that
cones from preferred clones reflected more light over a wider range of wavelengths than
those from non-preferred clones (Blatt and Borden 1999). The infrared differential between
cones and foliage in conifers is likely used as a long-range selection cue for L. occidentalis,
however the short-range influence of infrared cues is as of yet unclear (Tackacs et al. 2009).
Cones have significantly stronger infrared radiation than foliage, and this differential may
also provide a strong short-range navigation cue for L. occidentalis searching for feeding and
oviposition sites (Takacs et al. 2009). The role of chemical cues such as monoterpenes in
host attraction and host selection of L. occidentalis has not yet been explored.
The objectives of this study were to: 1) identify and rank the clones of lodgepole
pine favoured by L. occidentalis in 2008 and 2009 at the Kalamalka Seed Orchard, Vernon,
B.C. according to occupancy; 2) test Blatt and Borden’s (1996) finding that L. occidentalis
clonal preference on lodgepole pine is consistent from year to year, and; 3) assess clonal
differences among several candidate host-selection cues identified from the literature.
38

3.3 Methods
3.3.1 Study Area and Species
The study was done in lodgepole pine seed orchard 307 at the BC Ministry of Forests,
Lands, and Natural Resource Operations Kalamalka Seed Orchards, Vernon, B.C.
(50.238512° N , 119.276653° W, 489m above sea level) during the springs and summers of
2008 and 2009. The orchard is rectangular and consists of 1794 tree locations on 4.4 ha, with
1556 trees from 66 different clones as of spring 2009. Each clone was represented by 1 to 43
ramets. The oldest trees were planted in 1985 and the youngest in 2002. Each tree is
spatially referenced with an x, y coordinate corresponding to row and within-row position.
Row-spacing is 7 metres and within-row tree-spacing is 3.5 metres. On average, trees were
approximately 6 metres in height with an average crown width of 4 metres. The orchard is
planted in a randomized design with the caveat that a tree must be at least two rows and four
positions away from other ramets of the same clone.

3.3.2 Insect Monitoring
In 2008, monitoring for L. occidentalis commenced on 8 May and continued until
clones were classified on 8 July. In 2009, monitoring for insect presence was done 5 May to
11 July. Insect monitoring could not be done between the penultimate and final sampling
events in 2009 due to logistical constraints. As insects were encountered and captured (see
Chapter 2 for trapping methods), clone identity, spatial coordinates, and the number of L.
occidentalis found on each tree were recorded. Each adult insect was given a unique mark

39

when first captured to ensure they were not counted more than once (see Chapter 2 for
marking methods).

3.3.3 Clone Ranking and Clone Preference Classification
Favoured clones were identified by monitoring every tree in orchard 307 for the
presence of L. occidentalis adults or nymphs. Seven monitoring events were used in 2008
and ten in 2009. Preferred clones were identified in orchard 307 by ranking the proportion of
ramets harbouring insects in the 66 clones from highest to lowest. Individual ramets of
clones were then ranked according to the total number of insects found on them. Clones with
fewer than 20 ramets were excluded from the study as the lack of insects on these clones
could have been a function of clone scarcity, rather than physical and chemical attributes.
The rank of clones in 2008 and 2009 were recorded for comparison between years. I defined
three clone preference classes: favoured occupied (highest occurrence of L. occidentalis),
favoured unoccupied (with the same clone/genetic identities as the favoured occupied
clones), and non-favoured (absence or very low numbers of L. occidentalis). In both 2008
and 2009,1 selected three low-ranking clones and three high-ranking clones for further study.

3.3.4 Host-Selection Cues
Six experimental blocks were established in orchard 307 in 2008 and five in 2009. In
each block all ramets of the clones of interest were given a number, and three ramets of each
preference class were selected in each block using tables of random numbers. Between two
and five cones per ramet of the clones selected for study were chosen on 16 July 2008 to
measure potential host-selection cues. In 2009, clones were classified and ramets selected on

40

24 June. This was the earliest date insect preference could be determined. On each ramet
selected for study, three cone clusters bearing at least four cones and, when possible, evenly
distributed in the crown, were selected, labelled, and marked with flagging tape.
A list of host-selection cues for L. occidentalis reported in the literature was
established, and from the list the following cues were selected for further investigation: cone
infrared emittance (Takacs et al. 2009); cone monoterpenes (Dicke 2000); cone number
(measured by the number of cones per tree); cone size; and cone weight (Blatt and Borden
1999).

3.3.4.1 Infrared Radiation Measurements
Between two and five cones from 18 trees of the selected clones in each of the three
preference classes were photographed for infrared reflectance measurements on 17 July
2008. Four non-favoured trees were removed from the 2008 data set as the clone identities
were unclear. On each of the three sample dates in 2009 (26-28 June, 15-17 July, and 16-18
August), a single cone per cluster was photographed for infrared and for reference
photographs. In total, three cones per tree on 45 trees (15 trees per cone preference category)
were thus sampled.
All infrared photographs were taken with an Agema Thermovison 550 camera (FLIR
Systems Ltd., Burlington, ON, Canada), which photographs infrared in the 3-5 pm range.
Reference photographs were taken with a Nikon D90 camera. The photographs were taken
by StephenTakacs1 in 2008, and by Tracy Zaradnik2 in 2009. All photographs were taken

1Department of Biological Sciences, Simon Fraser University, Burnaby, British Columbia,
Canada.
2 Ibid.
41

from a distance of 1.5-3m from the tree. Cones had to be in direct sunlight with the sun
behind the photographer. A 15 x 15 cm square of cardboard covered in aluminum foil was
included in every photograph to measure absolute solar radiation and reflectance.
The photographs were analyzed by the respective photographer using ThermaCAM
Reporter 2000 Professional software (FLIR Systems 2004). The setting parameters were
emissivity at 0.90, which is the approximate emissivity for a pinecone (Tracy Zaradnik2,
personal communication); ambient temperature, as determined by the raw reading from the
average temperature of the total surface area of the aluminum-covered cardboard; distance at
1.5, 3.0, or 4.5 m, depending on distance from cone; atmospheric temperature, as taken from
the Environment Canada local temperature measurements to the closest 0.5 hour
(Environment Canada 2012); and relative humidity, as taken from Environment Canada to
the closest 0.5 hour (Environment Canada 2012). The average temperature of the cone
surface and the hottest point on cones visible in the image were recorded, and the average of
both measurements was calculated for each sample of 2-5 cones for each sampling event (the
number of cones sampled per ramet varied in 2008).

3.3.4.2 Host Terpenoid Sampling
In 2008, sampling for cone monoterpenes on ramets in each of the three preference
classes was done on 22 July. The circumference of the tree was divided into five sections,
and a single cone taken from each section about half-way up the crown, for a total of five
cones per ramet. In 2009, three cones per ramet, one from each of the three flagged cone
clusters, were selected for terpenoid sampling on each sampling occasion (26-28 June, 15-17
July, and 16-18 August). Pine cones collected for analysis were immediately placed on dry

42

ice and stored overnight in a -20 °C freezer. Cones were then shipped on dry ice to the
British Columbia Ministry of Forests and Range Analytical Chemistry Laboratory in
Victoria, BC3, where they were stored in a -70 °C freezer until extraction. For each sample
occasion, all cones from an individual ramet were pooled into a composite sample. Frozen
cones were immersed in liquid nitrogen, quickly milled in a large Wiley mill, and then
immediately returned to the -70 °C freezer until all samples were ready for extraction. A
subsample of frozen tissue was weighed and dried at 70 °C to determine moisture content.
For monoterpene extraction, approximately 0.4 g of ground sample was transferred to a glass
vial, and 4.0 ml of hexane with 250 ppm pentadecane as an internal standard was added.
The vial was tightly capped, left at 21-23 °C for 24 hours, shaken thoroughly, then left for
another 24 hours at 21-23 °C. After the 48 hour extraction, the sample was shaken again and
allowed to settle before transferring a 1 ml aliquot to a GC autosampler vial. The samples
were analysed on a Perkin Elmer Autosystem GC equipped with a flame ionization detector
using a J&W INNOwax, 25 M X 0.2 mm X 0.4 p. column. Instrument conditions were as
follows: Injector and detector temperatures: 200 °C and 300 °C respectively; helium flow
rate of 0.9 ml/min, ~20 psi; split ratio of 35:1; injection volume of 1 pi. Temperature
program: 60 °C for 1 min; then 3 °C/min to 85 °C; then 8 °C/min to 170 °C; then 20 °C/min
to 250 °C; then hold seven minutes (ASTM International, 2005). All of the pure analytical
standards used to identify and quantify each compound were purchased from Sigma Aldrich
Canada Ltd. (Oakville, ON, Canada), with the exception of sabinene, which was purchased
from Indofine Chemical Company (Hillsborough, NJ, USA). The terpene content was
recorded on an original dry weight basis.

3 P.O. Box 9536 Station Provincial Government, Victoria BC V8W 9C4 Canada
43

3.3.4.3 Physical Attribute Measurements and Cone Productivity
The width and diameter of all cones were measured using Vernier calipers
immediately following removal from the tree in both 2008 and 2009. Clone productivity data
(# cones/tree) were obtained from seed orchard staff at the end of August, following cone
collection in both years. Seed counts were not obtained in 2008, but were recorded in 2009
from one cone from each of the three cone clusters per tree and pooled into a single sample.
Seeds were extracted by dipping cones in boiling water for 15 seconds, baking in an oven at
52 °C for eight hours, shaking in a tumbler for three minutes, rubbing gently to de-wing,
screening to remove small debris, and rolling down an inclined plane to separate large debris.
Cleaned seeds were X-rayed (Hewlett-Packard Faxitron at 12 kV and 2.0 mA for 60 seconds
using Kodak A400 X-ray film) and filled, and empty seeds counted. Raw data were
expressed as “Total Seeds/ Cone” and “Filled Seeds/ Cone”. Filled seed/ cone is a generally
accepted measure of seed set in conifer seed orchards (Portlock 1996). Leptoglossus
occidentalis feeding can result in completely or partially empty seeds (Bates et al. 2000,
Connelly and Schowalter 1991). Only 14 to 18% of partially filled seeds are capable of
germination (Blatt and Borden 1998). Therefore, I considered any seeds that did not look
completely filled on the X-ray film to be empty. Seed X-ray photographs were taken by
W.B. Strong4.

4 B.C. Ministry of Forests, Lands, and Natural Resource Operations. Kalamalka Forestry
Center 3401 Reservoir Rd Vernon, BC Canada VIB 2C7
44

3.3.5 Statistical Analyses
To assess if insect preference rank by clone changed from 2008 to 2009,1 used
regression analysis to determine if there was a relationship between clone rank in 2008, and
clone rank in 2009. When ranking clones, if clone ranks were tied within a year, the rank
given to each of the tied ranks was the average of the ranks that they would have been
assigned had they not been tied (Zar 1999). Normality of errors and equality of variances
was examined by visually assessing residual plots. The response variable required a square
root transformation to meet the model assumptions that observations are independent across
treatments, variances are equal, and model errors are normally distributed (Zar 1999). I used
plots generated from classical discriminant function analysis to visually assess if quantitative
differences in the explanatory variables, when taken all together, would correctly classify
ramets into the predefined clone preference classes for the 2008 and 2009 data. In 2008, the
explanatory variables for all sampling events were cone monoterpenes, cone length, cone
diameter, clone productivity (# cones/ tree), and cone weight. In 2009, the explanatory
variables for all sampling events were cone monoterpenes, cone length, cone diameter, cone
weight, and cone temperature (determined from infrared photographs). Clone productivity
was omitted to allow comparison between sample periods. Seed set was measured after the
final sampling event in 2009. Clone preference class was used as the predefined grouping
factor. Statistical analyses were performed using the MASS library in R (Venables and
Ripley 2002).
Independent mixed effects analyses of variance (ANOVAs) were run with each
chemical or physical cone attribute, with clone preference class as a fixed factor and block
included as a random factor. Normality of errors and equality of variances was examined by

45

visually assessing residual plots. All response variables required a Logio transformation to
meet the model assumptions that observations were independent across treatments, variances
were equal, and model errors were normally distributed (Zar 1999). When the variable of
interest was found to vary significantly with cone preference class, a protected Mest was used
for post hoc comparison of means to determine significant differences between preference
classes (Carmer and Swanson, 1973). Statistical analyses were performed using R statistical
software (R Development Core Team, 2011).

3.4 Results
3.4.1 Clone Ranking and Clone Preference Classification
Clone rank in 2009 was consistent with clone rank in 2008, and can be modelled by
the equation V(clone rank 2009) = 2.106 + 0.103 * (clone rank 2008) where the slope of the
line indicates a significant relationship between clone rank in 2008 and 2009 (t = 35.10, d f 5 6,p< 0.0001) (Figure 3.1).
The clones assigned to preference classes in 2008 and 2009 are listed in Table 3.1 by
clone identification number, clone status, and clone provenance.

46

60 i

50 -

40 -

•M •

CD

O
O
2 30 ■
c
2
£

o
O

• ••
•• •

20 10

• •

-

0

10

30

20

40

50

60

Clone rank 2008
Figure 3.1 Correlation of clone rank in 2008 with that in 2009 in seed orchard 307 at
Kalamalka Lake Seed orchard in Vernon B.C., Canada. Clones with the same rank were
averaged.

47

Table 3.1 Clones sampled in 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed
orchard in Vernon B.C., Canada. Each clone that is from the same provenance is from a
different family. Numbers in parenthesis indicate how many ramets of that clone were
planted in a given year.
Year
ramets
planted

Clone
preference
class3

Provenance

Latitude
(N)

Longitude
(W)

Elevation
(m)

FO, FU

Champion Lakes

49.18

117.58

998

1510

1985(4)
1991(1)
1993(1)
1985(6)

FO, FU

Champion Lakes

49.18

117.58

998

1511

1985(6)

FO, FU

Champion Lakes

49.18

117.58

998

1505

1985(4)
1989(2)

NF

Champion Lakes

49.18

117.58

998

1506

1985(6)

NF

Champion Lakes

49.18

117.58

998

1520

1985(4)
1989(1)
1991(1)
1985(4)
1991(1)

NF

Settlers Road

50.52

115.73

1036

FO, FU

Champion Lakes

49.18

117.58

998

1511

1985(5)

FO, FU

Champion Lakes

49.18

117.58

998

1538

1985(4)

FO, FU

Larch Hills

50.70

119.18

777

1518

NF

Marl Creek

51.52

117.18

945

1528

1985(3)
1989(1)
1991(1)
1985(5)

NF

Inonoaklin

49.93

118.22

579

1534

1985(5)

NF

Larch Hills

50.70

119.18

777

Trial
Year

Clone

2008

1508

2009

1510

a Clones are designated as favoured occupied (FO), Favoured unoccupied (FU) or non-favoured (NF)

48

Canonical score plots from classical discriminant analysis revealed that the measured
host-preference cues contributed to the classification of clones into the pre-designated
preference in both years (Figures 3.2 a-d). There was overlap between occupied and
unoccupied favoured clones for the 17-22 July 2008 and 15-17 July 2009 sampling periods,
whereas these classes separated out almost completely for the 26-28 June, 2009 and 16-18
August, 2009 sampling periods.

49

b ) 26-28 June, 2009

a) 17-22 July, 2008
o o

•

<N
I

o

o
oo

o

o

0

o 13 o
o
o %
« % °°

-2 ■

•*
•

•
•

.... 1...... 1....."T.....1"
-6-4-2
0
3-

O

1 ...1
4 6

1-

o

----r ■■.. t----- -T..... —
1......n.......r-..... r..... —
T"
-6-4
-2 0 2
4 6 8
6-

▼
y*

8

▼

*

o* • * .

r

-3 -

7 -

0

d) 16-18 August, 2009

o
°° o
cp o
o

o

0-

V*T * t

▼

•• •
-2 •

o

4-

▼
f

°

1
8

c) 15-17 July, 2009

oo
CD
OO

2-

CM
o ots
to
UL
-1 -

1
2

-2 ■

• •
• •

A -

0

2

4

e

8

•
• •
-2

Factor 1

0

2

4

6

8

Factor 1

Figure 3.2 a-d. Canonical scores plots from classical discriminant analysis illustrating that the
measured host-preference variables contribute to the classification of clones into the
designated preference classes (o = favoured occupied; • = favoured unoccupied; ▼= non­
favoured clones) sampled from seed orchard 307 at Kalamalka Lake Seed orchard in Vemon
B.C., Canada.

