DISPERSAL AND FLIGHT BEHAVIOR OF TRYPODENDRON

LINEATUM (OLIVIER) (COLEOPTERA: SCOLYTIDAE) AS INFLUENCED

BY SEMIOCHEMICAL AND ENVIRONMENTAL FACTORS

By

SCOTT MICHAEL SALOM

B.S., Iowa State University, 1981 M.S., University of Arkansas, 1985

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF

THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

in

THE FACULTY OF GRADUATE STUDIES

(Department of Forest Sciences)

We accept this thesis as conforming

to the required standard

THE UNIVERSITY OF BRITISH COLUMBIA

April 7, 1989

© Scott Michael Salom, 1989 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.

Department of

The University of British Columbia Vancouver, Canada

Date

DE-6 (2/88) ii

ABSTRACT

Laboratory bioassays using a wind tunnel were developed to study flight behavior and orientation of the striped ambrosia beetle, Trvpodendron lineatum (Olivier). Factors that were studied in detail were windspeed, semiochemical concentrations, and semiochemical combinations. In the first of two experiments examining the effect of wind speed on T. lineatum response to a semiochemical-baited substrate, the highest % of males (21.4) and females (25.3) caught in a funnel trap, occurred at 0.0 m/s. As windspeed was increased from 0.0 to 0.9 m/s, the percent of beetles caught decreased linearly for both sexes. A second experiment showed that in the presence of wind, responding beetles oriented anemotactically to the semiochemical-baited substrate. With wind absent, beetles flew randomly and erratically. Upon reaching close to the baited substrate, a greater proportion of the beetles responded to the chemical stimuli and landed on the substrate than when an airflow was present. These results suggest that T. lineatum are capable of responding under varied wind conditions typically present in a forest, whereby they use wind to orient to olfactory stimuli, yet are best arrested to the stimuli under still conditions.

Flight response of T. lineatum to a multiple funnel trap baited with ethanol (1° attractant) and lineatin (2° attractant) at different release rates, indicated that only lineatin was effective in attracting beetles to this kind of

trap. Maximum response by both males and females occurred

at release rates of lineatin between 8 and 64 ug/24 h.

However, in a more detailed study of T. lineatum response to

ethanol and lineatin using modified drainpipe traps, serving

as a model of a host tree, ethanol did positively influence

male flight type, speed of reaction, and direction.

Nevertheless, lineatin was the most important semiochemical

in attracting males to land on and enter the traps. Ethanol

was more important for females than for males, and when

combined with lineatin, provided the optimal stimuli for

attracting females to land on and enter the traps.

Population movement of spring dispersing T. lineatum

was studied using mark-recapture techniques with lineatin-

baited funnel traps. In a first set of studies conducted in

an even-aged second-growth coastal forest in British

Columbia, beetle recapture distribution was compared with wind direction at distances between 5 and 500 m from the

beetle release site. At 5 and 25 m, beetle recapture was

predominantly upwind. With traps placed only at 100 m from

the release site, beetles were recaptured in all directions

irrespective of wind. However, with traps placed only 500 m

from the release site, beetles were only recaptured in the

downwind traps.

In mark-recapture experiments conducted in a valley,

beetles released from a forest margin influenced by

prevailing up-valley winds, flew upwind within the forest to iv

lineatin-baited funnel traps placed 25 m from the release

site. Beetle recapture in an open setting was higher along

the edge of the open setting than in its center, 325 m

closer to the release site. Beetles were recaptured 1 km

down-valley (upwind) and 1.9 km up-valley (downwind) from

the release site. In one experiment (two releases), 10.6

and 7.8% of the marked beetles recaptured were collected in

traplines > 700 m and > 1 km from the release site,

respectively.

In additional mark-recapture experiments in the valley,

beetles were released simultaneously from a windward and

leeward side of a forest margin in the valley through two

experiments of four releases each. With long distance

flight emphasized and no semiochemical-baited traps placed within 200 m of either release site, population movement was

predominantly downwind. Beetles also flew across the valley

to traps on the opposite facing slope at a fairly high

frequency (38% of the recaptured beetles), during the first

experiment. Beetles were recaptured at a much higher

frequency in traps placed within a forest as compared to

those in an open setting. This was likely a result of the

calmer wind conditions under the forest canopy, facilitating

better flying conditions and response to olfactory stimuli

for the beetles.

The implications these findings have on the general

knowledge of scolytid beetle dispersal and orientation to

olfactory stimuli are discussed. New considerations toward V improving pest management strategies for T. lineatum as a result of these sets of studies are presented. yi

TABLE OF CONTENTS

Page

ABSTRACT ii

TABLE OF CONTENTS vi

LIST OF TABLES ix

LIST OF FIGURES xii

ACKNOWLE DGEMENTS xv i i i

CHAPTER 1: OVERVIEW 1

1.1 Introduction 1

1.2 Objectives 7

1.3 Format 10

CHAPTER 2: THE EFFECT OF WIND SPEED ON TRYPODENDRON LINEATUM FLIGHT BEHAVIOR IN A WIND TUNNEL 12

2.1 Introduction 12

2.2 Materials and Methods 14

2.3 Results 21

2.4 Discussion 32

CHAPTER 3: RESPONSE OF TRYPODENDRON LINEATUM TO VARIED RELEASE RATES OF ETHANOL AND LINEATIN IN A WIND TUNNEL 39

3.1 Introduction 39

3.2 Materials and Methods 41

3.3 Results 44

3.4 Discussion 46

CHAPTER 4: WIND TUNNEL EVALUATION OF FLIGHT AND LANDING BEHAVIOR OF TRYPODENDRON LINEATUM (OLIVIER) IN RESPONSE TO A HOST ATTRACTANT AND AGGREGATION PHEROMONE 49

4.1 Introduction 49 vii

4.2 Materials and Methods 51

4.3 Results 56

4.4 Discussion 69

CHAPTER 5: INFLUENCE OF WIND ON THE SPRING FLIGHT OF TRYPODENDRON LINEATUM IN A SECOND- GROWTH CONIFEROUS FOREST 74

5.1 Introduction 74

5.2 Materials and Methods 75

5.3 Results and Discussion 82

CHAPTER 6: DISPERSAL OF TRYPODENDRON LINEATUM

(OLIVIER) WITHIN A VALLEY SETTING 96

6.1 Introduction 96

6.2 Materials and Methods 97

6.3 Results and Discussion 106

CHAPTER 7: POPULATION MOVEMENT PATTERNS OF TRYPODENDRON LINEATUM AS INFLUENCED BY

WIND AND VEGETATION WITHIN A VALLEY 12 9

7.1 Introduction 129

7.2 Materials and Methods 13 0

7.3 Results and Discussion 139

CHAPTER 8: CONCLUSIONS 161

REFERENCES 170 APPENDIX 1: RESPONSE OF TRYPODENDRON LINEATUM (OLIVIER) TO DIFFERENT SEMIOCHEMICAL- BAITED SUBSTRATES IN A WIND TUNNEL 181 APPENDIX 2: RESPONSE OF FLOWN AND UNFLOWN T. LINEATUM TO SEMIOCHEMICAL-BAITED FUNNEL TRAPS IN A WIND TUNNEL 184

A. 2.1 Introduction 184

A.2.2 Materials and Methods 185

A.2.3 Results and Discussion 189

APPENDIX 3: INFLUENCE OF DUSTING T. LINEATUM WITH viii

FLUORESCENT POWDER ON THEIR RESPONSE TO SEMIOCHEMICALS IN A WIND TUNNEL 193 ix

LIST OF TABLES

Page

Table 1. Frequency of Trypodendron lineatum flight in a wind tunnel at varied wind speeds in the presence of an unbaited and semiochemical-baited simulated log 24

Table 2 Statistical comparison of T. lineatum responses to first and second year control treatments in a wind tunnel. The numbers of the beetles released for each sex/ treatment = 124 57

Table 3. Flight frequency of T. lineatum to different semiochemical baits in a wind tunnel. Number of beetles released for each sex in each bait treatment = 124. 58

Table 4. The capture of T. lineatum in semiochemical-baited drainpipe traps in a wind tunnel (N = 124 released/treatment/ sex) 68

Table 5, Numbers of marked T. lineatum released and recaptured in the flight direction study on the University Endowment Lands, Vancouver, B.C., 1986 84

Table 6, Relationship of flight direction of T. lineatum to the wind direction, using a one-sample test for the mean directional flight angle on the University Endowment Lands, Vancouver, B.C., 1986 87

Table 7, Percent of T. lineatum caught at varied distances in the mark-recapture study, experiments 1 and 2, University Endowment Lands, Vancouver, B.C., 1986 90

Table 8. Comparison of wind direction and speed between forested and open sites within the Cedar Creek Valley (up-valley azimuth of 48 to 68°) in the Coquitlam Lake Watershed B.C., 1987 108

Table 9. Flight and recapture success of T. lineatum in the mark-recapture study at the Coquitlam Lake Watershed, B.C., 1987 (Day 1 and Day 2 data were pooled). 110

Table 10. The number of marked T. lineatum recaptured with respect to time in a X

dense old growth forest in the Coquitlam Lake Watershed, B.C., 1987 Ill

Table 11, Mean percent of T. lineatum recaptured at varied distances in the Cedar Creek Valley, Coquitlam Lake Watershed, B.C., 1987 , 113

Table 12 Comparison of T. lineatum flight and wind direction in an old growth forest in the Coquitlam Lake Watershed, B.C., 1987 114

Table 13 Weather measurements made between 113 0 and 1700 h PDT, from the Cedar Creek Valley in the Coquitlam Lake Watershed in British Columbia during 1988 141

Table 14, Flight and recapture success of Trypodendron lineatum in the mark- recapture study conducted at the Coquitlam Lake Watershed in British Columbia during 1988 148

Table 15, Comparison between up- and down-valley movement of adult Trypodendron lineatum during the mark-recapture study conducted in the Cedar Creek Valley within the Coquitlam Lake Watershed in British Columbia during 1988 150

Table 16, Movement of T. lineatum through a forest, flying up-valley (population A) and down- valley (population B), during four separate releases in experiment 2, within the Coquitlam Lake Watershed in British Columbia during 1988 154

Table 17 Comparison between an open and forested route for long distance flying Trypodendron lineatum, during 4 releases in experiment 2, within the Coquitlam Lake Watershed in British Columbia during 1988 , 156

Table 18, Mean recapture percentage of T. lineatum to different substrates baited with semiochemicals in a wind tunnel. Six replicates of 50 beetles, for a total of 3 00 beetles were released for each treatment by sex category 183

Table 19 Response of flown and unflown T. lineatum to semiochemical-baited funnel traps in a xi

wind tunnel 190

Table 20. The effect of marking treatments on T. lineatum response to a semiochemical- baited funnel trap in a wind tunnel 195 xii

LIST OF FIGURES

Page

Fig. 1 Body orientation directions of T. lineatum scored prior to and during flight take-off, as viewed from above the release tray (top); and flight take-off directions of the same beetles, as viewed from the side (bottom). All beetles taking off that did not fit into categories A to D were placed in the no-direction category. ...2 0

Fig. 2 The influence of wind speed on response of male (top) and female (bottom) T. lineatum to a semiochemical-baited funnel trap in a wind tunnel. Numbers next to boxes represent the number of data points. White and black boxes represent the % response for the baited and unbaited (control) treatments, respectively. Missing data accounts for some wind speeds having less than 9 observations 2 2

Fig. 3 Body orientation of T. lineatum, prior to and during flight take-off, at different wind speeds, in the presence of an unbaited and semiochemical-baited simulated log. Male and female data were pooled. Bars with different letters for each response by bait treatment category are significantly different (X ; p < 0.05) for individual comparisons 25

Fig. 4 Flight take-off direction of T. lineatum. at different wind speeds, in the presence of an unbaited and semiochemical- baited simulated log. Male and female data were pooled. Bars with different letters for each response by bait treatment category are significantly different (X2; p < 0.05) for individual comparisons 2 6

Fig. 5 Overall flight direction exhibited by T. lineatum, at different wind speeds, in the presence of an unbaited and semiochemical-baited simulated log. Male and female data were pooled. Bars with different letters for each response by bait treatment category are significantly different (X2; p < 0.05) for individual comparisons 28 xiii

Fig. 6 Type of flight exhibited by T. lineatum, at different wind speeds, in the presence of an unbaited and semiochemical-baited simulated log. Bars with different letters for each response by bait treatment by sex category are significantly different (X2; p < 0.05) for individual comparisons 2 9

Fig. 7 Positive lock-on response exhibited by T. lineatum. at different wind speeds, in the presence of an unbaited and semiochemical-baited simulated log. Male and female data were pooled. Bars with different letters for each response by bait treatment category are significantly different (X2; p < 0.05) for individual comparisons 31

Fig. 8 Mean percent (+) S.E. of T. lineatum caught in traps in response to combinations of varied release rates of lineatin and ethanol in a windtunnel: A) Males and B) Females. Lineatin release rate columns, pooled across ethanol treatments, with the different letters are significantly different (SNK; p < 0.05) 45

Fig. 9 A) Modified Norwegian drainpipe traps used as a baiting substrate for T. lineatum, with a typical smoke plume pattern produced at 0.15 m/s. B) Close-up of trap cover and lure holder, showing an upper and lower lineatin bait, and a medial ethanol bait 53

Fig. 10 Mean time interval + S.E. (indicated by a vertical line running through the top of each bar) between release and flight take-off of T. lineatum, in the presence of semiochemical baits in drainpipe traps, within a wind tunnel 60

Fig. 11 Flight take-off direction of T. lineatum, in the presence of semiochemical baits in drainpipe traps, within a wind tunnel. Male and female data were pooled. Stars on top of unbaited (1988) bars denotes a significant difference (X2; p < 0.05) with the unbaited (1987) treatment. Bars for each direction, with different letters, indicates significant differences in response (X ; p < 0.05), \

xiv

for individual comparisons. The numbers of observations for each treatment were: unbaited (1988) - 137; unbaited (1987) - 138; ethanol - 146; lineatin - 150; and lineatin and ethanol - 152 61

Fig. 12 Observations of flight type exhibited by T. lineatum. in the presence of semiochemicals in drainpipe traps, within a wind tunnel. Male and female data were pooled. Proportions are based on the total number of observations, where each ' individual could be scored for more than one type of flight. Bars for each type of flight, with different letters, indicates significant differences in in response (X2; p < 0.05), for individual comparisons. The numbers of observations for each treatment were: unbaited - 140; ethanol - 152; lineatin - 158; and lineatin and ethanol - 156 63

Fig. 13 Overall flight direction exhibited by T. lineatum. in the presence of semiochemical baits in drainpipe traps, within a wind tunnel. Male and female data were pooled. Bars for each direction, with different letters, indicates significant differences in response (X ; p < 0.05), for individual comparisons. The numbers of observations for each treatment were : unbaited - 141; ethanol - 158; lineatin - 166; and lineatin and ethanol - 166 62