50

3.4.2 Host-Selection Cues
3.4.2.1 Infrared Radiation Measurements
Mean cone temperature (mid-range infrared) did not differ by clone preference class
in either 2008, or any of the sampling periods in 2009 (Table 3.2). Maximum cone
temperature (mid-range infrared) was only measured in 2009 and only differed by clone
preference class in the 15-17 July sampling period (Table 3.2). The significance was due to a
higher mean maximum temperature in cones from favoured occupied clones than from non­
favoured clones (Figure 3.3).

51

Table 3.2. Results of mixed effects analyses of variance for infrared and physical attribute
measurements from all sampling events in 2008 and 2009 in seed orchard 307 at Kalamalka
Lake Seed orchard in Vernon B.C., Canada.
22
July 2008

26-28
June 2009

F2,46

P2,38

15-17
July 2009

16-18
August 2009

F

F

M easure
ment

Source o f
variation

Mean
length

clone
preference
class

25.02

<0.01

4.20

0.01

3.65

Mean
diameter

clone
preference
class

19.10

<0.01

7.90

<0.01

2.99

# Cones

clone
preference
class

1.87

0.17

n/a

n/a

n/a

Cone
weight at
harvest

clone
preference
class

21.09

<0.01

n/a

n/a

Mean
mid IR

clone
preference
class

1.74

0.19

0.58

High
mid IR

clone
preference
class

n/a

n/a

Seed per
cone

clone
preference
class

n/a

Percent
filled
seed per
cone

clone
preference
class

n/a

P

P

2,38

2,38

P

4.58

0.02

6.22

<0.01

n/a

1.05

0.36

n/a

n/a

3.83

0.03

0.57

1.57

0.22

0.99

0.38

0.18

0.84

5.88

<0.01

0.76

0.47

n/a

n/a

n/a

n/a

n/a

1.87

0.17

n/a

n/a

n/a

n/a

n/a

0.58

0.57

52

0.04

P

i0.06

©

z>
2
0
Q.
E
0

35

0 llT
C CO

30

-

25

-

+1
-3 o
X

0

E
c
0

0

20
Favoured
Occupied

Favoured
Unoccupied

NonFavoured

Figure 3.3. Average maximum surface cone temperature in “Celsius (± 1 S.E.) for the 15-17
July, 2009 sampling period in seed orchard 307 at Kalamalka Lake Seed orchard in Vemon
B.C., Canada. The presence of different lower case letters indicates a significant difference
between clone preference classes.

53

3.4.2.2 Host Terpenoids
Twenty-one known terpenoids were detected in cone samples from both 2008 and
2009 (see Appendix II). Only the results that differed significantly by clone preference class,
and were found in at least two sampling periods, are reported below. These compounds are
a-pinene, limonene, 5-3-carene, and p-cymene. For a complete list of results of analyses of
variance per sampling period see Appendix II.
The mean amount of 8-3-carene differed significantly by clone preference class in all
sampling periods across both years. Both favoured occupied and unoccupied clones had
significantly greater 5-3-carene than non-favoured clones in all sampling periods (Figure
3.4).
The mean amount of a-pinene differed by clone preference class in all 2009 sampling
periods (Table AII.1). Non-favoured clones had significantly greater a-pinene than favoured
occupied and favoured unoccupied clones in sampling periods 26-28 June and 15-17 July. In
the 16-18 August sampling period only favoured unoccupied clones differed from nonfavoured cloness (Figure 3.5).
The mean amount of limonene differed significantly by clone preference class in all
sampling periods (Table AII.l). Non-favoured clones had significantly less limonene than
both favoured clones classes in 2008 (Figure 3.6 a), but the opposite trend was observed in
all sampling periods in 2009; the mean level of limonene in non-favoured clones was
significantly greater than in favoured occupied and unoccupied clones (Figure 3.6 b-d).
The mean level of p-cymene differed significantly by clone preference class on
22 July 2008 and on 15-17 July 2009 (Table AII.l), with non-favoured clones having

54

significantly less p-cymene than both favoured clone classes. On 15-17 July 2009 nonfavoured clones and favoured unoccupied had a lesser amount of p-cymene than favoured
occupied clones (Figure 3.7).

55

22 July 2008
b

26-28 June 2009
-h

150

<?
to

150 -

15-17 July 2009
b

16-18 August 2009
b
100

-

Favoured
Occupied

Favoured
Unoccupied

NonFavoured

Figure 3.4 a-d) Mean levels of 8-3-carene in parts per million (± 1 S.E.) found in cones from
sampling events in 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in
Vernon B.C., Canada. The presence of different lower case letters indicates significant
difference between clone preference classes.

56

26-28 June 2009
LU

b

CO

+1

E
Ql
Q.
®
C

©

c
Q.
i
8
O
w

15-17 July 2009
100

-

CD

8>
8©
16-18 August 2009

100

-

Favoured
Occupied

Favoured
Unoccupied

NonFavoured

Figure 3.5 a-d) Mean levels of a-pinene in parts per million (± 1 S.E.) found in cones from
sampling events in 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in
Vernon B.C., Canada. The presence of different lower case letters indicates significant
difference between clone preference classes.

57

LU
CO

26-28 June 2009

60 H B

•H

E
Q.
3
©

C

a?
c

o
E
o

15-17 July 2009

w
©
©
c
©
©

16-18 August 2009
b

Favoured
Occupied

Favoured
Unoccupied

NonFavoured

Figure 3.6 a-d) Mean levels of limonene in parts per million (± 1 S.E.) found in cones from
sampling events in 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in
Vernon B.C., Canada. The presence of different lower case letters indicates a significant
difference between clone preference classes.

58

26-28 June 2009

i
15-17 July 2009

5

0.5

£

0.0
16-18 August 2009

I
Favoured
Occupied

Favoured
Unoccupied

NonFavoured

Figure 3.7 a-d) Mean levels of p-cymene in parts per million (± 1 S.E.) found in cones from
sampling events in 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in
Vernon B.C., Canada. The presence of different lower case letters indicates a significant
difference between clone preference classes.

59

3.4.2.3 Physical Attribute Measurements and Cone Productivity
Mean cone length differed significantly by clone preference class in all sampling
periods (Table 3.2). Mean cone length of the favoured unoccupied clone preference class
was consistently significantly greater than that of non-favoured clone preference classes
(Figure 3.8 a-b), whereas mean cone length of favoured occupied clones was only
significantly greater than non-favoured clones in sampling periods 17 July 2008 and 26-28
June 2009 (Figure 3.8 c-d).
Similarly, mean cone diameter differed by clone preference class except in sampling
period July 2009 (Table 3.2). The mean diameters of cones from the favoured clone
preference classes were greater than that of those from non-favoured clones on the remaining
three sampling periods, with the exception that the favoured occupied clone preference class
did not differ from non-favoured clones on 16-18 August, 2009 (Figure 3.9 a-b and d).
The estimated number of cones per tree did not differ by clone preference class in
either 2008 or 2009 (Table 3.2). Mean cone weight per tree differed by clone preference
class in both years (Table 3.2). The favoured clone preference classes did not differ from
each other, but weighed more than non-favoured clones, except that the favoured occupied
and non-favoured clone preference classes did not differ from each other in 2009 (Figure
3.10). Total seed per cone and percent seed filled did not differ by clone preference class in
2009 (Table 3.2).

60

22 July 2008

26-28 June 2009
b
LU

4.0 -

15-17 July 2009
b
®

4.0

16-18 August 2009
b

Favoured
Occupied

Favoured
Unoccupied

NonFavoured

Figure 3.8 a-d) Mean cone length in centimetres (± 1 S.E.) found in cones from all sampling
events 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in Vernon B.C.,
Canada. The presence of different lower case letters indicates a significant difference between
clone preference classes.

61

22 July 2008

26-28 June 2009
b

15-17 July 2009

16-18 August 2009
b

Favoured
Occupied

Favoured
Unoccupied

NonFavoured

Figure 3.9 a-d) Mean cone diameter in centimetres (± 1 S.E.) found in cones from all sampling
events in 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in Vernon B.C.,
Canada. The presence of different lower case letters indicates a significant difference between
clone preference classes.

62

Favoured
Occupied

Favoured
Unoccupied

Non_
Favoured

Figure 3.10 a-b) Mean cone weight in grams (± 1 S.E.) found in cones from 2008 and 2009 in
seed orchard 307 at Kalamalka Lake Seed orchard in Vernon B.C., Canada. The presence of
different lower case letters indicates a significant difference between clone preference classes.

63

3.5 Discussion
Host selection by insects involves a cantenation of responses to several potential
stimuli including visual, olfactory, gustatory, and mechanical cues (Visser 1986). My study
illustrates that feeding site selection by L. occidentalis is likely a complex process involving
a number of stimuli. While the differences between both favoured and non-favoured clone
preference classes are apparent, the slight separation seen in the discriminant analyses
between the occupied and non-occupied favoured clone preference classes suggests that even
small differences in levels of some of these stimuli may contribute to feeding site choice by
L. occidentalis. The higher degree of overlap between favoured occupied and favoured
unoccupied clone preference classes observed from the samples taken on 22 July 2008 and
on 15-17 July 2009 implies that feeding site selection in mid-July may be influenced by other
unmeasured characteristics of the trees and/or cones. In contrast, the greater degree of
separation between favoured occupied and favoured unoccupied clones on 26-28 June 2009
suggests that the levels of measured variables differed enough to influence feeding site
selection. Although this same separation was apparent in the 16-18 August samples, insect
counts were not conducted between the last two sampling events, so it is not possible to make
inferences regarding insect host choice for this period.
Genetically identical plants can exhibit substantial phenotypic variation related to site
and microclimate characteristics, this may explain some of the differences observed between
the favoured occupied and favoured unoccupied clone preference classes (Sultan 2000).
Seasonal variation in some or all of the monoterpenes may also have caused differences in
separation of favoured clone preference classes at different sampling events. For example,
seasonal changes in the release of terpenoid compounds have been observed in Japanese red

64

pine, (Pinus densiflora Siebold. et Zucc.), ponderosa pine (Pinusponderosa Douglas ex
Lawson), and bristlecone pine (Pinus longaeva Bailey) (Lim et al. 2008).
Consistent with Blatt and Borden’s (1996) work in the same seed orchard (orchard
307) in 1993 and 1994,1 found that highly infested clones in 2008 also had high infestation
levels in 2009. A few of the highly infested clones in my study were also highly infested in
the Blatt and Borden (1996) study. These clones may possess heritable characteristics
attractive to L. occidentalis that are consistent from year to year such as cone crop, or cone
size (Byram et al. 1986, Bilir et al. 2008). It is also possible that chemical defense systems
in clones preferred by L. occidentalis have lower levels of deterrent compounds than non­
favoured genotypes. The consistency in favoured clones observed over several years may
provide an effective way to select trees for monitoring and ultimately make management
decisions regarding L. occidentalis. Blatt and Borden (1996) also reached the same
conclusion. However, other highly ranked clones, including the most highly infested clones
in both 2008 and 2009, were not highly ranked in 1993 or 1994 (Blatt and Borden 1996).
This suggests that some of the differences observed in clone preference may be related to
physiological changes in the trees due to age (Boege et al. 2011). Trees in the orchard were
planted 8-9 years before the 1993-1994 studies, and 15 years elapsed between the two
studies. The carbon nutrient balance hypothesis of plant defense, advanced by Bryant et al.
(1983), includes the idea that as plants age their rate of growth declines freeing up certain
nutrients such as nitrogen and carbon. These can be used in secondary metabolism such as
inducible monoterpene production (Lewinsohn et al. 1991). Although there is much
evidence to support the carbon nutrient balance hypothesis, there are also several studies that
suggest otherwise (Herms and Mattson 1992, Stamp 2003). The growth-differentiation

65

balance hypothesis encompasses the ideas included in the carbon nutrient balance hypothesis,
and further advances the idea that other factors, notably environmental conditions such as
drought and average temperature, also play an important role in phenotypic variation in plant
defense (Herms and Mattson 1992, Stamp 2003).
Although L. occidentalis possesses infrared sense organs and has been shown to
orient to infrared cues in both laboratory and field experiments (Takacs et al. 2009), I did not
detect a significant relationship between mean cone temperature, measured as infrared
profiles of clones, and clone preference class in either 2008 or 2009. The average maximum
cone temperature differed by clone preference class only in the second sampling event in
2009 (July 15-17). Maximum cone temperature was not measured in the 2008 pilot study.
Although only a single analysis yielded significant results, both favoured clone preference
classes had a higher mean temperature than non-favoured clones consistently, albeit not
significantly, in all sampling events. Because cones are significantly warmer than foliage
and therefore provide a strong contrast, L. occidentalis could use infrared as a long- or
medium-range host location cue to identify cone-bearing trees (Takacs et al. 2009).
Populations of L. occidentalis in 2009 were so low that it was impossible to evaluate clone
preference early in the season. As a result, cones were likely sampled after host selection
occurred which means this study may not have been able to detect certain cues that are
important in the host-selection process. It has been noted that infrared navigation in the
haematophilic reduviid bug, Rhodnius proxilus Stal (Schmitz et al. 2000), and in the
pyrophilic buprestid beetle, Melanophila acuminata De Geer (Evans 1966, Evans 2010), are
used to orient to potential hosts, but are probably not used to ultimately determine their

66

suitability. Therefore, other factors, such as host volatiles may play a greater role in shortrange host selection and host acceptance.
Insects can be attracted or repelled by monoterpenes emitted from their host plants
(Keeling and Bohlmann 2006). These defensive compounds can be constitutive or induced
by insect damage. For example, feeding by Pissodes strobi Peck (Coleoptera: Curculionidae)
induces the release of monoterpene olefins in Sitka spruce (Picea sitchensis Bongard) (Miller
et al. 2005). Monoterpenes can play a role in mediating insect host location and acceptance.
For example, the rate at which Ips pint Say (Coleoptera: Curculionidae) entered phloembased media was greatest at medium concentrations of P-pinene (Wallin and Raffa 2000).
In this study, several monoterpenes, including bomyl acetate, myrcene and Pphellandrene, differed by clone preference class in a single sampling event. The level of ctpinene, limonene, o-3-carene, and p-cymene differed by clone preference class in at least two
sampling events. Since they are more likely to have influenced insect behaviour, I will
discuss only those monoterpenes differing by clone preference in two or more sampling
events.
S-3-Carene is known to be implicated in host defence against the carrot psyllid,
Trioza apicalis Forster (Hemiptera: Triozidae) (Nehlin et al. 1994), and thought to be
involved in host location in other Hemiptera (Visser and Yan 1995) and several other
coniferophagous insects (Rocchini et al. 2000, Duke and Lindgren 2006). In addition, attack
levels by the pine cone weevil, Pissodes validirostris Sahlberg (Coleoptera: Curculionidae))
on Scots pine {Pinus sylvestris L.) are positively correlated to levels of d-3-carene (Annila
and Hiltunnen 1977), and S-3-carene is also highly attractive to Dendroctonus valens Le

67

Conte (Coleoptera: Curculionidae) (Sun et al. 2004, Erbilgin et al. 2007). In the two years of
my study, 3-3-carene levels were consistently and significantly higher in the favoured clone
preference classes than in non-favoured clones, suggesting that 5-3-carene may be an
attractant for L. occidentalis. It seems unlikely that the higher levels are due to an induced
defensive response as both favoured occupied and favoured unoccupied clone preference
classes had a similar level of 3-3-carene.
a-Pinene is implicated in host location by several species o f Hemiptera (Chinta et al.
1994, Visser and Yan 1995, Visser 1996). a-Pinene is also used by the red shouldered bug,
Jadera haematoloma Herrich-Schaeffer (Hemiptera: Rhopalidae) in pheromone production,
and this monoterpene is known to influence response to aggregation pheromones of
numerous bark beetles (Borden et al. 1980, Hunt et al. 1989, Huber et al. 1989, Huber et al.
2000, Erbilgin et al. 2003). The levels of a-pinene did not differ significantly among clone
preference classes in 2008, but were higher in non-favoured than favoured classes in 2009,
with the exception of 16-18 August 2009, when favoured occupied and non-favoured clone
preference classes did not differ significantly. The relative abundance of L. occidentalis was
5.5 times higher in 2008 than in 2009. Crowding due to the greater numbers of L.
occidentalis observed in 2008 may have led to acceptance of higher levels of this compound
in 2008 than in 2009. Indeed, Wallin and Raffa (2002) found that above a certain population
density, I. pini tolerated higher levels of a-pinene. Limonene mediates several interactions
between insects and their host plants. It is used in host defence against the carrot psyllid, T.
apicalis Foerster (Hemiptera: Psyllidae) (Nehlin et al. 1994), host recognition by
Metopolophium dirhodum Walker (Hemiptera: Aphididae) (Visser and Yan 1995), as a
defensive compound in the burrower bug Sehirus cinctus cinctus Palisot de Beauvois

68

(Hemiptera: Cynidae) (Krall et al. 1997), and in pheromone production by the scentless plant
bug, Niesthrea louisianica Sailor (Hemiptera: Rhopaliodae) (Aldrich et al. 1979). Limonene
has also been shown to reduce the attraction of some insects to a-pinene (Hanula et al. 1985,
Nordlander 1990). In my study, limonene was found in significantly higher levels in both of
the favoured clone preference classes than in non-favoured clones in 2008, but in 2009 the
opposite trend was observed in all sampling events. As with a-pinene, it is possible that L.
occidentalis is more tolerant of high levels of limonene at a high population density. Wallin
and Raffa (2002) demonstrated that I. pini was more accepting of hosts with elevated
concentrations of limonene when the density of conspecifics was high. It is possible that the
reversal of limonene effects from 2008 to 2009 was due to the combined effect of one or
several compounds and insect crowding. For example, at low population densities high
levels of limonene may decrease L. occidentalis attraction to high a-pinene levels. Increased
levels of limonene mediating insect response to a-pinene has been observed in the pine
weevil, Hylobius abietis Linnaeus (Coleoptera: Curculionidae) (Nordlander 1990).
Alternatively, at a high population density there is a higher level o f feeding activity which
may have induced the higher level of limonene observed in 2008.
Insect response to host p-cymene levels has not been as well studied as other
monoterpenes. P-cymene, present in the essential oil of Eucalyptus grandis Hill ex. Maiden
(Myrtales: Myrtaceae), has been shown to elicit an electrophysiological response in the
eucalyptus brown looper, Thyrintaina amobia Stoll (Lepidoptera: Geometridae) (BatistaPereira et al. 2006). Lobesia botrana Denis and Shiffermuller (Lepidoptera: Tortricidae), a
pest in European grapes, is attracted to volatiles of a non-host, Tanacetum vulgare L.
(Asterales: Asteraceae), and also demonstrates an electrophysiological response to several

69

components including p-cymene (Gabel et al. 1992). A positive relationship between pcymene levels in Pinus sylvestris and cones attacked by P. validirostris has also been
established (Turgeon et al. 1994). Although the specific role of p-cymene in mediating
insect host location and/or acceptance is unknown, the significantly higher quantities found
in both of the favoured clone preference classes in 2008, and in the second sampling event in
2009, suggest p-cymene may be implicated in the attraction of L. occidentalis to favoured
clones.
While it is known that the monoterpene profile of a tree can change throughout its
lifetime, my data illustrate that change in cone terpene profiles also occur over the course of a
single growing season. These changes may be due to the seasonality of resource allocation to
growth and differentiation in lodgepole pine as there is experimental evidence for this in
several conifer species (Lim et al. 2008, Daly et al. 2010). The seasonality of host resource
allocation, pest physiological requirements, and environmental conditions all likely influence
L. occidentalis host choice.
I determined clone preference class in 2008 after initial host selection for
reproduction/oviposition had occurred. In 2009,1 was able to establish clone preference class
three weeks earlier than in 2008. However, it may be somewhat speculative to infer that the
host selection for reproduction/oviposition is linked to the terpene profile in clone preference
classes from the 26-28 June 2009 sample because I had, at this time, already sighted the
second generation. It is also important to note that direct comparisons between years should
only be made between for the 22 July 2008 data and the data from 15-17 July 2009 because
there may be seasonal changes in monoterpenes over the growing season (Lim et al. 2008,
Daly et al. 2010).

70

Blatt and Borden (1999) found that greater numbers oiL. occidentalis tended to be
found on lodgepole pine trees with greater numbers of cones, but that overall they preferred
clones with an intermediate cone crop. I found no difference in cone crop between clone
preference classes in either year. Rappaport and Roques (1991,1993) linked cone size with
clonal preferences of the Douglas-fir seed chalcid, Megastigmus spermotrophus Wachtl
(Hymenoptera: Torymidae), on Douglas-fir, but it is unclear how and at what stage clones
were selected for experiments in their studies. I used cone diameter, length, and weight as
stand-ins for cone size, and found that cones from both of the favoured clone preference
classes were significantly larger on average than non-favoured cones. Unfortunately, I was
not able to establish the favoured clones during peak oviposition which likely occurs midMay through mid-June. Because I measured cone size after oviposition, when cones were
ready for harvest, a definitive link between clone selection and a fitness benefit to offspring
cannot be assumed (Turgeon et al. 1994).
The percent of filled seed per cone was not statistically different between groups in
2009. It would have been useful to compare the damage in a low population year (2009) with
that o f a relatively higher population year, but unfortunately these data were not collected in
2008.
Overall, I did not find evidence for a specific primary host-selection signal used by L.
occidentalis. Multiple factors involving several sensory modalities, including cone terpenes,
cone infrared profile, and cone size, appear to interact in host location and selection by L.
occidentalis. Thermal cues, such as the cone and foliage infrared differential, may play a
role in mid- or long-range host selection as insects emerge to find food in spring. It is
believed that L. occidentalis use olfactory and/or gustatory cues (Takacs et al 2009) in

71

deciding to remain on a cone after landing on a suitable host. Similar scale-dependent
changes in host-orientation cues have been postulated for wood-feeding insects (SaintGermain et al. 2007).
To further identify the cues responsible for host selection in L. occidentalis, the next
step could be to conduct choice bioassays using cones from favoured and non-favoured
clones. Elucidating the cues involved in host selection by L. occidentalis has important
implications for pest management. If the key compounds that render a tree susceptible or
attractive to L. occidentalis can be identified, then perhaps clones could be phenotyped in
advance as a particular clone preference class. If the clones preferred by this insect are
known, a pesticide spray limited to these trees may control the majority of insects. Favoured
clone phenotypes could also have potential for use as trap crops: preferred plants and/or
cultivars grown close to the main crop that can be used to minimize damage (Shelton and
Badnes-Perez 2006). Pesticides could be applied to the trap crop thereby reducing the total
amount of pesticide required for control, the overall cost of pest management, and the
pesticide exposure of orchard workers (Cook et al. 2007). Determining the key compounds
that repel L. occidentalis may have important applications as well. Dormont et al. (1997) had
some success reducing the attack of seed-feeding insects on cones of mountain pine (Pinus
uncinata Ram.) using sprays made from oleoresin extracts of cones from Swiss stone pine (P.
cembra L.). It may be possible to manipulate the behaviour of L. occidentalis by spraying
trees with the extracts of repellent materials, or by baiting trap trees with attractive materials
such as 5-3-carene. Either of these approaches could ultimately reduce feeding damage to
seeds.

72

3.6 Literature Cited

Aldrich, J. R., Blum, M. S., Lloyd, H. A, Evans, P. H. and Burkhard, D. R. 1979. Novel
exocrine secretions from two species of scentless plant bugs (Hemiptera:Rhopalidae).
Entomologia Experimentalis et Applicata. 26: 323-331
Annila, E., and Hiltunnen, R. 1977. Damage by Pissodes validirostris (Coleoptera:
Curculionidae) studied in relation to the monoterpene composition in Scots pine and
lodgepole pine. Annales Entomologici Fennici. 43: 87-92
ASTM International, 2005: Standard Test Methods for Composition of Turpentine and
Related Terpene Products by Capillary Gas Chromatography, ASTM D 6387-99.
Bains, B., Isik, F., Strong, W. B., Jaquish, B., McLean, J. A., and El-Kassaby, Y. A. 2009.
Genetic resistance of spruce to gall-forming adelgids (Hemiptera: Adelgidae).
Canadian Journal of Forest Research. 39: 2536-2541.
Bates, S.L., Borden, J.H., Kermode, A.R., and Bennett, R.G. 2000. Impact of L. occidentalis
on Douglas-fir seed production. Journal of Economic Entomology. 93: 1441-1451.
Batista-Pereira, L.G., Fernandes, J.B., Correa, A.G., Da Silva, M.F.G.F., and Vieira, P.C.
2006. Electrophysiological responses of eucalyptus brown looper Thyrinteina amobia
to essential oils of seven Eucalyptus species. Journal of the Brazilian Chemical
Society. 17: 555-561.
Bilir, N., Prescher, F., Lindgren, D., and Kroon, J. 2008. Variation in cone and seed
characters in clonal Pinus sylvestris. New Forests. 36: 187-199.
Blatt, S.E., and Borden, J.H. 1996. Distribution and impact of L. occidentalis Heidemann
(Hemiptera: Coreidae) in seed orchards in British Columbia. The Canadian
Entomologist. 128: 1065-1076.
Blatt, S.E., and Borden, J.H. 1998. Interactions between the Douglas-fir seed chalcid,
Megastigmus spermotrophus (Hymenoptera: Torymidae), and the western conifer
seed bug, Leptoglossus occidentalis (Hemiptera: Coreidae). The Canadian
Entomologist. 130: 775-782.
Blatt, S.E., and Borden, J.H. 1999 Physical characteristics as host selection cues for
Leptoglossus occidentalis (Heteroptera: Coreidae). Environmental Entomology. 128:
246-254.
Boege, K., Barton, K.E., and Dirzo, R. 2011. Influence of Tree Ontogeny on Plant-Herbivore
Interactions. In Meinzer F.C. et al. (Eds.), Size and Age Related Changes in Tree
Structure and Function, Tree Physiology 4. (pp. 193-214). Springer: New York, NY.

73

Borden, J.H., Lindgren, B.S., and Chong, L. 1980. Ethanol and a-pinene as synergists for the
aggregation pheromones of two Gnathotricus species. Canadian Journal of Forest
Research. 10: 290-292.
Bryant, J.P., Chapin, III F.S., and Klein, D.R. 1983. Carbon/nutrient balance of boreal plants
in relation to vertebrate herbivory. Oikos. 40: 357-368.
Byram, T. D., Lowe, W. J., and McGriff, J. A. 1986. Notes: Clonal and annual variation in
cone production in loblolly pine seed orchards. Forest Science. 32: 1067-1073.
Cappuccino, N., and Kareiva, P. 1985. Coping with a capricious environment: a population
study of a rare pierid butterfly. Ecology. 66: 152-161.
Carmer, S. G., and Swanson, M. R. 1973. An evaluation of ten pair-wise multiple
comparison procedures by Monte Carlo methods. Journal of the American Statistical
Association. 68: 66-74.
Chinta, S., Dickens, J. C., and Aldrich, J. R. 1994. Olfactory reception of potential
pheromones and plant odors by tarnished plant bug, Lygus lineolaris (Hemiptera:
Miridae). Journal of Chemical Ecology. 20: 3251-3267.
Clark, E. L., Carroll, A. L., and Huber, D. P. 2010. Differences in the constitutive terpene
profile of lodgepole pine across a geographical range in British Columbia, and
correlation with historical attack by mountain pine beetle. The Canadian
Entomologist. 142: 557-573.
Connelly, A. E., and Showalter, T.D. 1991. Seed losses to feeding by L. occidentalis
(Heteroptera: Coreidae) during second year cone development in western white pine.
Journal of Economic Entomology. 84: 215-217.
Cook, S. M., Rasmussen, H. B., Birkett, M. A., Murray, D. A., Pyel, B. J., Watts, N. P., and
Williams, I. H. 2007. Behavioural and chemical ecology underlying the success of
turnip rape (Brassica rapa) trap crops in protecting oilseed rape (Brassica napus)
from the pollen beetle (Meligethes aeneus). Arthropod-Plant Interactions. 1: 57-67.
Courtney, S.P., and Kibota, T.T. 1990. Mother doesn’t know best: selection of hosts by
ovipositing insects. In Bemays E.A. (Ed), Insect plant interactions, Vol. II. CRC
Press, Boca Raton, FL.
Cronin, J.T., Abrahamson, W.G., and Craig, T.P. 2001. Temporal variation in herbivore
host-plant preference and performance: constraints on host-plant adaptation. Oikos.
28:312-320.
Daly, R. W., Helmig, D., Guenther, A. B., Baghi, R., Baroch, C., and Borke, C. 2010.
Seasonal biogenic volatile organic compound emission trends of four coniferous tree
species. In AGU Fall Meeting Abstracts. 1: 0227.

74

DeBarr, G. L., Merkel, E. P., O'Gwynn, C. H., and Zoerb, M. H. 1972. Differences in insect
infestation in slash pine seed orchards due to phorate treatment and clonal variation.
Forest Science. 18: 56-64.
Dicke, M. 2000. Chemical ecology of host-plant selection by herbivorous arthropods: a
multitrophic perspective. Biochemical Systematics and Ecology. 28: 601-617.
Dormont, L., Roques A., and Malosse, C. 1997. Efficiency of spraying mountain pine cones
with oleoresin of Swiss stone pine cones to prevent insect attack. Journal of Chemical
Ecology. 23: 2261-2274.
Duke, L.D., and Lindgren, B.S. 2006. Attack by Hylobius warreni on grafted lodgepole pine
and its relationships with monoterpene composition and scion:rootstock diameter ratio.
Agricultural and Forest Entomology. 8: 305-311.
Environment Canada. 2012. National Climate Data and Information Archive.
fwww.climate.weatheroffice.gc.ca~)
Erbilgin, N., Powell, J. S., and Raffa, K. F. 2003. Effect of varying monoterpene
concentrations on the response of Ips pini (Coleoptera: Scolytidae) to its aggregation
pheromone: implications for pest management and ecology of bark beetles.
Agricultural and Forest Entomology. 5: 269-274.
Erbilgin, N., Mori, S. R., Sun, J. H., Stein, J. D., Owen, D. R., Merrill, L. D .,... and Gillette,
N. E. 2007. Response to host volatiles by native and introduced populations of
Dendroctonus valens (Coleoptera: Curculionidae, Scolytinae) in North America and
China. Journal of Chemical Ecology. 33: 131-146.
Evans, W. G. 1966. Perception of infrared radiation from forest fires by Melanophila
acuminata De Geer (Buprestidae: Coleoptera). Ecology. 47: 1061-1065.
Evans, W.G. 2010. Reproductive role of infrared radiation sensors of Melanophila acuminata
(Coleoptera: Buprestidae) at forest fires. Annals of the Entomological Society of
America. 103: 823-826.
Finch, S. 1978. Volataile plant chemicals and their effect on host plant finding by the
cabbage root fly (Delia brassica). Entomologia Experimentalis et Applicata. 21: ISO159.
FLIR Systems. 2004. ThermaCAM Reporter 2000 Professional software. Danderyd, Sweden.
Fritz, R.S. and Simms, E.L (Eds.) 1992. Plant Resistance to Herbivores and Pathogens:
ecology evolution and genetics. University of Chicago Press.
Gabel, B., Thiery, D., Suchy, V., Marion-Poll, F., Hradsky, P., and Farkas, P. 1992. Floral
volatiles of Tanacetum vulgare L. attractive to Lobesia botrana Den. et Schiff.
females. Journal of Chemical Ecology. 18: 693-701.
75