Fig. 14 Positive lock-on response, exhibited as landing or non-landing behavior, by T. lineatum in the presence of semiochemical baits in drainpipe traps, within a wind tunnel. Bars for each behavior and sex, with different letters, indicates significant differences in response (X ; p < 0.05), for individual comparisons. The numbers of males observed for each treatment were: unbaited (U) - 67; ethanol (E) - 79; lineatin (L) - 83; and lineatin and ethanol (L+E) - 79. The numbers of females observed were; U - 80; E - 85; L - 90; and L+E - 89 65

Fig. 15 Mean time interval + S.E. (indicated by vertical lines running through the top of each bar) between flight take-off

and trap landing of T. lineatumf in the XV

presence of semiochemical baits in drainpipe traps, within a wind tunnel. Means were compared on combined data for the sexes, where different letters on top of the bars indicate significant differences between semiochemical baits (SNK; p < 0.05) 67

Fig. 16 Trap placement in mark-recapture studies of T. lineatum on the University Endowment lands in Vancouver, British Columbia. In the first experiment, 4, 8, and 16 lineatin-baited funnel traps were placed at distances of 5, 25, and 100 m, respectively, from the release point. Experiment 2 utilized the 100 m trapline only. In experiment 3, traps were set up 500 m from the release point only. The first replication had 4 traps placed in each cardinal direction, and the second and third replications has 2 traps at 45° intervals as shown here 77

Fig. 17 Directional data based on frequency distributions of T. lineatum flight and wind observations for experiment 1 in the University Endowment Lands study 85

Fig. 18 Directional data based on frequency distributions of T. lineatum flight and wind observations for experiments 2 and 3 in the University Endowments Land study. Dashed line in circles represents mean direction. Numbers in circles are the r-values. Those marked with a star are significant (p < 0.05). The dotted lines and associated values outside the circles are the scales for the frequency bars 91

Fig. 19 A perspective view of Cedar Creek Valley, Coquitlam Lake Watershed, British Columbia. Five mark-recapture experiments for studying T. lineatum dispersal were carried out in this valley from June- August, 1987. The up-valley azimuth ranges from 48 to 68°. The road moving from the southwest to the northeast covers a distance of 4 km in this view. Slope peaks reach 1200 m in elevation, while creek elevation averages 550 m 98

Fig. 20 Directional data based on frequency distributions of T. lineatum flight and xv i

wind observations for experiments 1 - 3, in the Cedar Creek Valley, Coquitlam Lake Watershed, B.C., 1987. Line in circles represent mean direction. Numbers in circles represents the r-values. Those marked with a star are significant (p < 0.05). The number and associated line outside the circles are the scales for the frequency bars 115

Fig. 21 Recapture of marked T. lineatum in pheromone traps, for experiment 3, in the Cedar Creek Valley, Coquitlam Lake Watershed, B.C., 1987. Traps were placed circularly around the release point and in traplines 50, 375, and 700 m east and 1000 m west of it. These catches are based on 4 releases where a total of 11 196 beetles flew. A) The total number of beetles recaptured on the day of release; B) The number of beetles recaptured during the next collection date (2 - 10 days later) 120

Fig. 22 Recapture of marked T. lineatum in pheromone traps for experiment 4 in the Cedar Creek Valley, Coquitlam Lake Watershed, B.C., 1987. Traplines were set up 50, 700, 1100, 1900 m east and 1000 m west of the release point. Additionally, line E-F was placed on the other side of the valley, ca. 1300 m east of the release point. These catches are based on 2 releases where a total of 6 698 beetles flew. A) The number of beetles recaptured on the day of release; B) The number of beetles recaptured during the next collection date (3-5 days later) 12 3

Fig. 23 Contour map of the Cedar Creek Valley in the Coquitlam Lake Watershed, with the designs of the first (A) and second (B) mark-recapture experiments, carried out in 1988. Dotted polygons represent regeneration openings surrounded by old-growth forest. The up-valley azimuth = 48 - 68° 131

Fig. 24 Wind direction and T. lineatum flight recapture for experiment 1, releases 1 (A) and 2 (B), represented by circular and xy axis histograms, respectively, placed relative to their location along the length xvii

of the valley. The x-axis represents the valley azimuth. Lines in the circles are the mean direction for the first 2 h (dotted), final 2 h (dashed), and all day (solid), following beetle release, with line length signifying the r-value for the vector 143

Fig. 25 Wind direction and T. lineatum flight recapture for experiment 1, releases 3 (A) and 4 (B), represented by circular and xy axis histograms, respectively, placed relative to their location along the length of the valley. The x-axis represents the valley azimuth. Lines in the circles are the mean direction for the first 2 h (dotted), final 2 h (dashed), and all day (solid), following beetle release, with line length signifying the r-value for the vector 144

Fig. 2 6 Wind direction and T. lineatum flight recapture for experiment 2, releases 1 (A) and 2 (B), represented by circular and xy axis histograms, respectively, placed relative to their location along the length of the valley. The x-axis represents the valley azimuth. Lines in the circles are the mean direction for the first '2 h (dotted), final 2 h (dashed), and all day (solid), following beetle release, with line length signifying the r-value for the vector 14 5

Fig. 27 Wind direction and T. lineatum flight recapture for experiment 2, releases 3 (A) and 4 (B), represented by circular and xy axis histograms, respectively, placed relative to their location along the length of the valley. The x-axis represents the valley azimuth. Lines in the circles are the mean direction for the first 2 h (dotted), final 2 h (dashed), and all day (solid), following beetle release, with line length signifying the r-value for the vector 146 xv iii

ACKNOWLEDGEMENTS

The education and experiences I have received during my tenure at the University of British Columbia have been extraordinary. Many individuals and organizations have provided invaluable assistance in helping me realize the goal of completing this dissertation, and I would like to thank them all.

Dr. J. McLean, as my teacher, research supervisor, and friend has been a constant source of inspiration during the past four years with his patience and overt willingness to help and advise me in a multitude of situations. Drs.

S. Lindgren, L. Safranyik, B. van der Kamp, and

D. Lavender have provided exemplary service as members of my graduate committee. My appreciation is extended to Drs. M.

Furniss, H. MacCarthy, J. Sweeney, R. Bridges, T. Shore, L.

Phelan, and J. Borden for reviewing individual chapters. N.

Ryant, J. Glaubitz, D. Challenger, L. Friskie, and E. Kovacs provided invaluable technical assistance in carrying out the studies.

Drs. A. Kozak, P. Marshall, and M. Stephens, Mr. G.

Therien, and F. Ho contributed significantly to the statistical analysis of much of these data. My thanks are extended to the Greater Vancouver Regional District, and G.

Joyce in particular, for allowing me to carry out the field studies during each of three summers on land they administer. Meteorological data during the summer of 1986 xix were provided to me by A. Neighbour of the U.B.C.

Climatological Station. The Ministry of Environment lent me a Stevenson screen and Dr. D. Golding and K. Rae lent me their CR 21 data logging equipment. Dr. J. Worrall graciously allowed me to use much of his workspace and walk-

in cooler in the basement of Header House. Dr. P. Murtha and R. Wiart provided technical assistance and access to the

Terrasoft program used to produce the valley perspective

(Fig. 19). N. Krajci, L. Friskie, and M. Salom contributed their drafting expertise to several figures.

I am grateful to the University of British Columbia, the I.W. Killam Foundation, and the Natural Sciences and

Engineering Research Council for providing me with financial support throughout the course of these studies, and Dr. J.

Wilson for helping to facilitate the funding process.

My sincerest thanks go to my family who have had trust in my abilities and have encouraged me to continue my education. Most of all my wife Mar has never let me consider anything other than the course that I have chosen for myself. Without her love and patience this project would have been much more difficult to complete. 1

CHAPTER 1: OVERVIEW

1.1 Introduction

The striped ambrosia beetle, Trypodendron lineatum

(Olivier) (Coleoptera: Scolytidae), is the most numerous (S.

Wood 1982) and most damaging of the ambrosia beetle species in British Columbia (Prebble and Graham 1957). These dark brown insects, 2.7 - 3.5 mm long and 1.7 mm wide, bore into the sapwood of logs, stumps, and windthrown trees of almost all species of conifers (S. Wood 1982). Tree genera especially susceptible to attack include Pseudotsuqa, Tsuqa,

Picea, Abies, and Thuja (Prebble and Graham 1957). Trees are generally susceptible to attack for a period of 3 - 12 months following tree mortality (Dyer and Chapman 1965;

Mathers 1935; Prebble and Graham 1957). Characteristic of the ambrosia beetle group,, adults of this species introduce a symbiotic fungi into the gallery, Monilia ferruginea

Mathiesen-Kaarik, upon which immature and adult beetles feed

(Fisher et al. 1953; Francke-Grossman 1963; Funk 1965).

Degrade of sawtimber, from the pinholes caused by gallery production and the black stain from the fungus, alone account for more than $63 million (Canadian) yearly loss to the British Columbia forest industry (McLean 1985).

Other costs include loss of export opportunities for high quality logs; repackaging bundles of logs to remove attacked material; modification of harvesting practices, inventory 2

management, and yarding procedures; and application of pest

management strategies (Borden and McLean 1981).

These economic realities have stimulated a great deal

of research on the life history, biology, and chemical

ecology of T. lineatum (Borden 1988a; Nijholt 1979), as well

as the development of pest management strategies for use in

attempting to minimize beetle attack (Borden 1988b; Lindgren

et al. 1983; Shore and McLean 1985). Mass-trapping beetle

and removal of nearby overwintering sites (ie. forest

margins) at log sorts and booming grounds (log grading and

storage areas), is an important contribution toward reducing

beetle populations. Yet it has not and will not solve the

problem unless inventory management, which includes removal

of log inventory from the forest as soon as possible, is

practiced. The evidence justifying this recommendation has

become apparent in recent years. For example, Gray and

Borden (1985) showed that of all the logs attacked at a log

sort on northern Vancouver Island, 45% of the logs had come

into the sort already attacked by ambrosia beetles. In a

survey of the log inventory of one woodlands division on

Vancouver Island, 126 000 m3 of inventory was left on the

ground during the spring. About 80% of this material was

attacked by ambrosia beetles before being taken from the

field, resulting in an estimated loss to the forest company

of over $500 000 (Daust 1985). In a study on northeast

Vancouver Island, semiochemical-baited funnel traps

(Lindgren 1983) placed in a commercial old-growth coniferous 3 forest more than 1.5 km from a recent logging setting, captured over 22 000 T. lineatum in 12 funnel traps during the summer (McLean and Salom unpublished). This emphasizes how prevalent the beetles are in these forests and their innate ability to find attractive host material.

Despite increasing awareness by private companies of the merits of moving their log inventory more quickly, there appear to be some unavoidable factors that constrain inventory removal before the spring beetle flight. These include: the difficulty in removing right-of-way logs until roads are completed and settled; loading equipment availability; manpower; need to maintain a spring inventory so that workers involved in loading and transport operations can remain employed; mill capacity; marketability of log species; and weather, such as early snowfall or late snowmelt.

It is clear that other woods management options must be considered. Borden (1988a) recommended that cut blocks be established far apart from each other in successive years, making it more difficult for beetles to disperse and find susceptible hosts. This could certainly be considered by harvest planners and would be even more practical as we start harvesting second-growth forests where road networks will already be established. However, this approach will be difficult to follow without knowing the dynamics of T. lineatum population movement. Previous studies of T. lineatum biology and physiology have provided some 4

understanding of their flight patterns and capabilities

during the spring (Bennett and Borden 1971; Chapman 1962;

Chapman and Kinghorn 1958; Dyer 1961; Graham 1959; Rudinsky

and Daterman 19 64) and late summer-early fall dispersal

periods (Dyer and Kinghorn 1961; Kinghorn and Chapman 1959).

A synopsis of this research is presented below.

In the spring, when temperatures rise above 15°C and

the snow on the ground has melted, T. lineatum will begin to

emerge and fly (Chapman and Kinghorn 1958; Rudinsky and

Daterman 1964) from their overwintering sites in the forest

duff and bark of stumps and trees (Dyer and Kinghorn 1961;

Kinghorn and Chapman 1959), in search of host material.

Mass flights, characteristic of this species, occur during

daylight hours when temperatures reach near 20°C or above, with optimal flight temperatures ranging from 19 - 2 6°C

(Rudinsky and Daterman 1964). Flight activity is highest under calm wind conditions, and is reduced markedly at wind

speeds of 5 km/h (Chapman 1962; Rudinsky and Daterman 1964), with essentially no flight observed above 6.4 km/h (Chapman

1962). Chapman (1962) also observed that beetles fly 3.0 -

4.5m above the ground within the forest, usually above the understory brush (Shore and McLean 1984), dropping to 1.8 -

2.4 m in exposed areas, and orienting upwind toward nearby

attractive host material (Chapman 1962). With respect to

flight distance capabilities, one brief reference by Dyer

(1961) indicated that in a mark-recapture study, one T.

lineatum beetle was recaptured 4 km away from the release 5

point, a year following the release. Furniss and Furniss

(1972) collected specimens of T. lineatum on snowfields,

1200 m above timberline.

The spring dispersal period is dominated by two types

of behavior. Initially, beetles tend to be photopositive

and will fly toward the sun (Graham 1959). The length of

time this behavior is exhibited varies greatly, averaging 30

min before the beetles become arrested to host stimuli

(Bennett and Borden 1971). Females are the pioneer beetles

and are attracted to susceptible logs by host volatiles, of

which ethanol has been identified as the most important

(Moeck 1970, 1971). After females tunnel into the host,

they will release a very potent aggregation pheromone, a

tri-cyclic ketal given the name lineatin (3,3,7-trimethyl-

2,9-dioxatricyclo [3.3.1.04'7] nonane) to which both sexes

respond (Borden et al. 1979; MacConnell et al. 1977).

Graham (1959) hypothesized that the initial response to

light by T. lineatum served as a mechanism to force at least

some beetles in a population to disperse in search of new hosts, usually scarce in nature. This is now considered a

fairly common behavior for many insects that disperse or migrate during their life cycle (Kennedy 1975). This behavior may be released at a fat content threshold below which photopositive dominance is lost and arrestment to host

stimuli occurs, as Atkins (1966a, 1969) and Bennett and

Borden (1971) proposed for the Douglas-fir beetle,

Dendroctonus ponderosae Hopkins. 6

From early July to late October, brood adults and some

parents will emerge from their breeding areas and fly to

overwintering sites (Kinghorn and Chapman 1959). Dyer and

Kinghorn (1961) found that beetles are photopositive when

emerging from the wood material and will fly ca. 15 m above

the ground before heading toward the nearest forest. They

also determined that this photopositive dominance was lost

as they enter the shade of the forest. Interestingly, and

to the dismay of pest managers, fall dispersing brood adults

are probably not responsive to semiochemicals, as they are

not inclined to mate at this time (Fockler and Borden 1972).

Yet conclusive tests are needed to verify this. With

respect to flight capacity, fall dispersers were found to

fly for much shorter periods than spring dispersers (Chapman

1956).

Since spring flying beetles appear to be the more

important of the two dispersal stages (Chapman 1956), and because beetles from this stage would be easier to study, due to their positive response to semiochemicals, I will hereafter, concentrate almost exclusively on this period in the life cycle of T. lineatum. At the present time, basic

flight characteristics of T. lineatum. with respect to seasonal and diurnal periods, has been well established in both North America (Chapman and Kinghorn 1958; Rudinsky and

Daterman 1964) and Europe (Annila et al. 1972), with little difference occurring between beetles from both regions.