Grevstad, F.S., and Klepetka, B.W. 1992. The influence of plant architecture on the foraging
efficiencies of a suite of ladybird beetles feeding on aphids. Oecologia. 92: 399-404.
Hanula, J. L., Berisford, C. W., and DeBarr, G. L. 1985. Monoterpene oviposition stimulants
of Dioryctria amatella in volatiles from fusiform rust galls and second-year loblolly
pine cones. Journal of Chemical Ecology. 11: 943-952.
Herms, D.A., and Mattson, W.J. 1992. The dilemma of plants: to grow or defend. The
Quarterly Review of Biology. 67: 283-335.
Huber, D.P., Gries, R., Borden, J.H., and Pierce, H.D. Jr. 2000. A survey of antennal
responses by five species of coniferphagous bark beetles (Coleoptera: Scolytidae) to
bark volatiles of six species of angiosperm trees. Chemoecology. 10: 103-113.
Hunt, D.W.A, Borden, J.H., Lindgren, B.S., and Gries, G. 1989. The role of auto-oxidation of
a-pinene in the production of pheromones of Dendroctonus ponderosae (Coleoptera:
Scolytidae). Canadian Journal of Forest Reasearch. 19: 1275-1282.
Joshi, A., and Thompson, J.N. 1995. Trade-offs and the evolution of host specialization.
Evolutionary Ecology. 9: 82-92.
Judd, G.J.R., and Borden, J.H. 1991. Sensory interaction during trap-finding by female onion
flies: implications for ovipositional host-plant finding. Entomologia Experimentalis et
Applicata. 58: 239-249.
Keeling, I.C., and Bohlman, J. 2006. Genes, enzymes and chemicals of terpenoid diversity in
the constitutive and induced defence of conifers against insects and pathogens. New
Phytologist. 170: 657-675.
Krall, B.S., Zilkowski, B.W., Kight S.L., Bartlet, R.J., and Whitman, D.W. 1997. Chemistry
and defensive efficacy of secretion of burrowing bug {Sehirus cinctus cinctus).
Journal of Chemical Ecology. 23: 1951-1962.
Lara, F.M., and Tanzini, M. R. 1997. Nonpreference of the lace bug Leptopharsa heveae
Drake & Poor (Heteroptera: Tingidae) for rubber tree clones. Anais da Sociedade
Entomologica do Brasil. 26: 429-434.
Lewinsohn, E., Gijzen, M., and Croteau, R. 1991. Defense mechanisms of conifers:
differences in constitutive and wound-induced monoterpene biosynthesis among
species. Plant Physiology. 96: 44-49.
Lim, J., Kim, J., Kim, K , Son, Y., Sunwoo, Y., and Han, J. 2008. Seasonal variations of
monoterpene emissions from Pinus densiflora in East Asia. Chemosphere. 73: 470478

76

Machial, L. A., Lindgren, B. S., and Aukema, B. H. (2012). The role of vision in the host
orientation behaviour of Hylobius warreni. Agricultural and Forest Entomology. 14:
286-294.
Miller, D.R., and Borden, J.H. 1990. The Use of monoterpenes as kairomones by Ips latidens
LeConte (Coleoptera: Scolytidae). The Canadian Entomologist. 122: 301-307.
Miller, B., Madilao, L.L., Ralph, S., and Bohlmann, J. 2005. Insect-induced conifer defense.
White pine weevil and methyl jasmonate induce traumatic resinosis, de novo formed
volatile emissions, and accumulation of terpenoid synthase and putative octadecanoid
pathway transcripts in Sitka spruce. Plant Physiology. 137: 369-382.
Mopper, S. 1996. Adaptive and genetic structure in phytophagus insect populations. Trends
in Evolutionary Ecology. 11: 235-238.
Nehlin, G., Valterova, I., and Borg-Karlson, A. 1994. Use of conifer volatiles to reduce
injury caused by carrot psyllid, Trioza apicalis Forster (Homoptera, Psylloidea).
Journal of Chemical Ecology. 20: 771-783
Nordlander, G. 1990. Limonene inhibits attraction to alpha-pinene in the pine weevils
Hylobius abietis and H. pinastri. Journal of Chemical Ecology. 16: 1307-1320.
Ott, D. S., Yanchuk, A. D., Huber, D. P., and Wallin, K. F. 2011. Genetic Variation of
Lodgepole Pine, Pinus contorta var. latifolia, Chemical and Physical Defenses that
Affect Mountain Pine Beetle, Dendroctonus ponderosae, Attack and Tree Mortality.
Journal of Chemical Ecology. 37: 1002-1012.
Papaj, D. R., and Rausher, M. D. 1987. Components of conspecific host discrimination
behavior in the butterfly Battusphilenor. Ecology. 68: 245-253.
Portlock, F.T. 1996. A field guide to collecting cones of British Columbia conifers. Victoria:
British Columbia Ministry of Forests. 91 p.
R Development Core Team. 2011. R: A language and environment for statistical computing.
R Foundation for Statistical Computing, Vienna, Austria.
Rappaport, N., Mori, S., and Roques, A. 1993. Estimating effect o f Megastigmus
spermotrophus (Hymenoptera: Torymidae) on Douglas-fir seed production: the new
paradigm. Journal of Economic Entomology. 86: 845-849
Rappaport, N.G., and Roques, A. 1991. Resource use and clonal variation in attack rate of the
Douglas-fir seed chalcid, Megastigmus spermotrophus Wachtl (Hymenoptera:
Torymidae) in France. The Canadian Entomologist. 123: 1219-1228.
Rocchini, L.A., Lindgren, B.S., and Bennett, R.G. 2000. Effects o f resin flow and
monoterpene composition on susceptibility of lodgepole pine to attack by the

77

Douglas-fir pitch moth, Synanthedon novaroensis (Lepidoptera: Sesiidae). Journal of
Applied Entomology. 124: 87-92.
Roques, A. 1987. Interaction between visual and olfactory signals in cone recognition by
insect pests. In Proc. Int. Symp. Insect-Plant Relationships. 6th. Pau 1986. Ed. V
Labeyrie, G Fabres, D Lachaise, pp. 153-60.
Saint-Germain, M., Buddie, C. M., and Drapeau, P. 2007. Primary attraction and random
landing in host-selection by wood-feeding insects: a matter of scale? Agricultural and
Forest Entomology. 9: 227-235.
Schmitz H., Trenner S., Hofmann M.H., and Bleckmann, H. 2000. The ability of Rhodinus
proxilus (Hemiptera: Reduviidae) to approach a thermal source solely by its infrared
radiation. Journal of Insect Physiology. 46: 745-751.
Seybold, S.J., and Tittiger, C. 2003. Biochemistry and molecular biology of de novo
isoprenoid pheromone production in the Scolytidae. Annual Review of Entomology.
48: 425-453.
Shelton, A.M., and Badnes-Perez, F.R. 2006. Concept and applications of trap cropping in
pest management. Annual Review of Entomology. 51: 285-308.
Stamp, N. 2003. Out of the quagmire of plant defense hypotheses. Quarterly Review of
Biology. 78: 23-55.
Sultan, S. E. 2000. Phenotypic plasticity for plant development, function and life history.
Trends in Plant Science. 5: 537-542.
Sim, J., Miao, Z., Zhang, Z., Zhang, Z., and Gillette, N. E. 2004. Red turpentine beetle,
Dendroctonus valens LeConte (Coleoptera: Scolytidae), response to host
semiochemicals in China. Environmental Entomology. 33(2): 206-212.
Takacs, S., Bottomley, H., Andreller, I., Zaradnik, T., Schwarz, J., Bennett, R., Strong, W.,
and Gries G. 2009. Infrared radiation from hot cones on cool conifers attracts seedfeeding insects. Proceedings of the Royal Society B-Biological Sciences. 276: 649—
655.
Thompson, J.N. 1988. Evolutionary ecology of the relationship between oviposition
preference and performance of offspring in phytophagus insects. Entomologia
Experimentalis et Applicata. 47: 3-14.
Tokushima, I., Orankanok, W., Tan, K. H., Ono, H., and Nishida, R. 2010. Accumulation of
phenylpropanoid and sesquiterpenoid volatiles in male rectal pheromonal glands of
the guava fruit fly, Bactrocera correcta. Journal of Chemical Ecology. 36(12): 13271334.

78

Turgeon, J.J., Roques, A., and de Groot, P. 1994. Insect fauna of coniferous seed cones.
Annual Review of Entomology. 39: 179-212.
Venables, W. N., and Ripley, B. D. 2002. Modem Applied Statistics with S. Fourth Edition.
Springer, New York.
Visser, J.H. 1986. Host odor perception in phytophagous insects. Annual Review of
Entomology. 31: 121-144.
Visser, J.H., and Yan, F.S. 1995. Electroantennogram responses of the grain aphids Sitobion
avenae F and Metopolophium dirhodum Walk (Horn, Aphididae) to plant odor
components. Journal of Applied Entomology. 119: 539-548
Wallin, K.F., and Raffa, K.F. 2000. Influences of host chemicals and internal physiology on
the multiple steps of postlanding host acceptance behavior of Ips pini (Coleoptera:
Scolytidae). Environmental Entomology. 9: 442-453.
Wallin K.F, and Raffa K.F. 2002. Density-mediated responses of bark beetles to host
allelochemicals: a link between individual behaviour and population dynamics.
Ecological Entomology. 27: 484-492.
Yanchuk, A. D., Murphy, J. C., and Wallin, K. F. 2008. Evaluation of genetic variation of
attack and resistance in lodgepole pine in the early stages of a mountain pine beetle
outbreak. Tree Genetics and Genomes. 4: 171-180.
Zar, J. H. 1999. Biostatistical analysis (Vol. 564). Upper Saddle River, NJ: Prentice hall.

79

Chapter 4: Calibration of walk-through field survey estimates of Leptoglossus
occidentalis (Hemiptera: Heteroptera: Coreidae) with abundance estimates based on
mark-release-recapture techniques

4.1 Abstract
An important component of an effective pest management strategy is the development
of an economic threshold. This requires standardized population census methods, which have
been successfully calibrated to estimate population. This could be achieved by establishing
the mathematical relationship between abundance estimates, using mark-release-recapture
methods, and walk-through surveys. Leptoglossus occidentalis is a pest insect in both
lodgepole pine and Douglas-fir seed orchards for which neither a standardized population
census method nor an economic damage threshold exists. Concurrent mark-release-recapture
studies and walk-through surveys for L. occidentalis were conducted in 2009 in lodgepole
pine and Douglas-fir seed orchards at the British Columbia Ministry of Forests Kalamalka
Seed Orchard and the Kalamalka Research Station respectively. Mark-release-recapture
surveys indicated that spray applications should be scheduled in mid- to late June, when
insect populations peaked in both years. The mathematical relationships between population
estimates and walk-through surveys in either location was not successfully determined. The
Jolly-Seber Method alone was used for population estimates in the lodgepole pine orchard
and the Schnabel, Schumacher-Eschemeyer, Jolly-Seber, and whole-tree methods were used
for population estimates in the Douglas-fir block. Data from mark-release-recapture surveys
indicated that the walk-through survey was unreliable at low populations.

80

4.2 Introduction
Accurate population density estimates of insect pests are critical for cost-effective
implementation of pest management strategies. An economic damage threshold for an
industry is obtained by setting the cost of pest management equal to the economic loss from
damage caused by a pest, it is an important decision-making tool that can help reduce the cost
of pest management actions as well as prevent excessive use of chemical pesticides (Pedigo
and Rice 2009). Economic damage thresholds have been successfully implemented for
several pest insect species, including the cabbage looper Trichoplusia ni Hubner
(Lepidoptera: Noctuidae) in cabbage crops (Greene 1972), and the potato leafhopper
Empoasca fabae Harris (Homoptera: Cicadellidae), which is an important pest on alfalfa and
several other plant crops in Canada and the Eastern United States (Cuperus et al. 1983).
The development of reliable economic thresholds for insect pests depends on
standardized population census methods. Standardized census methods utilizing markrelease-recapture methods have been developed for pests in several insect orders, including
Orthoptera and Lepidoptera (Pollard 1977, Thomas 1983, Haddad et al. 2008, Larsen and
Franzen 2009). In mark-release-recapture studies designed to estimate population size,
insects are captured, individually marked with an identification code so that data for
individual insects can later be collected, and then released back into the population. The
study area is re-sampled one to several times and captured insects are examined for marks.
The proportion of captured insects with marks is recorded (Krebs 1999, Southwood 2000,
Manly et al. 2005) and various statistical methods, such as the Jolly-Seber Method described
by Krebs (1999), are used to generate abundance estimates and confidence intervals of the
target insect species at a given time within a given area. These methods are commonly used

81

for organisms that are highly mobile. Mark-release-recapture studies can also provide insight
into other populations parameters such as births, deaths, immigration and dispersal, all of
which influence population size. For example, the Jolly-Seber model estimates probability of
survival and immigration rates as well as population size (Krebs 1999, Pollock and AlpizarJara 2005).
Mark-release-recapture methods are considered the most accurate approach to
population estimation (Williams et al. 2002). However, the population parameters used to
generate an estimate by these models may be highly inaccurate if low numbers of animals are
marked and/or recaptured (Haddad et al. 2007). While they can provide accurate data, these
methods can be costly as a considerable amount of time and effort is necessary to conduct
frequent sampling (Krebs 1999). Behavioural changes and physical damage to animals
caused by capture methods and marking can also influence the accuracy of population
estimates generated by mark-release-recapture studies. Additionally, most mark-releaserecapture studies have a set of very restrictive assumptions about the properties of the
populations being studied, for example several methods assume a closed population (no
immigration or emigration), which is unrealistic for highly mobile species (Krebs 1999,
Southwood 2000).
Validation of walk-through surveys using mark-release-recapture has been conducted
for several insect species. This generally involves one or more surveyors sampling a portion
of an area of interest. The relative abundance of insects determined from the walk-through
survey is then used to provide an estimate of the absolute population in the area of interest.
To provide greater accuracy, several paired relative (surveys) and absolute population
abundance estimates are conducted for different populations of the subject species. The

82

information can then be used to establish a mathematical relationship between the absolute
abundance and the relative abundance as measured by the surveys (Pedigo and Rice 2009).
For example, to develop census methods, Haddad et al. (2008) compared transect counts with
mark-release-recapture estimates for Neonympha mitchellii francisci French (Lepidoptera:
Nymphalidae), and Larsson and Franzen (2008) compared survey walks with mark-releaserecapture estimates for Andrena hattotfiana Fabricius (Hymenoptera: Andrenidae).
Determining if there is a stable relationship between field survey estimates and absolute
population abundance estimates is necessary for the development of an economic threshold.
This relationship can be estimated using multiple regression analysis on the results from
absolute population density estimates and visual monitoring. Once the relationship is
determined, survey data can be used in the equations to predict abundance at a given time in
a specific place (Zoebisch 1993, Larsson and Franzen 2008).
The potential seed loss per individual Leptoglossus occidentalis (Hemiptera:
Heteroptera: Coreidae) ranges from approximately 0.25 to 1.7 seeds per day (Bates et al.
2000, Strong 2006). The potential seed loss in an orchard can therefore be estimated by
multiplying the potential seed loss per individual, the estimated population density, and the
total area.. This information, together with seed prices and cost o f chemical controls, can be
used to obtain a quantitative economic threshold for the industry. The development of a
formula for economic threshold for L. occidentalis would be extremely valuable to seed
orchard managers, as this insect is considered a primary pest of conifer seed orchards (Strong
et al. 2001, Bates et al. 2002).

83

A repeatable visual monitoring system for L. occidentalis has been developed (see
methods) (W.B. Strong1, unpublished data), but a relationship between the relative
abundance generated from the surveys and the absolute abundance of insects in the field has
not yet been established. The goal of this study was to estimate absolute population
abundance in seed orchards by different types of population estimates, and to determine the
relationship between a visual monitoring system and absolute population density in seed
orchards.
4.3 Methods
4.3.1 Study Area
The study area is located in the Kalamalka Research Station and Seed Orchard in
Vernon, British Columbia, situated 3 km south of Vernon by Highway 97 on a hill between
the city and Kalamalka Lake (50.24° N , 119.28° W, 489m above sea level). The Kalamalka
seed orchard is in the Interior Douglas-fir biogeoclimatic zone (Pojar et al. 1987). Surveys
for population estimates of adult L. occidentalis were conducted from May through August in
a 4 ha lodgepole pine seed orchard (lodgepole pine orchard 307) in 2008 and 2009, and
during August in a 0.5 ha Douglas-fir orchard (Douglas-fir block 9) in 2009.

4.3.2 Trapping and Marking Methods
I used pheromone traps modified by Dr. Ward Strong1 (Unitraps, Contech Inc., Delta,
BC) to collect live insects. The top portion of the trap was removed and the entrance and
inside of the trap were painted with Insect-a-Slip™, a fluoropolymer resin (PTFE-30)
1 Research Scientist, B.C. Ministry of Forests, Lands, and Natural Resource Operations.
Kalamalka Forestry Center 3401 Reservoir Rd Vernon, BC Canada V1B 2C7

84

(BioQuip, Rancho Dominguez, CA), by brushing or wiping it onto the plastic. After drying,
the polymer forms a slippery surface that inhibits arthropods from gaining traction on the
treated area. When sighted, usually on a cone, L. occidentalis were captured by placing the
modified trap underneath the insect and tapping the top of the cone. When disturbed, L.
occidentalis drop before taking flight, causing them to fall into the ftuinel entrance of the
trap, where they are unable to gain enough traction to escape.. The entrance of the trap was
blocked using a piece of black nylon insect screening between captures to prevent L.
occidentalis from escaping.
In the 2008 mark-release-recapture studies, which required insects to have a unique
identifying code for analysis for the Jolly-Seber method, I used a technique that involved
painting the pronotum with Liquitex™ (Liquitex Artist Materials, Piscataway, New Jersey)
professional acrylic artist water-based colour paint and writing an alpha numeric code on top
of the paint mark with Pigma Micron™ (Sakura of America, Hayward, California) archival
ink pens (Figure 4.1). The paint colour identified the orchard in which the insect was caught,
and the alphanumeric code determined the capture/release date and the individual insect.

85

Figure 4.1. Marked Leptoglossus occidentalis. In a) 2008 insects were coded with Liquitex™
professional acrylic artist water-based colour paint and Pigma Micron™ archival ink pens
and in b) 2009 insects were coded with SRX Metallic Colorsharp™ permanent marker pens
and Pigma Micron™ archival ink pens. Photo a) was taken on 25 June 2008 by Ward Strong
in his lab at the Kalamalka Research Station, Vernon B.C. Canada. Photo b) was taken on 20
August 2009 in block 8 by Tamara Richardson at the Kalamalka Research Station, Vemon
B.C. Canada.

In 2009, to enable me to mark insects in the field, I replaced the Liquitex™
professional acrylic artist water-based colour paint with SRX Metallic Colorsharp™
permanent marker pens (MEGA Brands Inc., Montreal, Quebec). A unique number was
inscribed on the pronotum to identify the individual, and I used a colour dot on a forewing to
specify the date. In the mark-release-recapture studies done in Douglas-fir block in 2009,1
used mark-release-reacpature methods that only required the number of times and the date an
insect was sighted. For this study, insects were marked on their forewings with a dot in a dayspecific colour using SRX Metallic Colorsharp™ permanent marker. Each time an insect
was recaptured it was given an additional dot.

4.3.2.1 Mark durability and effect of marks on insects
Marks on insects did not degrade over the summer for both the studies conducted by
S. Desjardins2 and W.B. Strong1in 2006 and 2007, and those conducted by myself in 2008
and 2009. In fact, I recaptured a single marked insect in 2008 that had been released in 2007
and three more insects in 2009 that had been marked in 2008. The marks on these insects
were clear and confirmed the durability of the marking technique. Additionally, these
recaptures provided some circumstantial evidence that marking did not adversely affect L.
occidentalis.
In 2008, in order to examine the effect of marking on insect viability, I captured and
marked 15 L. occidentalis with metallic pens. I placed the marked insects in a Bug Dorm 2™
(BioQuip, Rancho Dominguez, CA), along with 15 unmarked insects, with some fresh cones

2 Professor Mathematics, Irving K. Barber School of Arts and Sciences, University of British
Columbia Okanagan, Kelowna BC

87

and Petri dishes containing water-soaked paper towels for water source. Survival and activity
levels were assessed after two weeks of containment. In 2009, to determine whether marking
influenced predator attraction to L. occidentalis, I collected 60 insects and subjected them to
one of six treatments: no mark, or metallic paint on the pronotum in silver, gold, metallic
green, metallic blue, or metallic purple. I randomly selected 10 trees in Douglas-fir block 8
at the Kalamalka Research Station, and tethered one insect from each of the six treatments to
a cone on six neighbouring branches of each tree. Tethers of dark green thread, 20 cm in
length, were tied immediately behind the pronotum. Insects were checked daily for ten days.
After ten days, only one control and one insect marked with purple paint were gone; the
remaining 58 insects were still alive. I did not observe what happened to the missing marked
insect but witnessed the unmarked insect being consumed by a predatory wasp
(Hymenoptera: Vespidae).

4.3.3 Population Estimation Methods
The Jolly-Seber method (Krebs 1999), which is an individual mark-release-recapture
method, was used for population estimates in lodgepole pine orchard 307 in both 2008 and
2009. In 2009, four different methods were used in the Douglas-fir block to compare
population abundance estimates: the Schnabel and the Schumacher-Eschemeyer methods,
both closed population methods; the Jolly-Seber method, an open population estimate (Krebs
1999); and whole tree counts, a type of quadrat method (Krebs 1999). The Leslie, Chitty and
Chitty test, designed for the Jolly-Seber model was used to test the assumption of equal
catchability (Krebs 1999). Regression analysis was used to establish if there was a
relationship between population estimates and the number of insects observed in concurrent

88

afternoon walk-through surveys in Douglas-fir block 9 and lodgepole pine orchard 307 in
2009. Walk-through surveys were conducted in both the morning and the afternoon. In
orchard 307, because very few L. occidentalis were observed, I combined afternoon and
morning walk-through surveys for the regression analysis. For Douglas-fir block 9, the
afternoon walk-through survey data was used in the analyses as more L. occidentalis were
observed during that time.
In lodgepole pine orchard 307, surveys for Jolly-Seber estimates involved visually
scanning every tree using orchard lifts. Each survey event required two days to complete.
During the summer of 2008, the surveys in lodgepole pine orchard 307 were initiated on May
5, 17,23, 28, and June 4,10, 20, and 28. During the summer of 2009, the surveys in
lodgepole pine orchard 307 began on May 8, 15, 23, 28, June 2,13, 18, 25, and on July 3 and
10. The six surveys for the Jolly-Seber estimates in Douglas-fir block 9 were conducted on
foot and were each completed in one day on August 12, 16, 23, 24,27 and 28 as Douglas-fir
block 9 had a much smaller area than orchard 307. Each time a marked insect was captured,
the code, a combination of colour and dots, the time of day, the clone upon which it was
found, and the location were recorded.
Three additional surveys were conducted between August 10 and 28, 2009 in
Douglas-fir block 9 to generate a dataset for both the Schnabel and the SchumacherEschemeyer population estimates. These involved surveying the entire block using three
scouts walking a set path, for a total of three separate paths. Each survey was done over two
days and every tree in the block was scanned. For each survey, each path was scanned
twelve times for the first two surveys, and fourteen times for the final survey. For these two
methods, since individual codes were not required, day-specific coloured dots were applied
to the forewings of L. occidentalis each time one was captured.
89

The whole tree method was employed three times between August 17 and 24,2009,
and involved a single scout examining 30 randomly selected trees in the Douglas-fir block.
Marks were not required for the whole-tree estimation method instead the numbers of L.
occidentalis observed on each tree was recorded at each sampling occasion.
In 2009, concurrent with all surveys in both lodgepole pine orchard 307 and Douglasfir block 9, visual walk-through monitoring surveys were conducted twice daily, in late
morning and mid-afternoon using the technique developed by W.B. Ward Strong1.
Monitoring was conducted at temperatures between 15 and 32 °C, wind speeds were below
15 km/hr, and there had been no rain for at least one hour. The visual monitoring method
involved walking diagonal transects through the orchard at approximately 1.5 km/hr,
examining cones on trees as they were passed, and recording the number of L. occidentalis
observed. This technique has previously been shown to provide repeatable monitoring results
with low variance (W. B. Strong1, unpublished data).

4.3.4 Statistical Analysis
Population estimates, using the methods described above and regression analyses,
were calculated using R statistical software (R Development Core Team 2012). Methods,
equations, and variables are defined in Appendix III.

4.4 Results
4.4.1 Jolly-Seber population estimates in lodgepole pine orchard 307
In 2008, a total of 732 Leptoglossus occidentalis were captured and marked; 23 of
these were recaptured once, and one was recaptured twice for a recapture rate of 3.28%. In
2009, a total of 179 L. occidentalis were captured and marked; 14 of these were recaptured
90

once, and one was recaptured twice for a recapture rate of 8.38 %. There were sufficient
recaptures of L. occidentalis to calculate population estimates and confidence intervals using
the Jolly-Seber method in both 2008 and 2009. Survey dates, population estimates, and
confidence intervals from the 2008 and 2009 surveys can be found in Table 4.1 and 4.2
respectively. The Leslie, Chitty and Chitty test for equal catchability indicated that the
number of new individuals entering the marked population was underestimated by 92% in
2008 and by 82% in 2009.
Walk-through surveys in lodgepole pine orchard 307 were conducted only in 2009.
Only two L. occidentalis were observed during these walk-through field surveys (Table 4.2)
so I was able to obtain population estimates for only six sampling intervals, all taken during
afternoon surveys. Consequently, I was unable to establish a relationship between the
population estimates of L. occidentalis, and afternoon walk-through surveys in 2009 using
regression analysis (F/,*=0.106, />>0.760, J?2=0.026). However, I was able to determine
general population trends for both 2008 and 2009 including the time at which the firstgeneration population peaked.

91

Table 4.1 2008 population estimates of Leptoglossus occidentalis per 4 ha in lodgepole pine
orchard 307 from mark-release-recapture data using the Jolly-Seber method of population
estimation at the Kalamalka Lake Seed orchard in Vernon B.C., Canada. Confidence
intervals are within parentheses.
Sample

Date

Jolly-Seber estimate3

1

17 May, 2008

-

2

23 May, 2008

-

3

28 May, 2008

-

4

4 June, 2008

1030 (266; 8 629)

5

10 June, 2008

2322(855; 11 003)

6

20 June, 2008

7378 (2 509; 39 146)

7
a

28 June, 2008
designates that no estimate can be made of this parameter from the data.

92

Table 4.2. 2009 population estimates of Leptoglossus occidentalis per 4 ha in lodgepole pine
orchard 307 from mark-release-recapture data using the Jolly-Seber method of population
estimation at the Kalamalka Lake Seed orchard in Vernon B.C., Canada. Also included is the
number of L. occidentalis found during walk-through surveys. Confidence intervals are
within parentheses.
Number of L.
occidentalis found
in morning walk­
Jolly-Seber estimate a through surveys

Number of L.
occidentalis found
in afternoon walk­
through surveys

1 8 May, 2009

-

0

0

2 15 May, 2009

-

0

0

3 23 May, 2009

-

0

0

4 28 May, 2009

241 (38, 5317)

0

0

5 2 June, 2009

36(11,351)

0

0

6

13 June, 2009

364 (82,4005)

0

0

7 18 June, 2009

453 (132,3341)

0

1

8 25 June, 2009

174(62, 1681)

0

0

9 3 July, 2009

42 (14, 559)

0

0

10 10 July, 2009

-

1

0

Sample

a

Date

designates that no estimate can be made of this parameter from the dat

93

Table 4.3. Total number of L. occidentalis captured during each sampling interval in 2008,
including single and double recaptures at the Kalamalka Lake Seed orchard in Vernon B.C.,
Canada. Surveys were done on 5 May and 10 May, 2008 but no L. occidentalis were sighted
or captured during this time
Total number of L.
occidentalis found
during sampling period

Number
of single
recaptures

Number of
double
recaptures

1 17 May, 2008

1

0

0

2 23 May, 2008

9

0

0

3 28 May, 2008

74

0

0

4 4 June, 2008

52

1

0

5

10 June, 2008

132

5

0

6 20 June, 2008

325

9

0

7 28 June, 2008

139

8

1

Date

Sample

Table 4.4. Total number of L. occidentalis captured during each sampling interval in 2009
including single and double recaptures in orchard 307 at the Kalamalka Lake Seed orchard in
Vemon B.C., Canada.

Sample

Date

Total number of L.
occidentalis found
during sampling period

Number of Number of
single
double
recaptures recaptures

1 8 May, 2009

5

0

0

2

15 May, 2009

4

0

0

3 23 May, 2009

11

0

0

4 28 May, 2009

21

0

0

5 3 June, 2009

7

0

0

6

13 June, 2009

22

1

0

7 16 June, 2009

38

2

0

8 23 June, 2009

52

1

1

9 3 July, 2009

13

8

0

10 11 July, 2009

18

3

0

94

4.4.2 Schnabel, Schumacher-Eschemeyer, Jolly-Seber and whole-tree methods o f population
estimates in Douglas-fir block 9
I obtained population estimates and confidence intervals using the Schnabel,
Schumacher-Eschemeyer, Jolly-Seber, and whole-tree methods in the Douglas-fir block
(Table 4.5). Both the Schnabel and Schumacher-Eschemeyer estimates (closed population
methods), as well as the whole-tree method yielded similar population estimates. The JollySeber method (open population method) yielded a higher estimate. The Leslie, Chitty and
Chitty test for equal catchability indicated that the number of new individuals entering the
marked population was underestimated by 95%.
Greater numbers of L. occidentalis were observed during walk-through surveys in the
afternoon than in the morning so I used the numbers from the afternoon surveys to conduct
the analyses. I averaged the number of insects sighted during walk-through surveys on both
survey days for all estimation methods except for the whole tree method since it only took
one day to complete the survey for this method. As in the lodgepole pine surveys, I was
unable to establish a relationship between population estimates of L. occidentalis and walk­
through survey results using regression analysis for the Schnabel method (F/j= 41.78 ,
p>0.098,Jf2=0.953), the Schumacher-Eschemeyer method (F/,/=8.736, p>0.208, R2= 0.795),
and the whole-tree estimate (Fj j = 6.322, p>0.241 , R2=0.726) due to low populations in
2009 and because I had only three sampling intervals,. Likewise, I was unable to use
regression analysis to establish a relationship between Jolly-Seber estimates and walk­
through surveys in Douglas-fir block 9 because I could obtain only two population estimates
using the Jolly-Seber method.

95

Table 4.5. Population estimates and of L. occidentalis per 0.5 ha in Douglas-fir Block 9
using the Schnabel, Schumacher-Eschemeyer and Jolly-Seber estimation methods at the
Kalamalka Lake Research Station in Vernon B.C., Canada. Also included are the numbers
of L. occidentalis found during walk-through surveys. Confidence limits are within
parentheses.
Number o f L.
occidentalis
found in
afternoon walk­
through surveys

Date

Schnabel
estimate

SchumacherEschemeyer
estimate

Jolly-Seber
estimate"

Number o f L,
occidentalis
found in
morning walk­
through surveys

10-11 Aug

853 (538,1579)

880 (547, 2 239)

-

1

8

21-22 Aug

410 (249, 797)

403 (251, 1 028)

625 (7 8 ,4 795)

2

6

25-26 Aug

3 0 6 (1 8 7 ,5 3 0 )

399 (227, 1 658)

781 (149, 1 737)

2

5

a

designates that no estimate can be made of this parameter from the data.