However, little is known about the dynamics of population 7 movement. Specifically, data are needed concerning flight distance and direction, and dispersal rate, as they relate to wind patterns, topography, vegetative cover, and host locations. Additionally, beetle response and flight orientation to wind and different semiochemical components have yet to be adequately examined (Borden 1977). The influence of wind, as it interacts with the environment, is clearly one of the most important factors influencing insect dispersal (Johnson 1969; Pedgley 1982).

The collection of data concerning population movement and flight orientation to hosts, can lead to significant pest management recommendations for harvesters and woodlands managers concerned with ambrosia beetle damage. Cut blocks can be distributed spatially to minimize risk of beetle attack. Also population suppression strategies with semiochemicals, similar to those used at the log sorts and booming grounds, may be applied in the forest settings. As discussed earlier, the focus of this problem needs to be extended from the log sort and yarding operations into the harvest areas.

1.2 Objectives

Flight orientation and response of T. lineatum, as well as other scolytids, has not been adequately investigated in detailed behavioral studies (Borden 1977). Lack of such studies often make interpretation of beetle trapping field experiments more difficult. Controlled behavior studies 8

concerning flight are usually carried out in wind tunnels

(Baker and Linn Jr. 1984). This bioassay has been

inherently difficult to use for the study of scolytid beetle

flight behavior (Schlyter et al. 1987a), with only one

successful behavior study published (Choudhury and Kennedy

1980). Therefore, my first objective was to determine the

optimal wind tunnel conditions needed for studying T.

lineatum flight behavior in the presence of semiochemicals.

Preliminary observations indicated the importance of wind

speed, and since this factor had yet to be examined, for

this or any other scolytid beetle (Borden 1977), it was the

subject of the experiments presented in Chapter 2 of this

report.

Another critical component for determining optimal wind

tunnel conditions include the determination of optimal

release rates of olfactory stimuli (Baker and Linn Jr.

1984; Schlyter et al. 1987b). This led to an experiment,

presented in Chapter 3, that examined the number of T.

lineatum captured in funnel traps baited with varied release

rates of ethanol and lineatin, alone and in combination.

Current pest management procedures for T. lineatum rely

heavily on semiochemical-based mass trapping programs

(Borden 1988b). The use of primary and secondary

attractants in trapping systems has come under debate in

North America (Borden et al. 1982; Salom and McLean 1988;

Shore and McLean 1983), which was stimulated by studies in

Europe, showing positive response of T. lineatum to host 9

attractants (Bauer and Vite 1975) that act synergistically

with the aggregation pheromone lineatin (Paiva and Kiesel

1985; Vite and Bakke 1979). Thus, it was felt that a

detailed investigation into the behavioral responses of T.

lineatum from North America to ethanol and lineatin, alone

and together, were needed to aid in interpreting previous

studies, as well as set a clearer path for further

improvements in trapping technology. This investigation is

presented in Chapter 4.

Our lack of knowledge concerning population movement of

T. lineatum prompted me to carry out a series of field

experiments of increasing complexity, in an effort to

characterize beetle flight patterns in the spring with

respect .to wind, topography, land cover, and location of

olfactory stimuli. Such field experiments generally employ

some form of mark-recapture techniques, which can include marking insects with fluorescent powders and paints, labels,

feeding with dyes, genotype differentiation, and radioactive

isotopes (Southwood 1978). The most common form of marking technique used for bark and wood boring beetles has been the use of fluorescent powders (Helland et al. 1984; Linton et al. 1987; Schmitz 1979; Shore and McLean 1988). This technique seemed suitable for use in these field studies.

The first set of field experiments was designed with the following objectives: 1) develop competence in using mark-recapture as a tool for studying T. lineatum movement; and 2) characterize short (5 - 25 m), intermediate (100 m), 10 and long distance (500 m) beetle flight in a second-growth coniferous forest. This work is presented in Chapter 5.

In a second set of field experiments, presented in

Chapter 6, the study site was moved to a valley setting, more typical of commercial forest situations along the coast of British Columbia. Some of the same short distance flight patterns were examined again, however, this time within an old-growth coniferous forest. Additionally, emphasis was placed on T. lineatum long distance flight capabilities and rate of dispersal.

The last two field experiments, presented in Chapter 7, were carried out in the same valley as were the previous set of experiments. They were designed to characterize long distance T. lineatum movement, without the presence of semiochemicals near the release sites, and with equal emphasis placed on up- and down-valley beetle movement.

Trap catch comparisons were made between forest and open settings, and use of roads as flyways by beetles was investigated. Equal numbers of beetles were released simultaneously from two separate release points, allowing for a better representation of beetle dispersal throughout the valley (Banks et al. 1988).

1.3 Format

Chapters 2-7 have been written up as separate papers in publication format. They should be able to stand alone 11

except all references are listed after Chapter 8. Chapters

2-4 are written for journals that separate the results and

discussion sections. In contrast, chapters 5-7 are written for journals that combine these two sections. All

other aspects of the format between chapters will be

consistent throughout the dissertation. While I have

attempted to exclude redundant presentations and discussions between the chapters, it was impossible to remove all of it

completely. Hopefully this will not detract too much from the reading. 12

CHAPTER 2: THE EFFECT OF WIND SPEED ON TRYPODENDRON LINEATUM FLIGHT BEHAVIOR IN A WIND TUNNEL

2.1 Introduction

Orientation behavior of flying insects in response to

semiochemicals has been a subject of recent investigations

and reviews (Bell and Carde 1984; Borden 1977; Farkas and

Shorey 1974; Kennedy 1978; Payne et al. 1986; Shorey 1973).

Olfactory orientation has been shown to vary with respect to

chemical gradients, wind, light, as well as physiological

requirements and capabilities of the insect. Upwind flight

is generally accepted as the behavior of most insects in

response to the detection of attractive material (Baker and

Linn Jr. 1984). Several species of scolytid beetles have

demonstrated upwind flight to attractive material in the

field (Botterweg 1982; Byers 1988; Chapman 1962; Gara 1963;

McMullen and Atkins 1962;).

Under more controlled conditions, wind tunnels have

become a valuable tool for deciphering the complex flight

patterns of insects (mostly Lepidoptera) (Baker and Linn Jr.

1984). They allow the investigator to test for individual

effects under a relatively rigid experimental design and

procedure not possible in the field. While upwind

orientation to semiochemicals has been demonstrated for many

species of Lepidoptera (Preiss and Kramer 1986), it has been

demonstrated for only two species of scolytids (Choudhury

and Kennedy 1980; Lindgren and McLean, unpublished data), 13

one anobiid species (Birch and White, 1988), and one clerid

species (Mizell et al. 1984).

The striped ambrosia beetle, Trypodendron lineatum

(Olivier), uses semiochemicals to find suitable hosts for

colonization, mating, and reproduction. During spring

flight dispersal, T. lineatum must detect and respond to

suitable host material, mainly coniferous logs that have been harvested for at least 3 months (Mathers 1935).

Temperature, time of day, and wind direction are known to

affect general flight activity, but little is known about

close-range flight behavior to host material, with respect to varied environmental conditions for this insect as well

as forest Coleoptera in general (Borden et al. 1986; Carde

1984).

Two wind tunnel studies presented here examine the effect wind speed has on close-range response of T. lineatum to attractive substrates. The first study was designed to determine the wind speed at which the highest proportion of beetles released would respond positively to a

semiochemical-baited funnel trap. It was hypothesized that the highest proportion of released beetles would be recaptured in the baited traps placed upwind, at the lower end of the wind speed range (0.0 to 0.9 m/s) tested. A second study was designed to see if the mechanisms by which

T. lineatum oriented to a semiochemical-baited substrate differed throughout the same range of wind speeds that were tested in study 1. One hypothesis tested was that chemically-induced anemotaxis would occur when wind is present and a second hypothesis was that a different orientation mechanism for attraction is used by the insect when wind is absent.

2.2 Materials and Methods

Both studies were conducted in a wind tunnel (Angerilli and McLean 1984) 3.6m long and 1.2 m high and wide. Cool white fluorescent lights (660 W) set 1.5 m above the wind tunnel simulated diurnal conditions and facilitated direct observations of flying beetles. A sheet of cellulose acetate was placed over the top of the tunnel to diffuse the incoming light and reduce glare. Incident illumination readings from within the tunnel were measured with a

Sekonic® model L-398 exposure meter. Vertical light measured 0.8 m from the ceiling in the exhaust, middle, and wind entrance sections of the tunnel were 112, 125, and 140 lx, respectively. Thus, a slight light gradient from the ceiling was present.

Beetles were collected daily during May and June of

1986 and 1988, in lineatin-baited, multiple-funnel traps, near a large log boom storage area at the mouth of the North arm of the Fraser River, British Columbia. Captured beetles were stored at 4°C, under a 14:10 h (L:D) photoperiod, in 1

L plastic jars containing slightly moistened absorbent cloth towels. The containers were checked twice weekly to monitor 15 moisture levels. Beetles survived for more than 100 days under these conditions. Since the beetles were captured in pheromone traps, they were judged to have had sufficient

flight exercise (Graham, 1959; Bennett & Borden, 1971) to

ensure subsequent response to semiochemicals.

Study 1

A multiple-funnel trap, placed in the center of the upwind section of the tunnel, ca. 0.6 m from the screen, was baited with two lineatin Biolures® (Consep Membranes, Bend,

Oregon) in the upper and lower portion of the trap, and a dispenser containing 95% ethanol in the middle. Release rates, determined at 2 0°C in the laboratory were 100 - 12 0 ug/24 h for each lineatin dispenser and 75 mg/24 h for the ethanol dispenser (B.S. Lindgren, personal communication)^

The lures and release rates were identical to those used in mass-trapping programs of T. lineatum throughout southwestern B.C.

A release platform consisting of a circular, flat glass plate 13 cm in diam, was suspended horizontally 35 cm above the tunnel floor and 1 m from the downwind screen.

Preliminary trials suggested that optimal response by T. lineatum occurred when the release platform was placed at this height. It was high enough for the beetles to be continuously exposed to the semiochemical plume prior to take-off, yet close enough to the striped floor to use ll. Research Director, Phero Tech. Inc., 1140 Clark Dr., Vancouver, B.C. V5L 3K3. 16

optomotor cues if needed. The glass plate was covered with

a white industrial paper towel to enhance traction for the

insects. A 16 cm2 x 0.7 cm deep plastic weighing boat,

covered by the same material, was taped to the upper side of

the glass plate. This allowed the beetles two perching

opportunities; the first around the edge of the weighing boat, and if they fell off that, a second opportunity around

the edge of the glass plate.

Beetles were sexed and given a walking test in the morning prior to flight in the wind tunnel. They were

allowed to acclimate to the warmer room temperatures (22 -

25°C) for 10 min before release, to ensure maximal response

of the beetles to the trap.

A nine-replicate, randomized complete block design was used, in which the treatments were wind speeds of 0.0, 0.15,

0.30, 0.45, 0.60, 0.75, and 0.90 m/s (0 - 3.2 km/h). For each treatment, males and females were tested separately.

Groups of 50 males and 30 - 50 females were placed on the release tray. All treatments within one block (replication) were run for each sex on the same day, and the experiment was run on 9 days over a 21 day period. Each replication for a given treatment lasted for 20 min. Beetles were never used more than once.

An additional three-replicate, control experiment assessed male and female catches at 0.0 m/s in unbaited traps. Prior to these replicates, the funnel trap, tunnel, 17

and release tray were cleaned with a 0.5% sodium

hypochlorite solution.

With treatment values set at equal intervals throughout

the wind speed range tested, the data were analyzed using

orthogonal polynomials, followed by regression analysis.

For males, the variances deviated significantly from

homogeneity, using Box's test (Fox and Guire 1976); thus a weighted regression analysis was conducted, with the weights

equaling the reciprocals of the error variances (SAS 1985).

Male and female regression equations were then compared

using an F-test of coincidental regressions (Zar 1984).

Study 2

A preliminary study was conducted to determine the

number of beetles landing in or on various semiochemical- baited substrates. These included a 28 cm diam x 71 cm long

log of western hemlock, Tsuga heterophylla (Raf.) Sarg.,

oriented either vertically, or horizontally across the width

of the tunnel; a simulated log (SL) of the same dimensions

as the real log, made of cardboard, cotton fiber tag paper,

and construction paper, oriented in the same two ways as the

real log; and a funnel trap. The beetles showed no preference for any substrate or orientation (Appendix 1).

Therefore a horizontal SL was used to assess the effect of wind speed on flight behavior, because it allowed for

control of released volatiles, not possible with a real log,

and assumed the position of most logs in the field. The SL 18 was placed on two concrete blocks with holes that reduced

the disruption of air flow. Two lineatin lures and an

ethanol dispenser, identical to those used in study 1, were hooked to the SL along the upwind side, providing a large,

relatively even downwind plume, as verified by releasing

TiCl4 smoke at the different wind speeds. The release tray

and its placement in the tunnel were the same as in study 1.

To study T. lineatum flight behavior, the overall

experiment was divided up into two parts. Part 1 examined beetle responses to a SL baited with lineatin and ethanol, while part 2 examined beetle responses to an unbaited SL,

serving as a control. Each part examined beetle responses at wind speeds of 0.0, 0.3, 0.6, 0.9 m/s. Parts 1 and 2 were conducted in 1986 and 1988, respectively. The tunnel and the release tray were carefully cleaned with a 0.5%

sodium hypochlorite solution prior to running part 2. The

0.0 m/s treatment in part 1 was divided into two treatments, one in an unaerated tunnel and the other in the same tunnel aerated for 10 min. This was carried out to determine if contamination occurred in the unaerated tunnel. No differences were observed between the two treatments, and as a result, the data were pooled, and are presented as one treatment hereafter. All other procedures used in conducting both parts of the experiment were the same.

During the experiment, two beetles of the same sex were released per replicate and observed for up to 8 min. Fifty replicates were run for each level of wind speed and bait 19 tested. Observations included occurrence of flight, body orientation prior to and during take-off (Fig. 1), take-off direction (Fig. 1), overall flight direction, type of flight, and lock-on behavior. Overall flight direction was scored as uptunnel, downtunnel, or non-directional, depending upon the predominant direction flown by the insect throughout the observation period. If beetles did not exhibit predominant up- or downtunnel flight, they were considered non-directional fliers. Type of flight was scored as: erratic - quick, sharp changes in movement, often occurring when beetles bounce off the tunnel walls; steady - a settled, stable movement typical of flight in the field

(personal observation); and direct - a fast, straight-line flight to some endpoint. Lock-on behavior can be described by steady flight in an upwind direction, or uptunnel toward the log in the absence of an airflow, with the insect apparently orienting specifically to the SL, either by casting up and downtunnel, hovering around the SL, or landing on it. All observations were recorded on a cassette tape and later transcribed on to data sheets.

The data were summarized as frequencies and analyzed separately for each behavioral category, as multidimensional contingency tables using the CATMOD procedure in SAS (1985).

This procedure uses a linear models approach to the analysis of categorical data, applying general weighted-least squares regression techniques to estimates of appropriate (linear and/or log linear) functions of the cell proportions in ORIENTATION PRIOR TO TAKE-OFF

WIND DIRECTION

FLIGHT TAKE-OFF DIRECTION

NO DIRECTION

Fig. 1. Body orientation directions of T. lineatum scored prior to and during flight take-off, as viewed from above the release tray (top); and flight take-off directions of the same beetles, as viewed from the side (bottom). All beetles taking off that did not fit in categories A to D were placed in the no-direction category. 21 complex categorical data layouts (Kleinbaum & Kupper, 1978).

This approach allowed me to test for first and second-order

interactions, main effects, and contrast between levels of main effects, for a specific response, using the chi-square

(X2) statistic.

2.3 Results

Study 1

The relationship between the numbers of T. lineatum caught at different wind speeds was similar for both sexes; the highest numbers of beetles were caught at 0.0 m/s. The proportion of males and females caught were 21.4 and 25.3%, respectively (Fig. 2). The percent response decreased linearly as wind speed increased. However, much of the variation in these data is not explained by the regression equations, with r2 values of 0.28 for both sexes. The high variation in these results may be due in part to the varied physiological states (e.g. varying energy reserves) of the beetles, resulting in a wide range of behavioral responses, or the conditions in the wind tunnel were not optimal for T. lineatum response, or both. The slopes and intercepts of the regression equations were not significantly different

(F120,115 = °-986'* P > 0.05) from each other, indicating no difference in response between sexes. Beetle responses to the unbaited control treatment at 0.0 m/s averaged only

3.3% for males and 2.0% for females (Fig. 2, black boxes), 22

.50

MALES Y = 21.4 - 17.2X

r2 = 0.2807 40 P < 0.0001

30

5! 't2 20 HI

10 - O < U W

tu 60 CQ FEMALES Y = 253 - 19.7X Q 50 - W r2 to ii-2 = 0.2753 P < 0.0001 I 40 o

WIND SPEED (M/S)

Fig. 2 The influence of wind speed on response of male (top) and female (bottom) T. lineatum to a semiochemical-baited funnel trap in a wind tunnel. Numbers next to boxes represent the number of data points. White and black boxes represent the % response for the baited and unbaited (control) treatments, respectively. Missing data accounts for some wind speeds having less than 9 observations. 23 indicating lack of contamination in the tunnel and little visual attraction to the trap under these conditions.

Study 2

The number of beetles taking flight was not influenced by wind speed (Table 1). Frequency of flight was analyzed by sex and bait separately because of a significant

interaction occurring between these factors (X^ = 17.1; p <

0.05), with males responding at lower frequencies than females to a baited SL.

Body orientation prior to and during flight take-off did not differ between males and females, therefore the data for the sexes were pooled. Uptunnel body orientation increased significantly in the presence of an airflow for both baited and unbaited treatments (Fig. 3). The presence of semiochemicals did not result in increased uptunnel orientation, but it did result in some increase in the occurrence of non-directional orientation, suggesting more active movement on the release tray.

Take-off direction of T. lineatum did not differ between males and females, and therefore the data for the sexes were pooled. At 0.0 m/s, take-off direction was basically uptunnel/vertical, changing to a downtunnel/vertical direction at 0.3 - 0.6 m/s, and then dominated by downtunnel take-off at 0.9 m/s in both baited and unbaited treatments (Fig. 4). A higher percentage of T. lineatum appear to take-off uptunnel in the presence of Table l. Frequency of Trypodendron lineatum flight in a wind tunnel at varied wind speeds in the presence of an unbaited and semiochemical- baited simulated log. 1,2

Wind speed % Flying (m/s) Males Females

Unbaited

0.0 61 67

0.3 70 74

0.6 77 68

0.9 70 80

Baited

0.0 56 72

0.3 44 74

0.6 47 76

0.9 45 67

100 beetles released for every level of wind speed and bait treatment (50 for each sex).

No significant differences in number of beetles flying, where each sex by bait treatment were analyzed separately (X2; P > 0.05). TOTAL NUMBER OF BEETLES OBSERVED WINDSPEED (M/S) UNBAITED BAITED

0.0 111 112 a a 0.3 125 113 0.6 135 108 b 0.9 137 100 B a a _a z b b a a w 80 -, b b b O BAITED w 60 to Ed 40 Z be 20 b ab ab a a AJ, o UNBAITED

o —i—i—i—i— —i 1 1 1— —i 1 1 1— —i 1 1 r 0.0 0.6 0.0 0.6 0.0 0.6 0.0 0.6 WIND SPEED 0.3 0.9 03 0.9 03 0.9 0.3 0.9 (M/S) UPTUNNEL CROSSTUNNEL DOWNTUNNEL NO DIRECTION

BODY ORIENTATION PRIOR TO TAKE - OFF

Fig. 3. Body orientation of T. lineatum, prior to and during flight take-off, at different wind speeds, in the presence of an unbaited and semiochemical- baited simulated log. Male and female data were pooled. Bars with different letters for each response by bait treatment category are significantly different (X2; p < 0.05) for individual comparisons. TOTAL NUMBER OF BEETLES OBSERVED

WINDSPEED (M/S) UNBAITED BATTED

0.0 118 116 0.3 135 114 0.6 135 118 0.9 139 106

B z w C BAITED Was (A w oz UNBAITED

~i 1 1 r ~l 1 1 r i r ~i r r 0.0 0.6 0.0 0.6 0.0 0.6 0.0 0.6 WIND SPEED 0.3 0.9 0.3 0.9 03 0.9 0.3 0.9 (M/S)

UPTUNNEL UPTUNNEL/ DOWNTUNNEL/ DOWNTUNNEL VERTICAL VERTICAL

FLIGHT TAKE - OFF DIRECTION

Fig. 4. Flight take-off direction of T. lineatum, at different wind speeds, in the presence of an unbaited and semiochemical-baited simulated log. Male and N3 female data were pooled. Bars with different letters for each response by bait treatment category are significantly different (X2; p < 0.05) for individual comparisons. 27 semiochemicals than without them, but this could not be tested statistically because of the occurrence of a significant wind speed by bait interaction (X g = 23.5; p <

0.05).

Overall flight direction changed from uptunnel and non- directional to downtunnel upon exposure to wind for both the baited and unbaited treatments (Fig. 5). Increases in wind speed above 0.3 m/s did not cause any further change in flight direction for baited treatments, but did cause an increase in downwind flight for unbaited treatments. A higher frequency of beetles exhibited uptunnel flight in the presence of semiochemicals, with statistical differences

2 occurring at windspeeds of 0.3 (X 2 = 7.1; p < 0.05), 0.6

2 2 (X 2 = 10.9; p < 0.05), and 0.9 m/s (X 2 = 8.8; p < 0.05).

2 Flight type data show significant speed by bait (X 6 =

2 20.3; p < 0.05) and speed by sex (X 6 = 16.4; p < 0.05) interactions. The frequency of erratic flight displayed by female beetles in response to both baited and unbaited SI/s. decreased significantly as wind speed increased from 0.6 to

0.9 m/s (Fig. 6). The frequency of erratic flight of male beetles varied significantly with no apparent pattern to changing wind speed in response to the baited SL, but was unchanged in response to the unbaited treatment (Fig. 6).

At a wind speed of 0.0 m/s, steady flight of both sexes was more frequent in response to the baited treatment than to the unbaited treatment, however this difference between baited and unbaited treatments declined at higher wind TOTAL NUMBER OF BEETLES OBSERVED WINDSPEED (M/S) UNBAITED BATTED

0.0 117 128 0.3 145 118 0.6 144 123 0.9 148 112

b ab, 100 b_b_ bj • B B / BAITED

50

b b b A b 0 b O S> C=J • S Bf UNBAITED —I 1 1 r— —i 1 1 1— —i 1 1 r 0.0 03 0.6 0.9 0.0 0.3 0.6 0.9 WINDSPEED 0.0 0.3 0.6 0.9 (M/S) UPTUNNEL NO DIRECTION DOWNTUNNEL OVERALL FLIGHT DIRECTION

Overall flight direction exhibited by T. lineatum, at different wind speeds, in the presence of an unbaited and semiochemical-baited simulated log. Male and female data were pooled. Bars with different letters for each response by bait treatment category are significantly different (X ; p < 0.05) for individual comparisons. TOTAL NUMBER OF BEETLES OBSERVED

MALES FEMALES WTNDSPEED (M/S) UNBAITED BATTED UNBAITED BAITED

0.0 61 56 66 69 0.3 64 40 72 74 0.6 75 44 64 73 0.9 65 41 74 67

a a

BAITED FEMALES

BAITED MALES

UNBAITED FEMALES UNBAITED MALES i 1 1 r ~i r WIND SPEED 0.0 03 0.6 0.9 (M/S) 0.0 03 0.6 0.9 STEADY DIRECT ERRATIC

TYPE OF FLIGHT

Fig. 6. Type of flight exhibited by T. lineatum. at different wind speeds, in the presence of an unbaited and semiochemical-baited M simulated log. Bars with different letters for each response by bait treatment by sex category are significantly different (X ; p < 0.05) for individual comparisons. 30 speeds. For baited males, it was an abrupt change, whereas for females, it was gradual, with similar frequencies of steady flight not occurring until wind speeds reached 0.6 m/s. These differences could not be tested statistically because of the speed by bait interaction, but they were apparent when plotted two dimensionally on a rectangular coordinate system.• Observed decreases in erratic and steady flight for both males and females, as wind speed increased, resulted from marked increases in direct flight, especially noticeable from 0.6 to 0.9 m/s.

No differences in positive lock-on responses were observed between sexes with respect to wind speed and bait; therefore male and female data were pooled. (Fig. 7). The frequency of T. lineatum exhibiting lock-on response to a baited and unbaited SL interacted significantly with wind

2 speed (X 6 = 32.9; p < 0.05), preventing statistical comparisons between the two. However, lock-on response was generally higher in the presence of semiochemicals at all wind speeds than when the semiochemicals were absent. The differences in response between these treatments were especially noticeable in the absence of wind, where 54% of the flying beetles locked on to the baited SL, with 29% landing on it, compared to only 8% locking on to the control

SL, of which less than 5% landed. Positive response decreased in the presence of an airflow, yet increasing wind speed from 0.3 to 0.9 m/s did not result in any further decreases. TOTAL NUMBER OF BEETLES OBSERVED W1NDSPEED (M/S) UNBAITED BAITED

i i 1 1 1 1 1 1 1 1 1 1— 0.0 03 0.6 0.9 0.0 0.3 0.6 0.9 0.0 03 0.6 0.9 WIND SPEED (M/S)

LOCK-ON WITH LOCK-ON WITHOUT NO RESPONSE LANDING LANDING

POSITIVE LOCK - ON RESPONSE

Fig. 7. Positive lock-on response exhibited by T. lineatum, at different wind speeds, in the presence of an unbaited and semiochemical-baited simulated log. Male and female data were pooled. Bars with different letters for each response by bait treatment category are significantly different (X p < 0.05) for individual comparisons. 32

2.4 Discussion

In study 1, the decrease in trap catch with increased wind speed was expected since T. lineatum prefer to fly under low wind speed conditions (Rudinsky and Daterman

19 64). However, the occurrence of highest beetle catch

frequency at 0.0 m/s was unanticipated. This suggests that

the beetles captured within an airflow oriented upwind to the trap, yet beetles captured without an airflow must have

responded by using a different orientation mechanism or

flight pattern.

In the absence of wind and semiochemicals, the 3:1

ratio of uptunnel vs. downtunnel take-off direction (Fig. 4) may be a result in part from the light gradient coming through the ceiling of the tunnel. Francia and Graham

(1967) found that T. lineatum were photopositive before and

after flight, although variation associated with this behavior was high. In addition, beetles may have visually detected the availability of more flying space uptunnel than downtunnel from the release tray. These explanations may

also account for the higher proportion of uptunnel vs. downtunnel fliers in the absence of wind and semiochemicals

(Fig. 5). In this case however, increased space uptunnel

from the release tray would not be just perceived through vision, but would be based on the beetle's actual flight

activities. 33

Both wind and semiochemicals significantly influenced beetle behavior in most of the response categories studied.

The higher frequency of positive response to a semiochemical-baited SL in the absence of wind may have resulted from initial photopositive flight or random flight bringing the beetles close to the SL by chance, followed by chemo-orientation near the SL. The increase in steady

flight, lock-on response, and landing in the presence of the

semiochemical-baited SL indicates some form of chemo- orientation. Wind appears to override any photopositive or random flight behavior the beetles might exhibit, as take• off direction (Fig. 4) and overall flight direction (Fig. 5)

shifted substantially from uptunnel to downtunnel upon their exposure to a wind speed as low as 0.3 m/s.

While upwind chemo-anemotaxis (a directed response to a chemical mediated by wind) appears to be the mechanism by which T. lineatum find and orient to a pheromone source

(Chapman, 1962; Chapters 5 and 6), close-range (i.e. < 1 m) orientation and landing on the host substrate does not appear to require wind. The chemo-anemotactic responses, reported in wind tunnel studies by Choudhury and Kennedy

(1980) for Scolytus multistriatus (Marsh.), Birch and White

(1988) for Anobium punctatum (Deg.), and Mizell et al.

(1984) for Thanasimus dubius (F.) did not incorporate a pheromone source within a no-wind environment. With

Lepidoptera, Farkas and Shorey (1972) demonstrated that the pink bollworm, Pectinophora gossypiella (Saund.), did not 34 require moving air to fly toward a pheromone source, and

suggested that the moths oriented to the chemical stimuli by sensing differences in concentration. Baker and Kuenen

(1982) had similar results with the oriental fruit moth,

Grapholitha molesta (Busk.), yet they showed that

longitudinal klinotaxis was the mechanism used by these moths to orient to the odor source. In both studies, flight was initiated within an airflow. In the study reported in this chapter, flight was initiated without an airflow. The mechanism of orientation used by T. lineatum possibly resulted from a combination of first photopositive or random

flight, and then chemically stimulated orientation. It is possible that beetles combined visual and olfactory stimuli to land on the SL (Borden et al. 1986; Kennedy 1986).

Combined use of visual and olfactory cues in Scolytidae have been reported by Pitman and Vite (1969) for the mountain pine beetle, Dendroctonus ponderosae Hopk., and by Tilden et al. (1983) for the western pine beetle, Dendroctonus brevicomis LeConte. Further experiments are needed to separate out the effects of each of these variables, before a precise description of T. lineatum orientation to a host in the absence of wind can be made.

With wind present, beetles tend to take-off and fly downwind, away from the source of semiochemicals. Despite the constant exposure of the beetles to the semiochemical plume, only 2 0% of the beetles appeared to respond chemo- anemotactically. The high frequency of non-responders may be a result of the artificial conditions imposed on the beetles by the wind tunnel, in which the small dimensions

(relative to a field setting) may have inhibited the beetles

from carrying out the full complement of their flight behavior, such as flight exercise, before responding positively to semiochemicals. However, further work has shown that additional flight exercise prior to testing of the already flown beetles in the wind tunnel did not enhance the frequency of positive response to semiochemicals

(Appendix 2).