96

Table 4.6. Population estimates of L. occidentalis per 0.5 ha in Douglas-fir
Block 9 using the whole-tree method at the Kalamalka Lake Research Station in Vernon
B.C., Canada.. Also included are the numbers of L. occidentalis found during walk-through
surveys. Confidence limits are within parentheses.
Number o f L. occidentalis
morning walk-through

Number o f L. occidentalis
afternoon walk-through

Date

Whole-tree
estimate*

17 Aug

250 (134,419)

8

9

20 Aug

3 1 2 (1 8 0 ,4 9 8 )

6

9

24 Aug

4 1 6 (2 5 9 , 626)

0

4

97

4.5 Discussion
Obtaining accurate estimates of absolute population density using mark-releaserecapture can be challenging, especially at low population densities when scouts have
difficulty finding the study organism in its environment. All of the mark-recapture methods
used assumed that all insects in the population have the same probability of capture. The
Leslie, Chitty and Chitty tests designed for use with Jolly-Seber methods demonstrated
clearly that this assumption was not met. The number of new individuals entering the
marked populations in lodgepole pine orchard 307 in 2008 and 2009, and in Douglas-fir
block 9 was underestimated by as much as 82% to 95%. The Jolly-Seber method, designed
for open populations, allows for high-mobility of L. occidentalis as well as immigration,
emigration, births, or deaths. However, both the Schnabel and Schumacher-Eschemeyer
methods assume a closed population so that deaths, immigration and emigration might have
contributed to uncertainty in population estimates. However, since the majority of L.
occidentalis eggs had hatched at the time of surveys, births would not have been an issue.
The influence of these factors may have been negligible given that the surveys were
conducted over a relatively short period of time. This would not have been the case with the
mobility of L. occidentalis since the majority of L. occidentalis dispersed rapidly following a
disturbance such as the marking process, or when a scout accidentally collided with a tree
while using an orchard lift. In my study, the uncertainty in population estimates, reflected by
the wide confidence intervals associated with the estimates was likely due in part to low
population densities.
Factors affecting population change in insects may include competition,
disease/parasitism, mortality overwintering, and availability of resources. In addition,
several populations of insects are known to go through somewhat predictable outbreak cycles
98

(Levia et al. 2011). For example, the western spruce budworm, Choristoneura occidentalis
Freeman (Lepidoptera: Tortricidae) exhibits a 20-33 year cycle (Swetnam and Lynch, 1993).
Since we do not have long-term data for L. occidentalis it is unknown if the species has
predictable population outbreak cycles.
The average daily temperature from December through March were higher in winter
2007/2008 (with a minimum winter temperature of -17.7 °C occurring 5 January 2008) than
in winter 2008/2009 (with a minimum temperature of -27.8°C occurring 8 December 2008)
which might explain the lower numbers of L. occidentalis observed during summer 2009
(Environment Canada 2012). Moreover, similar to other conophages, L. occidentalis
populations fluctuate with the availability of food resources (Turgeon et al. 1994). Long­
term studies have demonstrated that cone abundance and the size of larval populations are
linked. For example, the population of Strobilomya spp. in the French alps increases with
cone abundance (Yao et al. 1991).
Survey bias can also contribute to differences in population estimates (Jones III et
al. 1998). In order to minimize potential differences in the detection rate of L. occidentalis
with different surveyors, each scout conducted one full practice survey prior to starting the
actual study surveys. However, Leptoglossus occidentalis are relatively cryptic and can be
easily missed, particularly when they are not on cones. In 2008, the rate at which the scout
and I were capturing and marking insects after two surveys was approximately equal. The
scout working with me in 2009 took much longer to develop a search image, and was
capturing only 75% of what I was capturing (see Appendix III). As a result, it is likely that
the counts do not accurately reflect what was actually in the field.
During the walk-through surveys, the scout was able to scan only the bottom half of
most trees because the upper crown was not easily visible from the ground since trees on
99

average were six or more meters tall. Seed cones are present in greater numbers in the top
five whorls of a lodgepole pine tree (O’Reilly and Owens 1988), and L. occidentalis prefer to
be in the sun (Hedlin et al, 1981, personal observation). It is therefore possible that L.
occidentalis may be largely confined to the upper crown, and thus difficult to sight from the
ground. Studies on insect trap placement and efficacy have shown that certain insects are
trapped at particular heights more than others, indicating a height preference. For example,
greatest numbers of male Dioryctria abietivorella (Lepidoptera: Pyralidae) were caught in
pheromone traps placed at the top of host tree canopies (Whitehouse et al. 2011), higher
numbers of Ips duplicatus (Coleoptera: Curculionidae: Scolytinae) were caught at mid­
canopy level in spruce in China (Chen et al. 2009), Culicidae, in the genus Culex
differentially (Diptera: Culicidae), favoured bird-baited traps at distinct canopy levels in
New York state (Darbo and Harrington 2006), and Cydia pomonella (Lepidoptera:
Tortricidae) showed gender differences in canopy distribution in apple orchards in Michigan
(Epstein et al, 2011).
At high populations, individuals may be distributed more evenly throughout the
canopy, this would mean that sampling from the ground can be a problem at low, but not
high population densities. For example, predictive sampling methods for western spruce
budworm, Choristoneura occidentalis Freeman, have been developed based on the withinbranch and tree distribution of egg masses, larvae and pupae (Torgersen et al. 1993). In
addition, monitoring L. occidentalis from the ground in young seed orchards, where the
upper canopy is easily within sight of scouts, may also be viable.
Overall, the study in orchard 307 did provide some insight into seasonal population
trends that could be used to schedule pesticide applications for L. occidentalis in lodgepole
pine seed orchards. Both the 2008 and 2009 surveys demonstrated that L. occidentalis
100

populations peaked in mid to late June, despite differences in mean temperatures and
population density. In BC seed orchards, there is currently no reliable decision-making tool
used to schedule spray applications. The decision to apply insecticides at the Kalamalka
Seed Orchard is based on surveys conducted by seed orchard staff well-trained in sighting L.
occidentalis. Results differ in younger and older orchards as trees in the former are much
smaller which makes it easier to sight insects. In orchards with older trees, such as orchard
307, the timing of sprays is based on 30 minute surveys done concurrently by two scouts on
foot or in orchard lifts. At the time of my study, workers in orchard 307 were doing surveys
on foot. There is no specific threshold for insects but generally, if two or more insects are
sighted, pesticides are applied (Gary Giampa3, personal communication).
There have been numerous studies that clearly demonstrate the effect of the insect
on seed (Bates et al. 2002a, Bates et al. 2002b, Strong et al. 2001, Strong 2006), yet
substantial seed loss continues to be reported by seed orchards in spite of the application of
insecticides (Chris Walsh4 and Jim Corrigan5, personal communication). Adult females eat
more seed than any other life stages, and their feeding is greatest prior to mating and
oviposition which occurs from the end of May through early July, peaking in mid-June
(Strong 2006, Bates and Borden 2005). Seeds may also be more vulnerable at this time since
the seed coat is not hard yet (Krugman and Koerber 1969).
In both 2008 and 2009, sprays for L. occidentalis were applied in July well after the
peak numbers were observed in mid-June. By July, first generation adults are in decline and
would have already caused the majority of damage to seeds. Based on my findings, applying
pesticides in early June would provide more effective control. This conclusion is supported
3

Kalamalka Seed Orchard, BC Ministry o f Forests and Range, Vernon BC

4 Ibid.
5 Ibid.
101

by a population simulation model, based on published demographic parameters, that predicts
that peak populations would occur in early June (S. Desjardins2, personal communication).
My results also suggest that current survey methods are inadequate for management
decisions particularly at low population densities. Since L. occidentalis favours specific
clones in orchards regardless of tree species, and since preferences are consistent among
years, I propose that increasing the monitoring intensity and focusing on clones that are
favoured by L. occidentalis would be a more effective way to determine infestation levels in
seed orchards. Spray programs should also target favoured clones. This could improve pest
control and potentially reduce the size of the spray area.
The four abundance estimation methods used in Douglas-fir block 9 varied
substantially. It was therefore not possible to establish a relationship between walk-through
surveys and abundance estimates. This might be due to both the low populations observed in
2009, and the fact that I only conducted surveys on three sampling dates. Typically, several
surveys are used to establish a relationship between walk-through surveys and population
estimates (Pedigo and Rice 2009). Here again, it was difficult to sight the insects on cones as
L. occidentalis are able to hide within the dried-out Douglas fir cones (Figure 4.2b).
Both studies demonstrate the difficulty in accurately assessing the population size of an
insect population, particularly at low densities. To establish a survey method that reflects
actual population levels, it is necessary to repeat this research at a time when L. occidentalis
populations are moderate to high to determine if there is a lower threshold at which walk­
through surveys become more reliable.

102

Figure 4.2. L.occidentalis on a) lodgepole pine cones and b) Douglas-fir cones. On lodgepole
pine it can be easily overlooked as the colour of the sclerotized portion of the forewing and
dried male cones are very similar. On mature Douglas-fir cones the adult insect can hide
between cone scales rendering it difficult to sight. Photograph a) was taken by Tamara
Richardson on 25 June 2008 in orchard 307 at the Kalamalka Seed Orchard in Vernon B.C.,
Canada. Photograph b) was taken by Tamara Richardson on 24 August 2009 in block 9 at
the Kalamalka Research Station in Vernon B.C., Canada.

4.6 Literature Cited

Amstrup, S.C., McDonald, T.L., and Manley, B.F.J. 2005. Handbook of Capture-Recapture
Analysis. Princeton University Press, Princeton, NJ.
Bates, S.L., Borden, J.H., Kermode, A.R. and Bennett, R.G. 2000. Impact of L. occidentalis
on Douglas-fir seed production. Journal of Economic Entomology. 93:1441-1451.
Bates, S.L., Lait, C.G., Borden, J.H. and Kermode, A.R. 2002a. Measuring the impact of L.
occidentalis on seed production in lodgepole pine using an antibody-based assay.
Journal of Economic Entomology. 95: 770-777.
Bates, S.L., Strong, W.B. and Borden, J.H. 2002b. Abortion and seed set in lodgepole and
western white pine conelets following feeding by Leptoglossus occidentalis
(Heteroptera: Coreidae). Environmental Entomology. 31: 1023-1029
Bates, S.L. and Borden, J.H. 2005. Life table for Z,. occidentalis Heidemann (Heteroptera:
Coreidae) and prediction of damage in lodgepole pine seed orchards. Agricultural and
Forest Entomology. 7: 145-151.
Chen, G., Wang, Y., Liu, G., Zhou, X., Niu, J. and Schlyter F. 2009. Catching Ips duplicatus
(Sahlberg) (Coleoptera: Scolytidae) with pheromone-baited traps: optimal trap type,
colour, height and distance to infestation. Pest Management Science. 66:213-219.
Craig, C.C. 1953. On the utilization of marked specimens in estimating populations of flying
insects. Biometrika 40:170-176.
Cuperus, G.W., Radcliffe, E.B., Barnes, D.K. and Marten, G.C. 1983. Economic injury
levels and economic thresholds for potato leafhopper (Homoptera: Cicadellidae) on
alfalfa in Minnesota. Journal of Economic Entomology. 76:1341-1349.
Darbro, J.M. and Harrington, L.C. 2006. Bird-baited traps for surveillance of West Nile
mosquito vectors: effect of bird species, trap height, and mosquito escape rates.
Journal of Medical Entomology. 43:83-92.
Environment Canada. 2012. National Climate Data and Information Archive.
(www.climate.weatherofflce.gc.ca)
Epstein, D.L., Stelinski, L.L., Miller, J.R., Grieshop, M.J.and Gut, L.J. 2011. Effects of
reservoir dispenser height on efficacy of mating disruption of codling moth
(Lepidoptera: Tortricidae) in Apple Pest management Science. 67:975-979.
Greene, G.L. 1972. Economic damage threshold and spray interval for cabbage looper
control on cabbage. Journal of Economic Entomology. 65: 205-208.

104

Haddad, N.M., Hudgens, B., Damiani, C., Gross, K., Keufler, D. and Pollock, K. 2008.
Determining optimal population monitoring for rare butterflies. Conservation
Biology. 22: 929-940.
Hedlin, A.F., Yates, III H.O., Tovar, D.C., Ebel, B.H., Koerber, T.W. and Merkel, E.P. 1981.
Cone and seed insects of North American conifers. Canadian Forestry Service, US
Department of Agriculture,Forest Service and Secretaria de Agricultura y Recursos
Hidraulicos, Mexico.
Jolly, G.M. 1965. Explicit estimates from capture-recapture data with both death and
immigration-stochastic model. Biometrika. 52: 225-247.
Jones III E. L., Quinn IIT. J., and Van Alen, B. W. 1998. Observer accuracy and precision
in aerial and foot survey counts of pink salmon in a southeast Alaska stream. North
American Journal of Fisheries Management, 18: 832-846.
Krebs, C.J. 1999. Ecological Methodology. Addison Wesley Longman, Menlo Park, CA.
Krugman, S.L. and Koemer, T.W. 1969. The effect of cone feeding by Leptoglossus
occidentalis on ponderosa pine seed cone development. Forest Science. 15: 104-111.
Larsson, M. and Franzen, M. 2008. Estimating the population size of specialized solitary
bees. Ecological Entomology. 33: 232-238.
Levia, D.F., Carlyle-Moses, D. and Tanaka, T. 2011. Forest Hydrology and
Biogeochemistry: Synthesis o f Past Research and Future Directions: Volume 216 o f
Ecological Studies. Springer, New York, NY.
Manly, B.F.J.1984. Obtaining confidence limits on parameters of the Jolly-Seber model for
capture-recapture data. Biometrics. 40: 749-758.
Manly, B.F.J., McDonald, T.L. and Amstrup, S.C. K.H. 2005. Introduction to the Handbook.
In Handbook o f Capture-Recapture Analysis. Princeton University Press, Princeton,
NJ. Amstrup S.C., McDonald T.L., and Manley B.F.J., Eds. Princeton University
Press, Princeton, NJ.
Pedigo, L.P. and Rice, M.E. 2009. Entomology and Pest Management. Prentice-Hall, Upper
Saddle River, NJ.
Pojar, J., Klinka, K., and Meidinger, D.1987. Biogeoclimatic ecosystem classification in
British Columbia. Forest Ecology and Management. 22: 119-154.
Pollard, E. 1977. A method for assessing changes in the abundance of butterflies. Biological
Conservation. 12:115-134.
Pollock, K.H., Alpizar-Jara, R. 2005. Classical Open Population Capture-Recapture Models.
In Handbook o f Capture-Recapture Analysis. Princeton University Press, Princeton,
105

NJ. Amstrup S.C., McDonald T.L., and Manley B.F.J., Eds. Princeton University
Press, Princeton, NJ.
R Development Core Team. 2012. R: A language and environment for statistical computing.
R Foundation for Statistical Computing, Vienna, Austria. ISBN3-900051-07-0, URL
http://www.R-proiect.org.
Schnabel, Z.E. 1938. The estimation of a total fish population of a lake. American
Mathematician Monthly. 45: 348-352.
Schumacher, F.X. and Eschemeyer, R.W. 1943. The estimation of fish populations in lakes
and ponds. Journal of the Tennessee Academy of Sciences. 18:228-249.
Seber, G.A.F. 1982. The Estimation o f Animal Abundance and Related Parameters. 2nd ed.
Griffin. Londoa
Southwood, T.R.E. and Henderson, P.A. 2000. Ecological Methods. Chapman and Hall, New
York, NY.
Strong, W.B., Bates ,S.L. and Stoehr, M.U. 2001. Feeding by L. occidentalis
(Hemiptera:Coreidae) reduces seed set in lodgepole pine (Pinaceae). The Canadian
Entomologist 133: 857-865.
Strong, W.B. 2006. Seasonal changes in seed reduction in lodgepole pine cones caused by
feeding of L. occidentalis (Hemiptera:Coreidae). The Canadian Entomologist.
138:888-896.
Swetnam, T.W., and Lynch, A. M. 1993. Multicentury, Regional-Scale Patterns of Western
Spruce Budworm Outbreaks. Ecological Monographs. 63: 399-424.
Thomas, J.A. 1983. A quick method for estimating butterfly numbers during surveys.
Biological Conservation. 27:195-211.
Torgersen, T.R., Colbert, J.J. and Hosman, K.P. 1993. Patterns of occurrence and new
sampling implications for instar IV western spruce budworm (Lepidoptera:
Tortricidae). Forest Science 39: 573-593.
White, G.C. and Burnham, K.P. 1999. Program MARK: survival estimation from
populations of marked animals. Bird Study Supplement. 46: 120-138.
Whitehouse, C., A. Roe, Strong, W.B., Evenden, M. and Sperling, F. 2011. Biology and
management of North American cone-feeding Dioryctria species. The Canadian
Entomologist 143:1 -34.
Williams, B.K., Nichols, J.D. and Conroy, M.J. 2002, Analysis and management of animal
populations. Academic Press, New York, NY.

106

Yao, W.S., Fang, S.Y. and Roques, A. 1991. Specific composition, bio-ecological
characteristics and population dynamics of the larch cone fly (Strobilomyia
spp.;Dipt., Anthomyiidae) complex in the Da Khinggan and Xiao Khinggan
mountains in China. Journal of Applied Entomology. 112: 454-463.
Zoebisch, T.G., Stimac, J.L. and Schuster, D.J. 1993.Methods for estimating adult densities
of Liriomyza trifolii (Diptera: Agromyzidae) in staked tomato fields. Journal of
Economic Entomology. 86:523-528.

107

Chapter 5. General Discussion
Establishing and implementing effective and ecologically sound pest management
strategies requires comprehensive information on the life history of a target pest and the
interactions it has with its environment. Specific considerations include, but are not limited
to knowledge of pest life cycles (Pedigo and Rice 2009), insect migration into and out of
crops systems, pest distribution and dispersion within agricultural systems (Stanton 1983,
Vinatier et al. 2011), factors driving host-plant selection by pest insects (Turgeon et al. 1994,
Bemays and Chapman 1994), pest population abundance, and valid census methods (Haddad
et al. 2008, Larsson and Franzen 2008, Pedigo and Rice 2009). In my thesis, I sought to
address three specific questions that are important for the development of improved
management practices for the conifer seed pest Leptoglossus occidentalis in lodgepole pine
and Douglas-fir. First, I investigated the timing of spring colonization into a lodgepole pine
seed orchard by L. occidentalis and evaluated the spatial distribution ofZ. occidentalis
throughout the spring and early summer to detect potential spatial gradients and edge effects.
I then assessed clonal preferences, and analysed chemical and infrared cues that would
potentially influence host-selection behaviour of L. occidentalis on lodgepole. Finally, I
attempted to evaluate the accuracy of operational walk-through field survey estimates oiL.
occidentalis based on a simple visual monitoring system employed by seed orchard staff.
This was accomplished by comparing the visual monitoring results with population
abundance estimates based on mark-release recapture methods in a lodegpole pine seed
orchard at the Kalamalka Lake Seed Orchard and a Douglas-fir research block at the
Kalamalka Research Station in Vernon B.C.