Another explanation for low response is the variation in physiological state in individual beetles, considered an important factor in influencing orientation of bark beetles to olfactory stimuli (Borden et al. 1986; Wood 1982). These authors hypothesized that some beetles respond immediately to semiochemicals and colonize suitable host trees nearby, while others fly for a period of time before responding, and thus start new centers of attraction at greater distances.

Atkins (1966a, 1969) linked this dispersal pattern to fat content in Dendroctonus pseudotsuqae. Physical state of the beetles such as nematode or microorganism infection may also play a role in this. Although I am not sure of the mechanism in T. lineatum, I believe the variability in response behavior within a population may account for the high variation in response for this study. Apparently, low frequency of ambrosia beetle response to semiochemicals is quite common. In bioassays of walking insects within olfactometers, 40 to 50% response is the best that has been observed (Borden et al. 1977; Borden et al. 1980; MacConnell et al. 1977).

The decreases in T. lineatum response to semiochemicals at increasing wind speeds correspond well with observations

in field situations, where twice as many marked T. lineatum were recaptured in a forest when wind speed was half the normal average of 0.4 m/s (Chapter 6). Other scolytid beetles have also been captured in higher numbers during periods of calm, often associated with oncoming thunderstorms (Vite et al. 1964; Edson 1978).

It may be argued that by increasing the wind speed in the tunnel, a decrease in the size of the semiochemical plume would result, reducing the chance that beetles will respond while in flight. Sanders (1985) showed that a plume of smoke released from a point source decreased in size with an increase in wind speed. However, in our studies, the plumes were released from a much larger area than a point

source, and a test with TiCl4 smoke showed that the plume sizes did not decrease significantly between 0.25 and

0.75 m/s.

Another consideration is that at lower wind speeds the semiochemical concentration per unit volume of air in the tunnel would be increased, possibly accounting for the

increased response by the beetles. Yet release rates for both studies were 10 times higher than those which elicit optimal responses in the same tunnel (Chapter 3). Thus, at lower wind speeds when semiochemicals are above optimal

release rates, positive response would most likely decrease

rather than increase.

In summary, the presence of wind as well as its speed has a significant influence on T. lineatum response to a

semiochemical-baited substrate. In these experiments I was

able to study the influence of these factors at a

longitudinal range of ca. 3 m. Different behavior patterns by T. lineatum were observed in the presence and absence of wind. When airflow was absent, flight tended to be random,

although more beetles ended uptunnel near the SL possibly due to the light gradient in the tunnel, giving them an opportunity to be arrested by semiochemicals diffusing from the source. With wind present, the majority of beetles took off and flew downwind. Beetles that did respond to the semiochemicals flew upwind and oriented to the source

instead of being arrested by it.

In describing T. lineatum dispersal and subsequent attraction to host material, it is important to keep the environmental conditions as well as the physiological status of the beetles in context. In a field setting where wind

speeds were light (< 0.5 m/s), T. lineatum generally flew in all directions (Chapter 5), yet at long distances, such as

500 m, beetles were caught downwind, whereas at short distances (< 25 m) responding beetles flew upwind toward

semiochemical-baited traps. Field data, relating highest beetle catches to still wind conditions, as well as this 38 wind tunnel study, suggest that once T. lineatum arrive within 1 or 2 m of the host, arrestment, orientation, and

subsequent landing on the host is best facilitated during periods of light or nil wind conditions. In less optimal periods, when the wind speed increases, T. lineatum can

orient to and land on a host using chemo-anemotaxis, but will do so at lower frequencies. The different mechanisms used by spring dispersing T. lineatum as they search for

attractive host material allow them to optimize their search by adapting to the varying environmental conditions

encountered. 39

CHAPTER 3: RESPONSE OP TRYPODENDRON LINEATUM TO VARIED RELEASE RATES OF ETHANOL AND LINEATIN IN A WIND TUNNEL.

3.1 Introduction

The ambrosia beetle, Trypodendron lineatum (Olivier),

like most species of the family Scolytidae, relies on its

olfactory perception of chemicals for host attraction and mating (Borden 1985). Because this insect is a major pest

of logged timber on the Pacific coast of British Columbia, a substantial effort has been put into identifying and quantifying the chemicals to which the beetle responds

(Lindgren et al. 1983). This has resulted in.improved methodology in surveying and managing the pest (Borden and

McLean 1981; Lindgren and Borden 1983).

Three chemicals have been identified as attractants for

T. lineatum in North America; ethanol (Moeck 1970, 1971) and c<.-pinene (Nijholt and Schonherr 1976) as host attractants, and lineatin, a tri-cyclic ketal (MacConnell et al. 1977;

Borden et al. 1979) as the aggregation pheromone.

Combinations of these chemicals, for use in trapping programs, have been investigated in Europe and western North

America where T. lineatum occurs. Borden et al. (1982) confirmed reports from Europe (Paiva and Kiesel 1985; Vite and Bakke 1979) that ethanol, oC-pinene, and lineatin acted

synergistically in attracting T. lineatum. However, Borden et al. (1982) found that ethanol and o<-pinene did not 40

enhance the attraction of T. lineatum to sticky wire mesh

or drainpipe traps in British Columbia. In contrast, Shore

and McLean (1983) found that ethanol and £K.-pinene together

did act synergistically with lineatin in attracting North

American T. lineatum adults to drainpipe traps.

The different results may be attributed to several

factors. Firstly, Borden et al. (1982) used release rates

of ethanol and oC-pinene 3-6 and 2-3 times greater,

respectively, than did Shore and McLean (1983). Secondly,

Shore and McLean (1983) actually tested for interactions between the semiochemicals of Gnathotrichus sulcatus

(LeConte) and T. lineatum. These interactions may have

influenced beetle response to traps baited only with T.

lineatum semiochemicals. In addition, the experimental design used by Shore and McLean (1983) was not planned to differentiate the effects of ethanol and oC-pinene on beetle

response.

The conflicting results of these studies, along with

the development and widespread use of the multiple-funnel

trap (Lindgren 1983), prompted the investigation of the

importance of these semiochemicals in attracting T. lineatum

to the efficient new traps. A field study was first

conducted in British Columbia by Salom and McLean (1988) , where it was shown that lineatin was the only effective

chemical in attracting T. lineatum to funnel traps. No

synergistic effects resulted from adding ethanol (release

rate = 75 mg/24 h) or c<_-pinene (release rate = 30 mg/24 h) , 41

alone or together, to the traps. This was followed with the

study reported here which attempted to determine the optimal

release rates of semiochemicals for use in wind tunnel bioassays (Baker and Linn, Jr. 1984). The study was

designed to compare the number of T. lineatum caught in

traps baited with ethanol and lineatin, using various

combinations and release rates.

3.2 Materials and Methods

The attraction of beetles to funnel traps baited with various release rates of ethanol and lineatin, was studied

in 1987 in a wind tunnel described in Chapter 2. Three 8-

funnel traps were placed in the upwind section of the tunnel, 0.5 m from the screen. The three traps provided a

larger plume of semiochemicals within the tunnel than could be achieved with a single trap.

Treatments for testing the response of T. lineatum in

the wind tunnel included 4 lineatin release rates (0, 8, 64,

and 512 ug/24 h), each tested in all combinations with 3

release rates of 95% ethanol (0, 75, and 150 mg/24 h). This

resulted in 12 treatment combinations of lineatin and

ethanol. The 0,0 release rate was the control treatment.

Slow-release lineatin lures were used, in the form of

Hereon controlled release dispensers (Kydonieus and Beroza

1981), with release rates based on lure size (7.0 ug/24 42

h/cm2)!2-. The lures were aged for one week at ambient

temperatures to allow the release rates to stabilize. Six

lures were used in each lineatin treatment. A lure was placed in the top and bottom of each trap. The size of each

lure used for release rates of 8, 64, and 512 ug/24 h, were

0.19, 1.51, and 12.2 cm2, respectively.

Ethanol (95%) was released from a 40 mL plastic

container with a 1 mm-diam-aperture (release rate = 75 mg/24

Ii h)1^. For a release rate of 75 mg/24h, one lure was placed

in the central trap, and for treatments of 150 mg/24 h, one

lure was placed in each of the outer traps. Different placement of ethanol may have influenced the odor plume, yet

the type of dispenser used limited me to this design.

Traps used for the control treatments were not the same

as those used for the baited treatments. To see if baited

traps carried residue, a comparative test was run to compare beetle response to these traps following bait removal, with the response to the control traps.

The beetles used in this study had been collected

during the spring of 1987 with lineatin-baited multiple-

funnel traps at the same field site used in Chapter 2.

Beetles were collected daily and placed in 1 L plastic

containers with moistened cloth towels. The containers were

stored in a walk-in cooler at 4°C under a 14:10 h (L:D) photoperiod. The containers were checked twice weekly to

|2_ Phero Tech. Inc., 1140 Clark Dr., Vancouver, B.C. V5L 3K3. 43 monitor moisture levels. These beetles survived for more than 100 days.

Prior to testing, the beetles were removed from the plastic containers and selected for testing based on their healthy appearance (i.e. presence of all body parts) and their ability to walk normally. The selected beetles were then placed in petri dishes with cloth towels and stored at

4°C until needed for the test, always on the same day. The beetles were given a warm-up period of 15 min at room temperature prior to release. The beetles were released

from a horizontal tray, 3 5 cm above the tunnel floor, 1.5 m downwind from the traps. The wind speed was held constant at 0.15 m/s (Chapter 2).

A 4 X 3 factorial experiment was set up as a randomized complete block design with days serving as blocks. Males and females were tested separately. Because of the large number of treatments, one replication was tested per day for each of 15 experimental days. For each replication, 25 beetles were released and allowed 10 min to respond. The percentage of beetles caught in the traps was used as the dependent variable. The arc sine transformed data were analyzed by analysis of variance and mean separation between treatments was carried out with the Student-Newman-Kuel's test (P < 0.05) (SAS 1985). 44

3.3 Results

The number of T. lineatum caught in funnel traps in the wind tunnel was not significantly affected by the presence

of ethanol either for males (F2 154 = 0.4; P > 0.05) or

females (F2 ^54 = 1.4; P > 0.05). In contrast, a

significant increase in the number of beetles caught

occurred in the presence of lineatin alone for both males

(Fig 8A) and females (Fig. 8B). Males responded best at the

lowest release rates of lineatin with an average trap catch

of 2 3.5 and 21.6% for the 8 and 64 ug/24 h release rates,

respectively. A significant reduction in response occurred

at 512 ug/24 h with a mean catch of 16.6% of the beetles

tested. Despite somewhat lower catches, similar responses were found for females, in which the highest catches, 20.0

and 16.1%, were obtained at 8 and 64 ug/24 h, respectively.

Mean catch at 512 ug/24 h was 14.3 %, a significantly lower

response than for the 8 ug treatment.

No significant differences in response by T. lineatum were observed between the control traps and unbaited traps

that had previously held ethanol and lineatin during the

experiment (F5 72 = 0.9; P > 0.05). Thus, there was no

evidence for contamination of the traps. 45

Fig. 8. Mean percent (±) S.E. of T. lineatum caught in traps in response to combinations of varied release rates of lineatin and ethanol in a wind tunnel: A) Males and B) Females. Lineatin release rate columns, pooled across ethanol treatments, with the different letters are significantly different (SNK; p < 0.05). 46

3.4 Discussion

Ethanol was hypothesized to act as an arrestant/boring

stimulant for beetles in close proximity to a suitable host

(McLean and Borden 1977). Therefore, since drainpipe traps

require that beetles land on the trap and crawl through

small diam holes in order to enter the trap (McLean et al.

1987; Vite and Bakke 1979), ethanol would appear to serve an

important function in capturing beetles. In contrast,

funnel traps capture beetles that fly into the traps, and

don't require them to land. This type of behavior would

explain our data, which suggests that there is no gain in

adding ethanol baits to funnel traps, as the lineatin alone

is sufficient for attracting and capturing T. lineatum.

It is possible that ethanol and oC-pinene may be

important in attracting T. lineatum to funnel traps, but

that the dispensers normally used for releasing them may

account for our results. This has been investigated by

Cushon (unpublished datajl3-. He has shown in preliminary

studies that newly developed closed-system host lures work

better than the old host lures in capturing T. 1ineatum.

However, further testing is needed to compare beetle capture

in funnel traps baited with lineatin plus new host lures to

traps baited with lineatin alone. The argument against the

use of the old ethanol lure (plastic container with the

[3 Mr. Geoff Cushon, Research Technician, Phero Tech. Inc., 114 0 Clark Dr., Vancouver. B.C. V5L 3K3. small hole in the cover) is that rain water gets into the container diluting the ethanol solution, reducing the attraction of the bait. Yet this is not applicable to this wind tunnel study. In addition, the same dispensers were

shown to be effective in enhancing the attraction of T.

lineatum in Europe as well as another ambrosia beetle species, Gnathotrichus sulcatus (LeConte) to sticky wire mesh traps (Borden et al. 1982).

Response of T. lineatum to various release rates of

lineatin has been previously examined in a field setting in

British Columbia by Lindgren et al. (1983). They found that the number of beetles caught on cylindrical sticky wire mesh traps baited with lineatin, increased as release rates increased, from 10-40 ug/24h, and remained the same between

40 and 800 ug/24h. In a later experiment described in the same paper, release rates of 40 ug/24h were found to be optimal for funnel traps, yet a release rate of only 10 ug/24 h of lineatin was adequate as long as the remaining lineatin (30 ug/24 h) was placed within 1.5 - 2 m of the trap. Our results in the wind tunnel with low release rates correspond well with those from Lindgren et al. (1983).

However, a decrease in response was observed in the wind tunnel at the higher release rates. The differences in sensitivity of the beetles at the higher rates may have resulted from artificial factors imposed by the wind tunnel, an enclosed environment with a constant, unidirectional air flow. Both factors resulted in continuous exposure of the 48 beetles to the pheromone, whereas in the field, pheromone plumes are often broken up by turbulence and wind shifts

(Fares et al. 1980), resulting in non-continuous exposure.

For the above reasons, as well as the potential of attracting beetles from a larger area, and because the lures weaken with age, it is probably best for mass-trapping in the field to keep the release rates of lineatin higher than for studying beetle flight behavior in the wind tunnel.

Host volatiles may well be important in initial host recognition by T. lineatum in a natural forest situation.

However, from the results in Salom and McLean (1988) and this chapter, lineatin baits alone appear to be sufficient for running an effective mass-trapping program with multiple-funnel traps around log booms and dryland sorts, where host volatiles are likely to be present anyway. If the new host lures are to be used in these programs, further studies should be carried out to verify and quantify their enhanced effect on beetle capture. 49

CHAPTER 4: WIND TUNNEL EVALUATION OF FLIGHT AND LANDING BEHAVIOR OF TRYPODENDRON LINEATUM (OLIVIER) IN RESPONSE TO A HOST ATTRACTANT AND AGGREGATION PHEROMONE

4.1 Introduction

The effect of olfactory stimuli on forest Coleoptera is

best understood for the bark and ambrosia beetles

(Scolytidae), where orientation to host material is

considered to be a set of sequential events (Borden et al.