108

The temporal variability and spatial spread of a pest population may have a
significant impact on population estimates and ultimately, therefore, on pest management
practices (Stanton 1983, Krebs 1999). The movement of L. occidentalis from overwintering
sites into seed orchards in the spring is an important stage o f colonization as the seeds in the
immature cones are highly susceptible to damage at this time (Kovaleski et al. 1999, St.
Toepfer et al. 1999). The migration of L. occidentalis into seed orchards and subsequent
movement through them has not been well studied. In the study described in Chapter 2, L.
occidentalis were detected in lodgepole pine seed orchards in early to mid-May in 2008 and
2009. One of the goals of this study was to determine if previous reports that L. occidentalis
were initially concentrated along an orchard edge could be demonstrated. While I did detect
spatial gradients in the orchard in four of seven surveys conducted in 2008 and in one in
2009, the direction of the spatial gradients that were observed in 2008 were not consistent
throughout the season. The spatial gradients in the orchard may have been easier to detect in
2008 because 65% more L. occidentalis were observed in surveys that year than in 2009. It
is probable that population density, spatial distribution, and behaviour are linked in L.
occidentalis. This phenomenon has been demonstrated for several insect species (Morris
1955, Srivastava et al. 1984, Torgersen et al. 1993). Morever, it can be very difficult to
detect spatial patterns at low insect population densities (Taylor et al 1978). It is possible
that at higher population densities of L. occidentalis a spatial gradient/edge effect might be
detectable at the beginning of the spring migration. Additional studies done at differing
population densities could help determine at what population density an edge effect is
detectable, however without an effective trapping method, such a study would require
extensive labour and may be innacurate. L. occidentalis is a highly mobile insect that
disperses quickly when disturbed. This influences the results of any study that relies on
109

enumerating insects. The developmet of an effective trap for Z. occidentalis should be a
research priority as the development of a functional trap for this insect will reduce current
logistical constraints associated with working with L. occidentalis.
Phenotypic variation in host plants elicit differential responses from phytophagus
insects, and significant impacts on the fitness of an insect and its offspring can result from
differences in susceptibility or suitability of host plants (Fritz and Simmons 1992, Lara and
Tanzini 1997, Cronin et al. 2001). If host plants are not equally suitable for insect growth,
reproduction, and survival, females should select for oviposition those host phenotypes that
confer the greatest fitness to their progeny (Thompson 1988). The cues mediating L.
occidentalis host selection in lodgepole pine are not completely understood, but earlier
research has shown that L. occidentalis exhibits preference for certain clones and particular
physical attributes (Blatt and Borden 1999). In Chapter 3 ,1 confirmed that L. occidentalis do
prefer specific clones, and I established that clone preference remained consistent between
2008 and 2009. I did not find evidence for a specific primary host-selection signal used by
L. occidentalis, but I did ascertain that clone preference classes had significantly different
levels of the monoterpenes a-pinene, 5-3 carene, and p-cymene. I also found that L.
occidentalis preferred clones with cones of greater diameter and weight. Infrared radiation,
which has recently been shown to be significant in host orientation (Tackacs et al. 2009), did
not differ between preferred and non-preferred clones. My results suggest that several
factors, involving multiple sensory modalities, such as cone terpenes, cone infrared profile,
and cone size, appear to be involved in host location and selection by L. occidentalis on
lodgepole pine. Cone and foliage infrared differentials may play a role in medium or longrange host selection, which could be important for locating the host. However, host

110

acceptance by L. occidentalis is likely a function of olfactory and/or gustatory cues. Similar
scale-dependent changes in host orientation cues have been proposed for wood-feeding
insects (Saint-Germain et al. 2007). Further inquiry into the key compounds that render a
tree or clone susceptible or attractive to L. occidentalis is required. There are several
possible avenues for future research, for example clones might be phenotyped according to
susceptibility to pests, and orchards planted utilizing susceptible clones as perimeter trap
crops. Most importantly, identifying and testing the response of L. occidentalis to various
cues could help in the development of a much needed effective trapping system.
Standardized population census methods are essential tools for effective pest
management plans as they can help establish an economic damage threshold for the industry.
Standardized population census methods have been developed for several insect species
(Pollard 1977, Thomas 1983, Haddad et al. 2008, Larsson and Franzen 2009). In this study,
I calculated population abundance but was unable to establish the mathematical relationship
between the population estimates and the results of walk-through surveys required for the
development of standardized census methods (Pedigo and Rice 2009). This is due in large
part to the low numbers of L. occidentalis observed in the field in 2009, the high mobility of
L. occidentalis, and inconsistency in the ability of individual scouts to detect the insect
during walk-through monitoring. In addition to low insect counts, surveys were conducted
on foot in experimental blocks with trees that were up to six metres in height, which
seriously limited the ability of scouts to sight L. occidentalis since it is found more
frequently on cones in the sun than in the shade (Hedlin et al. 1981, personal observation).
Lodgepole pine seed cones are present in greater numbers in the upper crown of trees
(O’Reilly and Owens 1988) and those are more likely to be exposed to the sun. Many

111

studies indicate that several insects have height preferences and are trapped at particular
heights more frequently than others; this may also change with insect densities (Torgersen et
al. 1993, Darbro and Harrington 2006, Epstein et al. 2011, Whitehouse et al. 2011). My
results suggest that existing survey methods are unsatisfactory for management decisions, at
least at low population densities. Further research is required to evaluate the effectiveness of
different methods or patterns of walk-through surveys.
Systematic sampling can be problematic as it can produce biased results and very
narrow confidence limits for population estimates, especially if the sampling intervals
correspond to the distribution of the study organism in its environment (Sutherland 2006).
Nonetheless, systematic sampling methods are sometimes used for walk-through surveys and
may be advantageous in relatively homogenous crop-type systems, particularly when trying
to establish the spatial variation in abundance. It would therefore be useful to determine a
sampling method for L. occidentalis that is effective at different population densities. This
would require a comparison of systematic and other sampling methods with the diagonal
transects discussed in Chapter 4 (Sutherland 2006, Pedigo and Rice 2009). The development
of an effective trap for L. occidentalis might also facilitate the use of alternative, less labourintensive, methods for assessing population. Combined with improved census methods, this
may be sufficient to establish a mathematical relationship between population estimates and
walk-through surveys at different densities, and lead to the development of an economic
damage threshold.
Because it is responsible for substantially reduced seed yields in conifer seed
orchards, Leptoglossus occidentalis continues to be a serious economic pest in British
Columbia and throughout North America (Koerber 1963, Krugman and Koerber 1969, Alter

112

and Sexton 1990, Bates et al. 2000, Bates et al. 2002, Strong 2006). The impact ofZ.
occidentalis is not limited to North America. Leptoglossus occidentalis was first detected in
Europe in Italy in 2001 (Taylor and Villa 2001), and has since been found in several
countries including, but not limited to, Slovakia (Barta 2009), Japan (Ishikawa and Kikuhara,
2009), Norway (Mjos 2010), Turkey (Fent and Kment 2011), and Sweden (Lindelow and
Bergsten 2012). In Europe, L. occidentalis is a serious threat to seed orchards and pine nut
industries (Roversi et al. 2011, Rosenberg et al. 2012, Bracalini et al. 2013). As this pest
spreads and establishes itself into new areas with few natural enemies, but susceptible and
economically valuable crops, it is paramount that we develop effective census techniques,
trapping methods, and monitoring protocols.

113

5.6 Literature Cited
Alter, S. T., and Sexton, J. M. 1990. Effect of Leptoglossus occidentalis (Heteroptera:
Coreidae) on seed development of Douglas-fir at different times during the growing
season in western Oregon. Journal of Economic Entomology. 83:1485-1486.
Barta, M. 2009. New facts about distribution and host spectrum of the invasive nearctic
conifer pest, Leptoglossus occidentalis (Heteroptera: Coreidae) in south-western
Slovakia. Forestry Journal. 55:139-144.
Bates, S. L., Borden, J. H., Kermode, A. R., and Bennett, R. G. 2000. Impact of Leptoglossus
occidentalis (Hemiptera: Coreidae) on Douglas-fir seed production. Journal of
Economic Entomology.93:1444-1451.
Bates, S. L., Strong, W. B., and Borden, J. H. 2002. Abortion and seed set in lodgepole and
western white pine conelets following feeding by Leptoglossus occidentalis
(Heteroptera: Coreidae). Environmental Entomology. 31:1023-1029.
Bemays, E. A., and Chapman, R. F. 1994. Host-plant selection by phytophagous insects.
Springer.
Blatt, S. E. and Borden, J. H. 1999. Physical characteristics as potential host selection cues
for Leptoglossus occidentalis (Heteroptera: Coreidae). Environmental Entomology.
28:246-254.
Bracalini, M., Benedettelli, S., Croci, F., Terreni, P., Tiberi, R , and Panzavolta, T. 2013.
Cone and Seed Pests oiPinus pinea: Assessment and Characterization of Damage.
Journal of Economic Entomology. 106:229-234.
Cronin, J. T., Abrahamson, W. G., and Craig, T. P. 2003. Temporal variation in herbivore
host-plant preference and performance: constraints on host-plant adaptation. Oikos,
93:312-320.
Darbro, J.M. and Harrington, L.C. 2006. Bird-baited traps for surveillance of West Nile
mosquito vectors: effect of bird species, trap height, and mosquito escape rates.
Journal of Medical Entomology. 43:83-92.
Epstein, D.L., Stelinski, L.L., Miller, J.R., Grieshop, M.J.and Gut, L.J. 2011. Effects of
reservoir dispenser height on efficacy of mating disruption of codling moth
(Lepidoptera: Tortricidae) in Apple Pest Management Science. 67:975-979.
Fent, M., and Kment, P. 2011. First record of the invasive western conifer seed bug
Leptoglossus occidentalis (Heteroptera: Coreidae) in Turkey. North-Western Journal
of Zoology. 7:72-80.

114

Fritz R.S. and Simms E.L (Eds). 1992. Plant Resistance to Herbivores and Pathogens:
ecology evolution and genetics. University of Chicago Press.
Haddad, N. M., Hudgens, B., Damiani, C., Gross, K., Kuefler, D., and Pollock, K. 2008.
Determining optimal population monitoring for rare butterflies. Conservation
Biology. 22:929-940.
Ishikawa, T., and Kikuhara, Y. 2009. Leptoglossus occidentalis Heidemann (Hemiptera:
Coreidae), a presumable recent invader to Japan. Japanese Journal of
Entomology. 12:115-116.
Koerber, T. W. 1963. Leptoglossus occidentalis (Hemiptera, Coreidae), a newly discovered
pest of coniferous seed. Annals of the Entomological Society of America. 56:229234.
Kovaleski, A., Sugayama, R. L., and Malavasi, A. 1999. Movement of Anastrepha
fraterculus from native breeding sites into apple orchards in Southern Brazil.
Entomologia Experimentalis et Applicata. 91:459-465.
Krebs C.J. 1999. Ecological Methodology. Addison Wesley Longman, Menlo Park, CA.
Krugman, S. L., and Koerber, T. W. 1969. Effect of cone feeding by Leptoglossus
occidentalis on ponderosa pine seed development. Forest Science. 15:104- 111.
Lara, F.M., and Tanzini, M. R. 1997. Nonpreference of the lace bug Leptopharsa heveae
Drake & Poor (Heteroptera: Tingidae) for rubber tree clones. Anais da Sociedade
Entomologica do Brasil. 26: 429-434.
Larsson, M., and Franzen, M. 2008. Estimating the population size of specialised solitary
bees. Ecological Entomology. 33:232-238.
Lindelow, A., and Bergsten, J. 2012. The invasive western conifer seed bug, Leptoglossus
occidentalis (Heteroptera: Coreidae), established in Sweden. Entomologisk Tidskrift.
133:55-58.
Mjos, A. T., Nielsen, T. R., and 0degaard, F. 2010. The western conifer seed bug
(Leptoglossus occidentalis Heidemann, 1910) (Hemiptera, Coreidae) found in SW
Norway. Norwegian Journal of Entomology. 57:20-22.
Morris, R. F.1955. The development of sampling techniques for forest insect defoliators,
with particular reference to the spruce budworm. Canadian Journal of Zoology.
33:225-294.
O'Reilly C., and Owens, J. N.1988. Reproductive growth and development in seven
provenances of lodgepole pine. Canadian Journal of Forest Research, 18:43-53.

115

Pedigo, L. P., Rice, M.E. 2009. Entomology and pest management (No. Ed. 6). Prentice-Hall
International.
Pollard, E. 1977. A method for assessing changes in the abundance of butterflies. Biological
Conservation. 12:115-134.
Rosenberg, O., Ylioja, T., Ravn, H. P., Krokene, P., and Voolma, K. 2012. Valuable seed
destroyed by insects. Scandinavian Journal Forest Research News & Views 1: 100101 .

Roversi, P. F., Strong, W. B., Caleca, V., Maltese, M., Sabbatini Peverieri, G., Marianelli, L.
and Strangi, A. 2011. Introduction into Italy of Gryon pennsylvanicum (Ashmead), an
egg parasitoid of the alien invasive bug Leptoglossus occidentalis Heidemann. EPPO
Bulletin. 41:72-75.
Saint-Germain, M., Buddie, C. M., and Drapeau, P. 2007. Primary attraction and random
landing in host-selection by wood-feeding insects: a matter of scale? Agricultural and
Forest Entomology 9: 227-235.
Srivastava, N., Campbell, R. W., Torgersen, T. R., and Beckwith, R. C.1984. Sampling the
western spruce budworm: fourth instars, pupae, and egg masses. Forest Science. 30:
883-892.
Stanton, M.L.1983. Spatial patterns in the plant community and their effects upon insect
search. In “Herbivorous Insects”. S. Ahmad, (ed.)., pp. 125-157. Academic Press,
New York.
St. Toepfer, G.H., and Dom, S. 1999. Spring colonisation of orchards by Anthonomus
pomorum from adjacent forest borders. Entomologia Experimentalis et Applicata,
93:131-139.
Strong, W. B. 2006. Seasonal changes in seed reduction in lodgepole pine cones caused by
feeding of Leptoglossus occidentalis (Hemiptera: Coreidae).The Canadian
Entomologist. 138: 888-896.
Sutherland, W. J. (Ed.). 2006. Ecological census techniques: a handbook. Cambridge
University Press.
Takacs, S., Bottomley, H., Andreller, I., Zaradnik, T., Schwarz, J., Bennett, R., and Gries, G.
2009. Infrared radiation from hot cones on cool conifers attracts seed-feeding insects.
Proceedings of the Royal Society B: Biological Sciences. 276: 649-655.
Taylor, L. R., Woiwod, I. P., and Perry, J. N. 1978. The density-dependence of spatial
behaviour and the rarity of randomness. The Journal of Animal Ecology. 47: 383406.

116

Taylor, S. J., Tescari, G., and Villa, M. 2001. A Nearctic pest of Pinaceae accidentally
introduced into Europe: Leptoglossus occidentalis (Heteroptera: Coreidae) in
northern Italy. Entomological News. 112:101-103.
Thomas, J.A. 1983. A quick method for estimating butterfly numbers during surveys.
Biological Conservation. 27:195-211.
Thompson, J.N. 1988. Evolutionary ecology of the relationship between oviposition
preference and performance of offspring in phytophagus insects. Entomologia
Experimentalis et Applicata. 47:3-14.
Torgersen, T. R., Colbert, J. J., and Hosman, K. P. 1993. Patterns of occurrence and new
sampling implications for instar IV western spruce budworm (Lepidoptera:
Tortricidae). Forest Science. 39: 573-593.
Turgeon, J.J., Roques, A., and de Groot, P. 1994. Insect fauna of coniferous seed cones.
Annual Review of Entomology. 39: 179-212.
Vinatier, F., Tixier, P., Duyck, P. F., and Lescourret, F. 2011. Factors and mechanisms
explaining spatial heterogeneity: a review of methods for insect populations. Methods
in Ecology and Evolution. 2:11-22.
Whitehouse, C., A. Roe, Strong, W.B., Evenden, M. and Sperling, F. 2011. Biology and
management of North American cone-feeding Dioryctria species. The Canadian
Entomologist. 143:1-34.