1986). For the striped ambrosia beetle, Trypodendron

lineatum (Olivier), it is hypothesized that the pioneer

females cue into the host attractants (primary attractant phase), and upon successfully boring into a suitable log,

emit a pheromone that results in the mass attack of the logs by males and females (secondary attraction phase)

(Borden 1985). As mentioned in Chapter 1, the host

attractant for North American T. lineatum has been

identified as ethanol and the aggregation pheromone has been

identified, and is known to be a tri-cyclic ketal called

lineatin.

Over the past decade, development of a facile synthesis

for the pheromone, along with the evolution of trap technology (Lindgren 1983; Lindgren et al. 1983), has led to

semiochemical-based pest management strategies around log processing areas (Borden 1988b). In an effort to increase

efficiency of this program, investigators have looked at the

intra-specific response of T. lineatum to various 50 combinations of host attractants and its pheromone (Borden et al. 1982; Salom and McLean 1988). Inter-specific response of T. lineatum and the closely related

Gnathotrichus spp. to their respective pheromones and host attractants has also been investigated (Borden et al. 1981;

Shore and McLean 1983). Host attractants were concluded to be necessary components of the semiochemical bouquet for attracting North American T. lineatum by Shore and McLean

(1983) whereas they were not in two other studies (Borden et al. 1982; Salom and McLean 1988). This has resulted in some confusion as to the actual roles of these chemicals in

North American T. lineatum host finding, orientation, and attack. Adding to the confusion, studies conducted in

Europe have consistently shown that T. lineatum respond to host attractants alone (Bauer and Vite 1985) and respond synergistically to lineatin when host attractants are added to the bouquet (Borden et al. 1982; Paiva and Kiesel 1985;

Vite and Bakke 1985), regardless of the traps employed for capturing beetles (Borden et al. 1982; McLean et al. 1987).

In this chapter, I addressed the differences in results for North American T. lineatum by conducting a detailed behavioral study in a wind tunnel to examine beetle flight and landing response to olfactory stimuli. By conducting the study in the wind tunnel, I was able to control most factors and make detailed behavioral observations of individual beetles. 51

4.2 Materials and Methods

The study was conducted in a wind tunnel described in

Chapter 2, using a constant wind speed of 0.15 m/s, at a temperature inside the tunnel of 23 + 2°C. The beetles used

in this study were collected daily during May and June of

1987 and 1988, at the same log boom site described in

Chapter 2. Storage procedures were also similar to the description given in Chapter 2.

Traps

Three modified Norwegian drainpipe traps (Borregaard

Ind. Ltd., 1701 Sarpsborg, Norway) were set out at equal spacings across the upwind sector of the wind tunnel, ca.

0.6 m from the screen. This trap design was chosen over other designs because it was judged to best simulate a log.

External ridges projecting 3 mm from the trap surface, circle the full length of the trap concentrically, allowing beetles to land and subsequently walk on the trap, as they would on a log. Numerous holes (3 mm diam), punched between the ridges, simulate gallery holes in a log. They provide:

1) the openings needed for semiochemical emission, and 2) entrances into the trap for responding beetles. As a result, beetles that land on these traps are required to go through a searching phase to find the semiochemical source,

similar to a field situation where males search for mates 52 and females search for suitable boring sites (McLean et al.

1987; Vite and Bakke 1979).

Several modifications were made on these traps from those used in the field (Fig. 9). Traps were painted a light green-gold color to allow for easier observation of beetle movement near and on the trap. Lindgren et al.

(1983) found no differences in response by T. lineatum to funnel traps of 5 different colors, suggesting that painting our traps would not alter beetle response. The bottom section was fitted with a funnel that tightly connected the trap and the collecting jar.

The semiochemical plume emitted from the trap was

simulated using TiCl4 smoke. Initial tests indicated that a funnel fitted into the collecting jar and the trap cover lifted 1 cm above the top of the trap, achieved emission of the plume over the full length of the trap (Fig. 9A). The semiochemical dispensers were placed in the trap by a 1 cm - diam wooden dowel suspended from the trap cover (Fig. 9B).

The length of each trap was 1.1 m, producing a composite plume from the three traps which covered most of the volume within the tunnel.

Treatments

Traps unbaited (control), and baited with ethanol and lineatin, alone and in combination, served as the four treatments of study. In treatments containing ethanol, 95 % ethanol was released from two 40 ml dispensers (total B

Fig. 9. A) Modified Norwegian drainpipe traps used as a baiting substrate for T. lineatum. with a typical smoke plume pattern produced at 0.15 m/s. B) Close-up of trap cover and lure holder, showing an upper and lower lineatin bait, and a medial ethanol bait. 54 release rate = 150 mg/24 each placed in the center of the two outer traps. In treatments containing lineatin, the pheromone was dispensed from Hereon® controlled release devices (Kydonieus and Beroza 1981), where release rates were based on lure size (7 ug/24h/cm2)l2-. Two 0.76 cm2

lineatin lures were placed in the top and bottom of all three traps, resulting in a total release rate of 32 ug/24 h

(Chapter 3). All lures had been aged for one week at ambient temperatures to allow release rates to stabilize.

Removal of all traces of semiochemicals from the wind tunnel as well as the traps for the control treatments would have been very difficult to carry out during the run of the experiment in 1987. This led to a repeat of the control treatments only, during 1988, to test for evidence of contamination during the first set of unbaited control treatments. Prior to the second set of control treatments, the charcoal filter in the wind tunnel was reactivated, the old dust filter was replaced with a new one, the wind tunnel was completely cleaned with a 0.5% hypochlorite solution, and previously unused modified drainpipe traps were used.

All subsequent procedures described below were the same for both the first (all treatments) and second (controls only) experimental sets.

Test Procedure

Beetles were sexed and given a walking test in the morning prior to flight in the wind tunnel. They were 55 allowed to acclimate to the warmer room temperatures (22 -

25°C) for 15 min before release, a period sufficient to ensure maximal response of the beetles to the traps.

Beetles were released from the same platform at the same location as was used in the studies reported in Chapter

2. Two beetles, with males and females tested separately, were released for each treatment/replication. All treatments were selected randomly within each replication.

In 1987, a range of 1 to 4 replications were conducted/day over 24 experimental days, whereas in 1988, 8 to 12 replications of the second set of unbaited treatments were conducted over 6 experimental days. In total, 124 beetles of each sex were tested in each treatment.

The detailed observations of beetle response to the different treatments included; occurrence of flight, time to

flight take-off, take-off direction, flight type, overall

flight direction, lock-on behavior, landing on the traps, time to landing, and behavior on the traps. Flight take-off direction was scored as upwind, vertical, or downwind. Type of flight was scored as steady or erratic (see Chapter 2 for descriptions). Overall flight direction, upwind, downwind, or non-directional, was based on the predominant direction flown by the insect throughout the observation period. If beetles did not exhibit predominant up- or downtunnel flight, they were considered non-directional fliers. Lock- on behavior was also described in Chapter 2. For beetles landing on the traps, observations were made to see if they entered the traps, and if they did, how long it took. All observations were recorded on a cassette tape and later transcribed onto data sheets.

Response behavior data were summarized as frequencies and analyzed as described in Chapter 2. Response time data were analyzed with a General Linear Model (GLM) (SAS 1985), with subsequent mean separation between treatments carried out with the Student-Newman-Kuel's test (p < 0.05).

4.3 Results

No significant differences were found between the first and second year unbaited control treatments for all behaviors and time responses observed, except for flight take-off direction (Table 2). The different results in flight take-off direction did not apparently affect any other set of results. Therefore, the unbaited treatment results from year 1 were considered a valid control for the baited treatments for all behavior categories studied except flight take-off direction. In the latter case, both year 1 and year 2 unbaited treatment results are presented.

Occurrence of flight was found to be independent of semiochemical treatment and sex (Table 3). Although not significant, slightly more beetles flew in the presence of baited traps vs. unbaited traps. Females did exhibit a higher tendency to fly than males (X^ = 5.72; p < 0.05), although the difference could not be detected at individual 57

Table 2. Statistical comparison of Trypodendron lineatum responses to first and second year control treatments in a wind tunnel. The numbers of beetles released for each sex/treatment = 124.

Behavior Test Statistic1 Probability Observed (Degrees of Freedom) Level

Occurrence of Flight x2(l)= 0.69 0.4054

Time of Flight F(l,283) = °-01 0.9094

Flight Take-off x2(2) = 8-33 0.0155 *2 Direction

Flight Type x2(l)= 0,00 0.9451

Overall Flight x2(2) = 2,10 0.3495 Direction

Lock-on Response x2(2) = °-98 0.6111

Time Between Take-off 3 and Landing

Captured in Traps as x2(l)= 0,0° 0.9491 % of Beetles Locking-on

1 Degrees of freedom for the X2 statistic are a product of treatment (t-1) by response levels (r-1).

2 Significant probability level is < 0.05. 3 • Not enough responders to make time comparisons. 58

Table 3. Flight frequency of Trypodendron lineatum to different semiochemical baits in a wind tunnel. Number of beetles released for each sex in each bait treatment = 124.

Attractant % Flying Males Females

Unbaited 54.0 (67)2 64.5 (80)

Ethanol 63 .7 (79) 68.5 (85)

Lineatin 66.9 (83) 72.6 (90)

Lineatin + Ethanol 63.7 (79) 71.8 (89)

Significant differences in response were found between sex (X ; p < 0.05), but not among baits (X2; p > 0.05).

Number of flown beetles are presented in the parentheses. 59 treatment levels. Response time from release to initial take-off was found to be independent of treatment and sex

(Fig. 10). Mean take-off time for all treatments averaged

80 seconds.

Differences in upwind and vertical take-off directions were observed between the first and second year unbaited controls (Fig. 11). Upwind take-off was higher in year 1 and vertical take-off was higher in year 2 treatments. In year 1, no differences in upwind take-off direction were observed for any treatments. An increase in downwind take• off was observed in the presence of lineatin and ethanol together. These differences in responses between the 1987 and 1988 control treatments for this behavior did not appear to affect other aspects of T. lineatum behavior in the wind tunnel.

Frequencies of steady and erratic flight were influenced by the presence of semiochemicals, with no differences observed between males and females (Fig. 12).

Ethanol caused a significant increase in steady flight from

27 to 37% and a decrease in erratic flight from 72 to 63%, when compared to the unbaited treatment. With lineatin alone, the frequency of steady flight increased significantly to 48%. Ethanol combined with lineatin did not further enhance the occurrence of steady flight.

No differences between males and females were observed in overall flight direction (Fig. 13). Upwind flight direction increased significantly from 15 to 31% in the 120

110 rZ71 MALE 1X5 FEMALE 100 •—s CO 90 | 80 H 70 CO 60 O N

CO 50

2 40

30 1 20 10

0 UNBAITED ETHANOL LINEATIN LIN + ETH

SEMIOCHEMICAL BAITS

Fig. 10. Mean time interval + S.E. (indicated by a vertical line running through the o top of each bar) between release and flight take-off of T. lineatum. in the presence of semiochemical baits in drainpipe traps, within a wind tunnel. 100

(ZZ1 UNBAITED (1988) FxXl UNBAITED (1987)

UPWIND VERTICAL DOWNWIND

FLIGHT TAKE-OFF DIRECTION

Fig. 11. Flight take-off direction of T. lineatum. in the presence of semiochemical baits in drainpipe traps, within a wind tunnel. Male and p female data were pooled. Stars on top of unbaited (1988) bars denotes a significant difference (X ; p —< 0.05) with 2th e unbaited (1987) treatment. Bars for each direction, with different letters, indicates significant differences in response (X ; p < 0.05), for individual comparisons. The numbers of observations for each treatment were: unbaited (1988) - 137; unbaited (1987) - 138; ethanol - 146; lineatin - 150; and lineatin and ethanol - 152. 100 [771 UNBAITED 90 fV\l ETHANOL 80 W7A LINEATIN LIN + ETH 70

CO 60 O 0- co 50

40

30

CQ 20

10

0 STEADY ERRATIC

FLIGHT TYPE

Fig. 12. Observations of flight type exhibited by T. lineatum. in the presence of semiochemicals in drainpipe traps, within a wind tunnel. Male and female data were pooled. Proportions are based on the total number of observations, where each individual could be scored for more then one type of flight. Bars for each type of flight, with different letters, indicates significant differences in in response (X ; p < 0.05), for individual comparisons. The numbers of observations for each treatment were: unbaited - 140; ethanol - 152; lineatin - 158; and lineatin and ethanol - 156. [771 UNBAITED

UPWIND DOWNWIND NO DIRECTION

OVERALL FLIGHT DIRECTION

Overall flight direction exhibited by T. lineatum, in the presence of semiochemical baits in drainpipe traps, within a wind tunnel. Male and female data were pooled. Bars for each direction, with different letters, indicates significant differences in response (X ; p < 0.05), for individual comparisons. The numbers of observations for each treatment were : unbaited - 141; ethanol - 158; lineatin - 166; and lineatin and ethanol - 166. 64 presence of ethanol when compared to the unbaited traps, while a corresponding reduction in downwind flight from 67 to 50% occurred. Further increases in upwind flight occurred in the presence of lineatin, yet they were not statistically significant for ethanol alone. However, non- directional flight decreased significantly. Responses to ethanol and lineatin combined did not differ from responses to lineatin alone.

Positive lock-on response patterns differed between sexes when exposed to the different semiochemical treatments

(Fig. 14). The landing of males on drainpipe traps was not influenced significantly by the presence of ethanol, but with lineatin present, landing frequency increased from 6 to

27%. The number of beetles landing did not increase significantly when lineatin and ethanol were combined, but the frequency did rise to 34%. Lock-on response without landing did not appear to be affected by the presence of semiochemicals.

Landing by females was significantly greater in the presence of ethanol, with an increase in response from 1 to

12% when compared to female response to unbaited traps (Fig.

14). The presence of lineatin also significantly enhanced landing, with 16% of the flown females responding. The combination of ethanol with lineatin did not result in an

increased response over lineatin or ethanol alone. Non- landing positive responses increased slightly from 18 to 24% with ethanol present, and significantly increased over the a a 100 MALE z

o 50 - a ab ab to to* (/} FEMALE Z D 0 -i—i—i—i— —i—i—i—i— o U E L L+E CM U E L L+E

LOCK-ON WITH LOCK-ON WITHOUT NO RESPONSE LANDING LANDING

POSITIVE LOCK-ON RESPONSE

Fig. 14. Positive lock-on response, exhibited as landing or non-landing behavior, by T. lineatum in the presence of semiochemical baits in drainpipe traps, within a wind tunnel. Bars for each behavior and sex, with different letters, indicates significant differences in response (X2; p < 0.05), for individual comparisons. The numbers of males observed for each treatment were: unbaited (U) - 67; ethanol (E) - 79; lineatin (L) - 83; and lineatin and ethanol (L+E) - 79. The numbers of females observed were; U - 80; E - 85; L - 90; and L+E - 89. 66 unbaited control to 32% when only lineatin was present. The

combination of ethanol and lineatin did not improve this

response.

The time interval between take-off and landing on the trap also differed with semiochemical treatments, yet was

independent of sex (Fig. 15). The small number of ethanol

only responders would make it difficult to detect the presence of statistical differences between sexes, therefore male and female response times are presented as separate bars. Time response to ethanol-baited traps was

significantly shorter than lineatin-baited traps, averaging

81 and 191 s, respectively. Intermediate responses were observed when lineatin and ethanol were combined, with an average response time of 128 s, but not significantly different from ethanol.

The presence of lineatin was critical in inducing males to enter the traps. In terms of total beetles released, 12% were captured; of the beetles that flew, 18% were captured; and of the beetles that showed a positive lock-on response,

43% were captured (Table 4). Responses to ethanol were not significantly higher than responses to unbaited traps, and responses to lineatin and ethanol together were only slightly higher than lineatin alone.

The presence of ethanol and lineatin, however, were both found to be critical in stimulating females to enter the traps (Table 4). For traps baited with ethanol only, 7% of the total number of beetles released were captured, 9% of 280 260 1771 MALE 240 [\3 FEMALE 220 H CO 200

180 H

160

140

120

100

80

60

40

20 A 0

ETHANOL LINEATIN LIN + ETH

SEMIOCHEMICAL BAITS

Fig. 15. Mean time interval + S.E. (indicated by vertical lines running through the top of each bar) between flight take-off and trap landing of T. lineatum. in the presence of semiochemical baits in drainpipe traps, within a wind tunnel. Means were compared on combined data for the sexes, where different letters on top of the bars indicate significant differences between semiochemical baits (SNK; p < 0.05). Table 4. The capture of Trypodendron lineatum in semiochemical-baited drainpipe traps in a wind tunnel (N = 124 released/treatment /sex).

No. Beetles % of % of % of % of Beetles1 Attractant Entering Beetles Beetles Beetles Landing On a Trap Released That Flew Locking On a trap

Males

Unbaited 1 0.8 b2 1.5 b 9.1 b 100.0

Ethanol 3 2.4 b 3.8 b 13.0 b 60.0

Lineatin 15 12.1 a 18.1 a 42.9 a 71.4

Lin + Eth 19 15.3 a 24.1 a 47.5 a 70.4

Females

Unbaited 0 0.0 b 0.0 b 0.0 c 0.0

Ethanol 8 6.5 a 9.4 a 27.6 ab 80.0

Lineatin 7 5.6 a 7.8 a 16.3 b 50.0

Lin + Eth 13 10.5 a 14.6 a 37.1 a 81.3

Individual comparisons between treatments could not be carried out due to low numbers.

For each column, within each sex, percentages followed by different 2 letters are significantly different (X ; p < 0.05), for individual comparisons. the number that flew were captured, and 28% of the number that showed a positive lock-on response were captured.

Responses to traps baited solely with lineatin were very similar to those baited with ethanol alone. The proportion of beetles entering the traps, with respect to all categories in Table 4, was higher when ethanol and lineatin were combined compared to ethanol and lineatin alone, yet this increase was not statistically significant.

Although the time interval between landing on and then entering the traps was recorded, the number of observations were too few to be analyzed. No pattern could be discerned from the few beetles that were recorded.

4.4 Discussion

Neither flight frequency (Table 3) nor time of flight take-off (Fig. 10) were affected by the different treatments tested. The influence of semiochemicals on flight take-off direction could not be determined from our data, due to the varying results obtained between the unbaited controls in year 1 and year 2. The differences between the two control treatments (Fig. 11) are difficult to explain because of the lack of differences in response observed between these treatments for all other behavior categories tested.

Flight type was recorded to see if a characteristic behavior could be observed that was necessary for beetles to respond to semiochemicals. Increased steady and decreased 70 erratic flights were observed when beetles were exposed to semiochemicals (Fig. 12), suggesting that flight type may have potential in describing T. lineatum response in wind tunnel studies. Steady flight may indicate that beetles are in a receptive mode to respond to attractants. Although some beetles responded positively following a period of erratic flight, most positive responses came from beetles that initially flew in a steady manner. From field observations, I suspect that beetles exhibiting steady flight are unaffected by the wind tunnel, and those that exhibit erratic flight may be responding to the unnatural conditions imposed upon the insects by a constrained environment. Also these individuals might be the ones that are lost to nature. Flight type was not reported in the only detailed account of a study of a scolytid beetle in a wind tunnel (Choudhury and Kennedy 1980). Bioassays within wind tunnels that account for a vertical wind flow (Phelan and Miller 1982) may be needed to improve beetle response.

Males and females responded similarly in terms of overall flight direction (Fig. 13). Ethanol and lineatin were both important in influencing upwind flight of beetles to the baited traps, with lineatin causing a decrease in non-directional flight. While upwind flight response to ethanol was expected for females, as they are the pioneering sex for this species, it was also important with respect to upwind male response. Male response to ethanol did not carry over to any positive lock-on behavior, although a few more males did landed on ethanol-baited traps than on unbaited traps (Table

4). Lineatin was by far the most powerful compound to induce males to land on the traps (Fig. 14). By contrast, females required ethanol almost as much as lineatin to induce trap landing.

Beetles of both sexes responded in decreasing time to traps baited with lineatin alone, ethanol and lineatin together, and ethanol alone (Fig. 15). Ethanol plays a role not only in attracting females, but also to a lesser extent in attracting males to host material. Ethanol and lineatin were of equal importance in inducing females to enter traps, while only lineatin was effective for males (Table 4).

Lineatin was the critical compound for influencing males to fly upwind, land on, and enter the traps. For females, lineatin was more important in stimulating steady upwind flight than ethanol, however, both compounds were equally important in inducing landing on the trap. Trap capture increased slightly, yet not significantly, when both compounds were used together. These results suggest that the host stimulus may serve partially as a close-range attractant and the pheromone may serve as a long-range attractant. A similar hypothesis has been made for the host attractant and pheromone of the spruce bark beetle, Ips tvpographus L., based on electrophysiological studies

(Dickens 1981). It was found that I. typographus has a lower threshold and wider range of perception of pheromone concentration, compared to the concentration of the host attractantc<-pinene, indicating the insect's ability to perceive the pheromone at greater distances than the host volatiles.

The results of male and female T. lineatum responses to semiochemicals in the wind tunnel differ from field studies performed by Borden et al. (1982). These authors found that both male and female capture in drainpipe traps was inhibited by the presence of ethanol. However, the release rates of ethanol in their experiment was almost 7x higher than those reported here, not including additional ethanol volatiles emitted from nearby logging slash. In fact, they reported that high release rates of ethanol, at g/24 h, inhibited male response when compared to lower release rates, at 120 mg/24 h. In contrast, Shore and

McLean (1983) reported increases in male and female catches to drainpipe traps baited with ethanol, plus <*=-pinene

(another host volatile), and lineatin, when compared to traps baited solely with lineatin. However, Shore and

McLean (1983) did not test ethanol and oC-pinene separately.

Salom and McLean (1988) determined in both field and laboratory experiments that neither ethanol nor cxCrpinene were important in capturing T. lineatum in funnel traps.

However, beetles must enter holes for capture in drainpipe traps, and lack of this behavior requirement in funnel trap may account for the difference in the results between the two studies. This study reports the first detailed observations of flight behavior for this insect. Using optimal release rates (Chapter 3), the results confirm the hypothesis that primary females rely partially on the host attractant ethanol, while males, although not inhibited by ethanol, appear to rely mostly on lineatin in searching for female mates. Ethanol appears to speed response by both males and females (Fig. 15), resulting in less exposure of these beetles to predators and other mortality agents, during their spring dispersal flight period. If males were as attracted to ethanol as females are, they would possibly find alternative host material instead of homing in on tunnelling females who are ready to mate and produce the brood needed to perpetuate the population. While males cannot afford to search for potential suitable hosts, and take the chance of not finding a mate, females can afford to respond to alternative but equally suitable host material, since they produce the aggregation pheromone. 74

CHAPTER 5: INFLUENCE OF WIND ON THE SPRING PLIGHT OF TRYPODENDRON LINEATUM IN A SECOND-GROWTH CONIFEROUS FOREST

5.1 introduction

Investigations into insect dispersal and flight behavior have become important in forest pest management research during the past 2 5 years, due to the recognition of dispersal as an obligatory phase in the life cycle for many pests (Wellington 1983), and as an important process in their population dynamics and spread (Stinner et al. 1983).

For the striped ambrosia beetle, Trypodendron lineatum

(Olivier), many advances have been made in our knowledge of its biology over this time (Nijholt 1979). Nevertheless, a recent study on the occurrence of ambrosia beetle populations in commercial forest settings has emphasized the need to learn more about the dispersal dynamics of T. lineatum (McLean and Salom unpublished data). Quantitative information is needed concerning flight distance and direction of T. 1 ineatura, as it relates to weather patterns, topography, and other environmental factors. Such information can play an important role in developing a pest management program for T. lineatum in the forest, through modification of harvest planning and log inventory management, based on where beetle populations may be expected to disperse. 75

The importance of wind in influencing insect dispersal has been well summarized by Johnson (1969) and Pedgley

(1982). For T. lineatum, upwind flight towards attractive material was observed by Chapman and Kinghorn (1958) and documented from trap catches by Chapman (1962) in relation to an average prevailing wind. To improve our understanding of the pattern of T. lineatum spring dispersal, more information is needed to relate the dynamics of wind direction to that of beetle flight toward sources of semiochemical attraction. A mark-recapture study was undertaken to evaluate the influence of wind on T. lineatum dispersal to olfactory stimuli.

5.2 Materials and Methods

The mark-recapture study, consisting of three experiments, was conducted from June to August, 1986, in a dense, second-growth, closed canopy forest of Douglas-fir,

Pseudotsucra menziesii (Mirb.) Franco and western hemlock,

Tsuga heterophylla (Raf.) Sarg., naturally regenerated ca.

75 years ago following a clearcut and burn of the area

(Thompson 1985). The experimental site was located in a 140 ha stand on the University Endowment Lands (UEL) at the

University of British Columbia, Vancouver, B.C. It could only be accessed by small walking trails. This location was preferred over an industrial forest site where the presence of competing host material (logs) might have a confounding influence on the recapture of released beetles.

The first experiment in the study was designed to examine beetle dispersal over a 100 m radius (3.1 ha), where olfactory stimuli were present in all directions at three different distances from the release point. Twenty-eight multiple-funnel traps, baited with the aggregation pheromone lineatin, were set up in three concentric rings at distances of 5, 25, and 100 m from the release point, with 4, 8, and

16 traps placed at the three distances, respectively (Fig.

16). This experiment was replicated six times. The first hypothesis tested was that beetles fly upwind to olfactory stimuli. The second hypothesis tested was that most beetles would fly upwind to the closest source of olfactory stimuli.

Thus the frequency of catches was compared at the three distances.

The second experiment examined beetle dispersal at only a 100 m distance from the release point. This followed the removal of the 5 and 25 m traps. This experiment was replicated four times. The hypotheses tested were that beetle flight would still be upwind even though close-range attractants (ie. 5 and 25 m traps) were not present, and secondly, that a higher proportion of released beetles would fly the longer distance than when close-range attractants were in place, as in experiment 1.

The third experiment evaluated beetle dispersal at a

500 m radius from the release point, equaling 78.5 ha in 77

Fig. 16. Trap placement in mark-recapture studies of T. lineatum on the University Endowment lands in Vancouver, British Columbia. In the first experiment, 4, 8, and 16 lineatin-baited funnel traps were placed at distances of 5, 25, and 100 m, respectively, from the release point. Experiment 2 utilized the 100 m trapline only. In experiment 3, traps were set up 500 m from the release point only. The first replication had 4 traps placed in each cardinal direction, and the second and third replications has 2 traps at 45° intervals as shown here. 78 area. Two different plot designs were used. The first design had four traps in each of the four cardinal directions, 500 in from the release point. The second design, with two replications, had pairs of traps placed at

45° intervals around the circle, also at a 500 m distance.

For purposes of presentation, the two designs were combined

as one experiment. The hypothesis tested for this distance was that dispersal would be non-directional or downwind, but not upwind, because beetles would not perceive the pheromones 500 m from the release point.

Trypodendron lineatum were collected during their mass

flight period of May and June, 1986, using lineatin-baited multiple funnel traps, placed adjacent to a large log boom

storage area at the mouth of the north arm of the Fraser

River in Vancouver, B.C. Beetles were collected daily and placed in 1 L plastic containers with slightly moistened cloth towels at 4°C, to minimize their activity and preserve them for experimental use. This storage technique was found to be satisfactory, as mortality was generally low for the

first 50 days.

Beetles were marked with fluorescent powder (Ferro

Industrial Products Ltd., Surrey, B.C.) using the vacuum duster technique described by Linton et al. (1987). Marked beetles were placed in a cooler for the period following

application of the powder, and remained there until their

release in the forest, 2 - 3 h later. A brief experiment was run prior to this study, to look at the effect marking 79

T. lineatum with fluorescent powder may have on mortality

and short distance flight response to semiochemicals in a wind tunnel (Appendix 3). No effect of marking was observed

for up to one week following its application. Hence, the marking technique was considered satisfactory for use in this study.

The release-recapture procedure was the same for all three experiments. On flight days, beetles were released

from a shady site under the canopy usually at 12 00 h PDT or when air temperature rose above 15°C, the low temperature

flight threshold for T. lineatum (Chapman and Kinghorn

1958). Beetles, ranging in number from 1750 to 3650, were p placed on a 1 m wooden tray, with a small frame ridge, 0.2 m from the edge, effectively giving the beetles two ridges

from which to take-off. The tray was set 0.5 m above the ground. An absorbent paper, lined on one side with cellophane, was placed underneath the tray to catch any beetles that fell off. Similar paper was used as a bedding

for the inner square of the tray, where the beetles were placed.

An experiment was terminated 4 - 6 h following beetle release, when temperatures declined toward 15°C. The beetles were then collected from all the traps in the plot.

In an effort to see how many marked beetles flew for more than 1 day within the plot, traps were checked late in the afternoon of the second day post-release provided temperatures had exceeded the flight threshold of 15°C. 80

The beetles from the traps were brought back to the

laboratory for counting. An ultra-violet lamp was used to

aid in identifying the marked specimens. Unmarked beetles would sometimes have a little powder on them from contact with the marked beetles, yet the originally marked ones had

the powder somewhat impregnated on to their exoskeleton,

enabling easy distinction between marked and unmarked

beetles.

Various weather measurements were made while the

experiments were in progress. Wind direction and wind speed

data were collected at 10 min intervals, with the use of a bee-smoker placed over a circular grid 1 m in radius and broken up into 16 intervals of 22.5° (360° circle). The

spout of the smoker was orientated vertically at 0.5 m above

the forest floor. Wind speed was measured by timing the movement of a puff of smoke for a distance of 1 m. Wind parameters were measured 10 m east of the release tray.

Temperature measurements were also made every 10 minutes with a mercury thermometer set in the shade. Relative humidity was measured hourly with a sling psychrometer.