117

Appendix I
Table AI.l. Results of logistic regression models used to determine spatial trends in 2008
s u r v e y d a te

m odel

Zgp

pm

Z g ,x

Peix2

Z b i:

P bix

Z girv

Psim,

May 5, 2008

null

-

-

-

p BIx_______Z g2x2
-

-

-

-

-

-

-

-

*

-

-

-

-

-

-

-

-

-

-

X

May 17, 2008

y
x + x2
x*y
null

-

-

-

-

-

-

-

-

-

-

<0.01
<0.01
0.09

-

-

-

-

-

-

-

-

0.68

0.50

-

-

-

-

-

-

y

-7.49
-2.98
0.01

-

-

-

-

-0.02

0.99

-

-

18.98
20.44
13.63

x + x2

-0.13

0.09

0.13

0.90

0.90

-

-

-

-

15.30

**y

<0.01

0.09

<0.01

0.99

-

-

-0.01

0.99

<0.01

1.00

17.08

<0.01

-

-

-

-

-

-

-

-

93.62

<0.01
<0.01
<0.01

-0.49

0.62

-

-

-

-

-1.17

0.24

<0.01

0.94

0.09

<0.01

X

May 23,2008

null
X

y
x+x2
x*y
May 28, 2008

null
X

y

x + x2
x *y
June 4, 2008

null
X

y

x + x2
x*y
June 10,2008

null
X

June 20, 2008

14.63
-7.45
-5.57
-5.49
-3.79
26.03
13.09
12.97
-8.92
-6.38
24.86
12.56
12.23
-8.18

-6.66
27.75
14.43

0.13

-

-

0.29
-

-2.64

0.01

0.34

1.07
-

-0.79

0.43

-1.67

-

-

-

-

-

-

-

-

567.48

<0.01

0.39

0.70

-

-

-

-

-

-

569.33

<0.01

-

-

-

-

-0.20

0.84

-

-

569.44

<0.01
<0.01

0.01
-0.07

0.99
0.95

0.09
-

0.93
-

-0.37

0.71

0.31

0.76

571.32
573.19

<0.01

-

-

-

-

-

-

-

-

465.69

<0.01

2.22

0.03

-

-

-

-

-

-

462.62

<0.01

-

-

-

-

-1.05

0.30

-

-

466.59

<0.01
<0.01

0.49

0.63
0.04

0.04

0.97
-

0.63

-

-

0.48

-

2.02

-1.03

0.30

464.62
464.45

<0.0

-

-

-

-

-

-

-

-

834.87

<0.01

1.80

0.07

-

-

-

-

-

-

833.58

<0.01

-

-

-

-

2.23

0.03

-

-

831.83

0.03

-

-

-

-

830.78

-

1.69

0.09

-0.75

0.46

831.98

95.37
75.10
96.26
75.33

y

14.53

x + x2

-9.99

<0.01

2.46

0.01

x*y

-7.78

<0.01

1.48

0.14

2.13
-

<0.01

-

-

-

-

-

-

-

-

1309.70

<0.0

3.51

<0.01

-

-

-

-

-

-

1299.10

<0.01

-

-

-

-

0.93

0.35

-

-

1310.80

<0.01

3.75

<0.01

0.00

-

-

-

-

1291.00

<0.01

1.36

0.17

-

0.06

0.95

0.41

0.68

1302.10

null
X

y
x + x2
x*y

27.47
15.42
14.22
11.13
-7.66

118

3.08
-

Table AI.2. Results of logistic regression models used to determine spatial trends in 2009
survey date_________ m odel

May 8, 2009

null
y

-12.19
-5.90
-5.83

<0.01
<0.01
<0.01

x + x2

-1.66

<0.01

x*y
null

-2.82
-17.27
-8.72
-8.72

<0.01
<0.01
<0.01
<0.01

x + x2

0.99

<0.01

1.13

0.26

x*y
null

<0.01
<0.01
<0.01
<0.01

-0.36

y

-4.37
-16.82
-8.44
-8.54

0.72
0.88
-

x + x2

-4.92

<0.01

1.75
-0.39

y

-4.15
-19.67
-9.86
-9.85

<0.01
<0.01
<0.01
<0.01

x+x2

-6.79

<0.01

x*y
null
y

-4.93
-14.63
-5.95
-7.61

<0.01
<0.01
<0.01
<0.01

x + x2

-2.23

<0.01

x*y
null
y

-3.29
-20.70
-10.36
-10.34

<0.01
<0.01
<0.01
<0.01

x+x2

-6.92

<0.01

x*y
null
y

-4.95
-22.95
-11.43
-10.96

<0.01
<0.01
<0.01
<0.01

x + x2

-7.12

<0.01

2.13

0.03

x*y

-5.47

<0.01

0.70

0.49

X

May 15,2009

X

y

May 23,2009

X

May 28, 2009

x*y
null
X

June 2, 2009

X

June 13,2009

X

June 18,2009

Z m_________Peo
p glx—
IX
----um ....... a~Z'Deix_____

X

0.38

0.71

1.41

0.16

0.27

0.79
0.57

-

-

-0.57

-

-

-0.15
-

-0.98
-

Z Mx2

-

-

-

-

-

-

-

-

0.20

-

1.29
-

1.82

0.07

-

-

-

1.47

0.14

-0.20

0.84

-

-

1.12

-

1.30

0.20

-0.57

0.57

-

-

-

0.12

-

-

1.57

-

-

119

1.39

1.90

-

-

53.64

0.74
-

-

0.90
-

0.56

-

-

64.44
146.09
147.77
147.74

-

-

-0.12

-

147.82

0.95
-

151.42
136.01
137.99
136.90

-

135.39

-

0.71
-

140.74
212.61
213.64
212.98

-

-

-

213.74

0.55
0.25

-

0.91
-

-

-

216.00
93.62
90.33
94.21

-

-

-

90.35

0.69

0.71
-

0.06

-

-0.83

-1.26

-0.60
-1.15

0.26

-

-

-

0.17

-

-

-

0.06

-

-

-

0.41
0.21

-

-

-

-

0.19
-

-

-

1.30

-

58.84
60.69
60.59

0.30

-1.04

0.03

-0.91

-0.69

-

-

-0.11

-0.41

0.38

-

-

92.75
248.11
250.07
249.34

-

-

-

249.95

0.58

0.36

0.21

-

-

253.00
346.89
346.39
347.30

-

-

-

344.28

0.88

348.77

-1.25

0.37

0.98

-0.87

A1C

-

0.63

-

-

-

-

-

-

-

0.21

-0.37

Z Blxv------ fMS/
p elxv

.
-

-0.59

-

0.65
0.04

1.25

0.33

-

-0.46
2.03

0.49

1.42
-

0.30

-

-

-

1.04

-

Zg/z— .... t'Di:
p e,.

0.16

0.08
0.70
0.33
-

Psix2

0.49

0.55

0.16

Table AI.3. Chi-squared test statistics and AIC values used to select best fit spatial point process
models for 2008 and 2009 surveys__________________________________________________
survey date 2008

model

X*

p>X2

AIC

survey date 2009

model

2

p>X2

AIC

05-May-08

X

-

-

-

08-May-09

X

0.143

0.706

86.294

y

-

-

-

y

0.240

0.624

86.196

x*y

-

-

-

x*y

0.383

0.826

88.054

null

-

-

-

null

-

-

84.437

X

0.542

0.462

26.840

X

0.233

0.629

223.559

y

7.448

0.006

19.934

y

0.290

0.590

223.502

x*y

7.768

0.021

21.614

x*y

0.523

0.770

225.269

null

-

-

25.382

null

-

-

221.792

X

0.249

0.618

140.180

X

0.022

0.883

206.446

y

20.412

0.000

120.018

y

1.103

0.294

208.424

x*y

20.659

0.000

121.771

x*y

1.125

0.570

207.343

null

-

-

138.430

null

-

.

206.446

X

0.145

0.703

994.016

X

0.960

0.327

335.405

y

0.040

0.842

994.122

y

1.609

0.205

334.756

x*y

0.185

0.912

995.977

x*y

2.569

0.277

335.795

null

-

-

992.161

null

-

-

334.365

X

4.942

0.026

790.481

X

4.908

0.027

135.991

y

1.066

0.302

794.358

y

1.535

0.215

138.430

x*y

6.008

0.050

791.415

x*y

6.439

0.040

135.521

null

-

-

793.423

null

-

-

138.894

x

3.085

0.079

1550.774

X

0.039

0.843

397.508

y

4.752

0.029

1549.106

y

0.760

0.383

396.788

x *y

7.835

0.020

1548.024

x*y

0.799

0.671

398.748

null

-

-

1551.859

null

-

-

395.548

X

11.161

0.001

2689.242

X

1.965

0.161

557.222

y

0.773

0.379

2699.630

y

1.403

0.236

557.784

x*y

11.933

0.003

2690.469

x*y

3.368

0.186

557.819

null

-

-

2698.403

null

-

-

557.187

17-May-08

23-May-08

28-May-08

04-Jun-08

10-Jun-08

20-Jun-08

120

17-May-09

23-May-09

28-May-09

02-Jun-09

13-Jun-09

18-Jun-09

Appendix II
Table AIL 1: Results of mixed effects analyses of variance for terpenoids from all sampling
events 2008 and 2009 in seed orchard 307 at Kalamalka Lake Seed orchard in Vernon B.C.,
Canada.

________ 22 July 2008
Source o f

26-28 June2009

15-17 July
16-18 August
2009____________2009________

Compound

variation

F 2.46

P

F2.38

P

F 2,38

P

F2.38

P

a-pinene

clone preference class

2.068

0.138

4.015

0.026

3.897

0.029

3.304

0.048

camphene

clone preference class

4.565

0.016

0.193

0.825

1.763

0.185

0.140

0.870

P-pinene

clone preference class

3.577

0.036

1.568

0.222

0.114

0.893

0.715

0.496

sabinene

clone preference class

9.216

<0.001

1.191

0.315

0.139

0.871

0.006

0.994

8-3-carene

clone preference class

3.998

0.025

11.624

<0.001

8.536

<0.001

7.715

0.002

myrcene

clone preference class

0.009

0.991

1.381

0.264

1.522

0.231

0.626

0.540

a-phellandrene

clone preference class

6.896

0.002

1.370

0.266

0.195

0.823

0.161

0.852

r-limonene

clone preference class

7.970

0.001

9.543

<0.001

7.327

0.002

4.024

0.026

P-phellandrene

clone preference class

14.860

<0.001

0.850

0.436

1.200

0.312

0.336

0.717

y-terpinene

clone preference class

0.560

0.575

0.345

0.710

n/a

n/a

0.763

0.473

p-cymene

clone preference class

17.404

<0.001

0.821

0.448

3.466

0.041

0.635

0.536

terpinolene

clone preference class

3.415

0.041

2.875

0.069

1.177

0.319

3.457

0.042

camphor

clone preference class

2.208

0.122

0.873

0.426

2.166

0.129

2.259

0.118

linalool

clone preference class

0.181

0.835

n/a

n/a

n/a

n/a

n/a

n/a

bomyl acetate

clone preference class

10.868

<0.001

2.286

0.116

0.678

0.514

2.276

0.117

P-caryphyllene

clone preference class

8.145

<0.001

0.750

0.479

1.548

0.226

0.821

0.448

pulgeone

clone preference class

0.164

0.849

n/a

n/a

1.762

0.185

2.257

a-caryophyllene

clone preference class

2.860

0.068

0.267

0.767

0.019

0.981

0.329

0.722

a-terpineol

clone preference class

2.569

0.088

n/a

n/a

n/a

n/a

n/a

n/a

bomeol

clone preference class

2.094

0.135

1.393

0.261

n/a

n/a

0.238

0.790

geranyl acateate

clone preference class

3.170

0.051

8.728

<0.001

n/a

n/a

n/a

n/a

121

Appendix III
Jolly-Seber Method

All animals captured in the first sample are by definition unmarked and in all following
samples they are either marked or unmarked. Jolly (1965) defined the following variables,
which are scored in a “Method B” table as described by Leslie and Chitty (1951) (from Krebs
1999):
m, = the number of marked animals caught in sample t
u, = the number of unmarked animals caught in sample t
n, = the total number of animals caught in sample t
= mt + nt
st = total number of animals released after sample t
= (n, - accidental deaths or removals)
mr, = the number of marked animals caught in sample t as caught in sample r
Rt = the number of st individuals released at sample t and caught later in another sample
Z, = the number of individuals marked before sample t not caught in sample t but caught in
some sample after sample t

122

The population size is estimated using Jolly’s (1965) equation;

Population size=

Size of marked population
Proportion of animals marked

(4.1)

The proportion of animals marked (a,) is estimated as;
at = mt + 1
nt + 1

(4.2)

Where the “+1” is a correction for bias in small sample sizes (Seber 1982);
The estimated size of the marked population (M, ) is estimated by (Seber 1982);
Mt = (St+ l)Z t + m,
R,+ 1

(4.3)

The estimated population size just before time t is;
Nt = Mi
at

(4.4)

Manly’s (1984) method of upper and lower confidence limits for population size were
calculated using the following transformation;
T,(N0 = loge (Nj + loge [ l - ( p t/ 2 ) + V(1 —p t )/2]

(4.5)

Where,
Pt =
Nt

Total caught at time t
Estimated population size at t

123

(4.6)

The estimated variance of the transformation is;
Var\T<(m-\ - ( M ,-m ,+ s,+ l ) f 1
Mt + 1
Rt + 1

-

1 )+
1 + 1
st + 1
mt +1 nt + 1

(4.7)

The upper and lower 95% confidence intervals for 7/ are calculated as follows;
7/w

= T, (N<) - 1.6 ^Var [UN,)]

(4.8)

TI upper

T, ( N ,) - 1.6 ^Var[T,(Nt)]

(4.9)

The confidence intervals for population size are then calculated from;
(\L + n,)2 <N,< (AU + n,)2
16Z
16C7

(4.10)

with;
L = elL
U = ew
e = base of natural logarithms

The Schnabel Method
Schnabel (1938) defines the following variables for calculations using two or
more sampling times (from Krebs 1999);
Ci = the total number of animals caught in sample t
R, = the number of individuals already marked when caught in sample t
Ut = the number of individuals marked for the first time and released in sample t
Provided the number is small, accidental deaths of marked or unmarked animals are
subtracted from U, then;
C, = R, + Ui

(4.11)

Marked animals in the population increase with the number of samples taken and;
124

M t = the number of marked animals in the population right before sample t.

M, = X U,

"

(4.12)

Population size (Nt) is estimated by;
N, = Y A C ,M ^

(4.13)

X'R'+l
The variance of the reciprocal of the population estimate is calculated by;
Variance (1) = Y R>
N
(Z C ,M ,f

(4.14)

with standard error;
Standard error (J_) =V [Variance (1)]
N
N

(4.15)

The total number of recaptures was less than 50 so upper and lower Poisson confidence
intervals for X Rt were obtained from tables (in Krebs 1999) and plugged into equation (4.13)
to obtain the confidence intervals for population size estimates.

The Schumacher-Eschemeyer Method
Using the variables defined for the Schnabel estimate above, Schumacher and
Eschemeyer (1943) estimated population size (N,) as;

Nt = X(CtM,2)
I*/

(4.16)

I (RMd

i-l

125

Variance is defined as;
Variance (I) = Y (R,2/C,) - ICFRMtflYCM?) 1
s- 2

m

(4.17)

The confidence interval for the Eschemeyer-Schumacher method is calculated;
1 ± ra S.E.
N

(4.18)

where S.E. is from equation 4.15 and ta from Student’s t- table for (100-a) confidence
interval with s- 1 degrees of freedom.

126

Appendix IV
Table AIV.l. Comparison of Number of L. occidentalis sighted by each scout in 2008 in
orchard 307 at Kalamalka Seed Orchard Vernon B.C. Canada. Survey dates included are
those where scouts spent an equal amount of time surveying.
Number of L. occidentalis sighted
Date

Tamara

Sydney

17 May, 2008

1

0

23 May, 2008

5

4

28 May, 2008

39

35

4 June, 2008

28

24

10 June, 2008

64

68

Table AIV.2. Comparison of Number of L. occidentalis sighted by each scout in 2009 in
orchard 307 at Kalamalka Seed Orchard Vernon B.C. Canada . Survey dates included are
those where scouts spent an equal amount of time surveying.
Number of L. occidentalis sighted
Date

Tamara

Nicole

8 May, 2009

5

0

15 May, 2009

12

2

23 May, 2009

6

5

28 May, 2009

12

9

3 June, 2009

7

0

13 June, 2009

20

10

127