Barometric pressure data were obtained from the University

of British Columbia Climatological Station located ca. 1 km

from the study site.

Beetle catches and wind direction data are presented in the form of circular histograms. The statistical procedure described by Batschelet (1981), is summarized as follows.

Mean angles (0^) of beetle catch from the release point and 81

mean wind direction (#w) were determined using a rectangular

coordinate system with x and y axes and origin 0. The mean vector length (r) was calculated for each mean angle of beetle flight and wind direction. In unimodal distributions, r serves as a measure of concentration of the

sample points around the mean angle, and thus indirectly as

a measure of dispersion. The values of r range from 0 to 1.

As r increases, the dispersion of the data around the mean

angle decreases. Since these data were grouped into angular

intervals, it was necessary to apply a correction factor to

r, because without it, r would be biased and have a tendency to be too small. Use of r alone as a comparative statistic

is cautioned because the size of a population tends to affect r inversely.

The data for each circular distribution were subjected to the Rayleigh test for randomness. The probability of obtaining a vector length greater than the measured r is given by -nr p = e

(M.A. Stephens personal communication)4. where n equals the number of observations. In circular distributions where this probability (p) was < 0.05, the distributions were considered directed and the mean angle

(Jd) considered significant.

To determine if beetle flight direction differed from wind direction, a one-sample test, analogous to a one-sample

4 Dr. M.A. Stephens, Professor of Mathematics and Statistics, Simon Fraser University, Burnaby, B.C. 82 t-test for data on a linear scale (Zar 1984), was conducted using wind direction as a pre-assigned angle to be tested against the beetle flight direction angles. A

+ d-value was determined for each test angle, serving as 95% confidence intervals (Batschelet 1981). If the pre-assigned angle fell within the confidence limits of the flight angle, then the angles were not considered to be significantly different.

To determine differences in the number of beetles caught at the three distances in experiment 1, a two-way analysis of variance (ANOVA) was run (SAS 1985). A test between beetle catches at 100 m between experiments 1 and 2 was run using a one-way ANOVA (SAS 1985).

5.3 Results and Discussion

Weather

Of the weather measurements made during the study, aside from temperature, only wind direction seemed to have a major influence on beetle flight. Wind speed was light averaging 1.4 - 1.9 km/h day-1. Even on very windy days outside the forest, the wind speed at 0.5 m above the forest floor remained relatively light. This phenomenon is presented in detail by Pedgley (1982). Weighting individual wind directions by wind speed had a minimal effect on the

computation of the average wind direction (jdw) , as the ranges of wind speed observed within replicates were very 83 small. The range of standard deviations for wind speed over the 13 releases was 0.4 - 0.6 km/h. In addition, comparisons between wind direction over the first hour

following beetle release and the overall flight period were not perceptively different. Barometric pressure generally decreased during each day the study was run. Relative humidity consistently ranged between 70 - 90% on flight days. Finally, sky conditions were almost always sunny, which was necessary for the temperature in the forest to rise above 15°C. Temperature never exceeded 21°C, well below the high temperature threshold of 3 0°C reported by

Rudinsky and Daterman (1964).

Experiment 1

The percentage recapture, based on the number of beetles that flew successfully from the tray, averaged 24.9%

(Table 5). The averages were fairly consistent for the six replications. Less than 10% of the recaptured beetles were collected during the 24 h that followed the first beetle collection. The overall high rate of recapture may be partially attributed to the absence of competing host material in the study area. This can be contrasted with

Shore and McLean (1988), where sticky vane traps caught only

6% of the marked beetles, in a dryland sort containing a considerable supply of susceptible host material.

Directional data are presented individually as circular histograms (Fig. 17). Within each circle is a dotted line Table 5. Numbers of marked T.lineatum released and recaptured in the flight direction study, on the University Endowment Lands, Vancouver, B.C., 1986.

No. Marked Beetles Recaptured No. Beetles No. Beetles % of No. Exp. Rep. Released That Flew N That Flew

1 1 1 750 1 470 417 28.4 2 2 400 1 919 411 20.4 3 2 000 1 543 383 24.8 4 2 000 1 877 643 34.2 5 2 000 1 880 379 20.2 6 2 000 1 846 394 21.3 Totals 12 150 10 535 2 627 24.9

2 1 2 000 1 796 207 11.5 2 1 950 1 584 199 12.5 3 2 000 1 691 223 13.2 4 2 000 1 709 214 12.5 Totals 7 950 6 780 843 12.4

3 1 2 500 1 985 35 1.8 2 3 650 2 732 51 1.8 3 3 550 2 497 24 1.0 Totals 9 700 7 214 110 1.5 N CIRCULAR TRAPLINES WIND DIRECTION REP 5m 25m 100m

Fig. 17 Directional data based on frequency distributions of T. lineatum flight and wind observations for experiment 1 in the University Endowment Lands study. 86 representing the mean angle. For beetles, it is their mean direction of flight. Although it is not known whether beetles flew directly to the traps or not, for purposes of this presentation, beetle flight direction represents the resultant direction of beetle catch. For wind, it is its mean direction. The numbers in the circle represent the r- value. Significant r-values are noted with a star. The dotted lines and associated values outside the circle match

frequency with bar length.

In replicates 1-3 (Fig. 17), wind direction was westerly to southwesterly and beetle flight direction was upwind, to the west, at all three distances. However, beetle flight direction at 100 m in replicate 3 was not significant. When the wind direction changed to the southeast in replicate 4, flight direction at 5, 25, and 100 m, was to the southeast, not significant, and to the northeast, respectively. In replicate 5, neither wind direction nor beetle flight direction at 5 and 25 m were significant. In the last replicate, wind direction was southeasterly, as was beetle flight direction at 5 and 25 m, but not at 100 m.

The one-sample test for mean beetle flight direction was used for comparisons with the wind direction. For the five replications with a significant wind direction, beetle flight direction was significantly different at 5 m once, at

25 m four times, and at 100 m three times (Table 6).

However, at 25 m, the beetle flight direction was close to Table 6. Relationship of flight direction of T. Iineatum to wind direction, using a one-sample test for the mean directional flight angle, on the University Endowment Lands, Vancouver, B.C., 1986.

Flight Wind Distance Direction Direction Exp. Rep. (m) \ ± d1 \ vs.^3

5 259 + 7 NS' 1 1 25 284 + 9 260 * 100 261 + 18 NS

5 286 + 10 NS 1 2 25 251 + 14 281 * 100 264 + 18 NS

5 256 + 22 * 1 3 25 242 + 16 218 * 100 NS

5 105 + 25 NS 1 4 25 NS 128 100 48 + 16 *

5 NS 1 5 25 142 + 45 NS 100 322 + 37

5 151 + 28 NS 1 6 25 149 + 28 155 NS 100 14 + 50 *

2 1 100 NS NS 2 2 100 12 + 23 275 * 2 3 100 236 + 29 186 * 2 4 100 8 + 41 235 *

3 1 500 90 + 35 94 NS 3 2 500 65 ± 10 117 * 3 3 500 67 + 25 95 *

J&k + d = mean flight direction of beetles (rounded to the nearest deg.) + 95% C.L.; NS = not significant.

J?W = mean wind direction for experiments 1 and 2; and mean wind direction - 180° for experiment 3; NS = not significant.

Significance levels indicated: * = p < 0.05; NS = not significant. 88 the wind direction on all occasions, whereas at 100 m, when beetle flight direction was not significant, it tended to be very different from wind direction.

These data indicate upwind flight at 5 m, and reduced directionality at increasing distances from the release point. We must also consider that although wind was significantly directed in 5 of the 6 replications, beetle catches did occur in all directions. This suggests that some beetles were not responding immediately to the pheromones and flew non-directionally with respect to wind until they perceive or respond to the pheromone plumes.

The numbers of beetles caught at 5, 25, and 100 m from the release point did not differ significantly despite increased intertrap spacings of 8, 20, and 32 m, respectively (Table 7). This suggests that although some beetles are immediately attracted to olfactory stimuli, others preferred to fly before responding or they may not have encountered the plume until reaching the outer traps.

It must be noted that although flight exercise has been documented as a necessary prerequisite for T. lineatum to respond to semiochemicals during spring dispersal (Graham

1959; Bennett and Borden 1971), the beetles used in this study had previously flown and responded to pheromone-baited traps. Thus it is possible that these beetles responded more immediately to the pheromone than beetles just emerging from the forest duff on their first flight might have done.

This was assumed not to be critical in interpreting the 89 data, because a wind tunnel experiment described in Appendix

2, shows minimal differences between flown and unflown T.

lineatum in the response to semiochemical-baited funnel traps.

Experiment 2

Incidence of recapture dropped by 50% when the 5 and 25 m traps were removed from the plot (Table 5). Less than 8% of the recaptured beetles were collected during the 24 h

following the first collection. However, more beetles were caught at 100 m in this experiment than in the previous one

(Table 7). Without the presence of olfactory stimuli near the release site, a higher percentage of beetles flew

farther to get to attractive material.

Directional data show that wind direction and beetle

flight direction were both random in replicate 1 (Fig. 18).

In replicates 2, 3, and 4, wind direction and beetle flight direction were all significant, however the mean flight directions were neither upwind nor downwind. They were significantly different from each other in all three cases

(Table 6). Despite statistically significant beetle flight directions, close examination of the circular histograms in

Fig. 18 shows that the beetles were caught in all directions

for each of the four replications. In general, there was no correlation between beetle flight direction and wind direction. The random movement of beetles may be explained by the absence of pheromones near the release site and the 90

Table 7. Percent of T. lineatum caught at varied distances in the mark-recapture study, experiments 1 and 2, University Endowment Lands, Vancouver, B.C., 1986.

No. of Distance % Released Beetles Exp. Replicates (m) Recaptured + S.E.

5 8.1 + 1.5 a2 1 6 25 9.7 + 0.8 a 100 7.2 + 0.7 a B3

2 4 100 12.4 + 0.4 A

Data subjected to arc sine transformation.

Different lower case letters represent significant differences within an experiment (SNK; p < 0.05).

Different capital letters represent significant differences in beetle catch at 100 m between experiments (SNK; p < 0.05). 91

CIRCULAR TRAPLINES WIND DIRECTION EXP REP 100m 500m

Fig. 18 Directional data based on frequency distributions of T. lineatum flight and wind observations for experiments 2 and 3 in the University Endowments Land study. Dashed line in circles represents mean direction. Numbers in circles are the r-values. Those marked with a star are significant (p < 0.05). The dotted lines and associated values outside the circles are the scales for the frequency bars. 92

consistent presence of low wind speeds, which averaged 1.5 km/h throughout this experiment. This contrasts somewhat with Gara (1963), who investigated Ips confusus (Lec.) dispersal, also in the absence of olfactory stimuli near his

release site. At a forest site that experienced twice the wind speed as ours, he determined that the weaker flying

I.confusus generally flew downwind until responding to

olfactory stimuli, whereupon they oriented upwind.

Differences between the directional data for

experiments 1 and 2 at 100 m may be due to the influence of

the close-range traps in experiment 1. With upwind flight

recorded at 5 and 25 m, we can assume that beetles flying

further distances may have also been influenced initially to

fly upwind by their detection of the traps closer to the

release point. As a result, upwind flight was recorded in

replicates 1 and 2 of experiment 1, although the other replicates did not exhibit the same responses.

Experiment 3

The proportion of beetles recaptured (1.5%) was greatly reduced when traps were placed 500 m from the release point

(Table 5). Less than 10% of the recaptured beetles were collected during the 24 h following the first collection.

Although the distance probably played a major role in the

low numbers recaptured, another cause may have been the distances between the traps along the perimeter of the plot, 93

(800 m in replicate 1 and 400 m in replicates 2 and 3), which left large areas of open space.

The directional data for the beetles that were caught

showed strong evidence of downwind movement. In replicate

1, the wind was blowing from the west and the beetles were

caught to the east of the release point (Fig. 18). In

replicate 2, direction of wind came from the northwest while beetle catch was to the northeast. Wind direction in

replicate 3, was from the west and beetle catch was to the northeast. Beetle catch direction at 500 m in replicate 1 was not significantly different from wind direction, yet was

in replicates 2 and 3 (Table 6). However in the latter two replications, flight and wind directions were fairly close.

These data suggest that beetles caught at the 500 m traps in one day within the homogeneous forest setting, were ones that were flying with the wind. It is assumed from experiment 2 data, that beetle dispersal pattern within the

forest, where light wind speeds predominate, is non-directed with respect to wind without the presence of olfactory stimuli near the release site. Beetles flying crosswind or upwind probably could not make it as far as the ones carried or helped by the wind.

In conclusion, experiment 1 showed that spring dispersing Tj, lineatum fly upwind toward proximate olfactory stimuli (ie. 0 - 25 m) within the forest. Upwind flight at

25 and even 100 m may have been influenced by the 5 m traps.

Catches of beetles in all traps, however, suggested that 94

some beetles did not perceive or respond at take-off to

olfactory stimuli and flew in an unknown manner until they

did respond. At 100 m, the pattern of trapped beetles

appeared more random and inconsistent in relation to wind

direction at the release point. In experiment 2, where no

olfactory stimuli were present near the release site,

numbers of beetles caught at 100 m were not related to wind

direction. At 500 m, the pattern of beetle recapture

suggested that the few recaptured were assisted in their

flight by the wind. Non-directional wind movement

corresponded with non-directional flight both times that it

occurred in the study.

Graham (1959) first noticed that physical activity,

such as flight, was needed for spring emerging T. lineatum before they would respond to olfactory stimuli. He

suggested that the initial domination of positive phototactic behavior served as a mechanism that forced the

insects to disperse away from the emergence sites to find

susceptible host material. This behavior has also been noted for other scolytids (Henson 1962; Atkins 1966a;

Choudhury and Kennedy 1980). These results allowed me to build on Graham's hypothesis. In a closed canopy forest at wind speeds averaging less than 2 km/h, T. lineatum appeared to disperse in a diffuse pattern when pheromone sources were no closer than 100 m away from their point of release. With pheromone distances set 500 m from their point of release, the furthest flying beetles were influenced by the direction 95 of prevailing winds in the form of downwind displacement.

These observations suggest that in natural forest settings, where wind speed is consistently low, and suitable host material such as windfalls, were widely scattered, beetles will fly in a non-directional manner with respect to wind.

Upon perception of the pheromones as far as 25 m away, physiologically ready T. lineatum respond by flying upwind, suggesting chemo-anemotactic behavior. Their initial flight pattern as well as response to the aggregation pheromone, serves to minimize their expenditure of energy, while maximizing their chance of successfully finding and colonizing a normally scarce resource. 96

CHAPTER 6: DISPERSAL OF TRYPODENDRON LINEATUM (OLIVIER) WITHIN A VALLEY SETTING

6.1 Introduction

Examination of the movement of spring flying

Trypodendron lineatum (Olivier) in a second-growth forest

(Chapter 5) served as an important initial study of the dispersal pattern of this species. Using these results, I designed a second set of studies that examined population dispersal in a more typical setting. Consequently, the

study was moved to a valley, physiographically characteristic of the forests along the coast of British

Columbia (Pojar 1983). This valley had alternate patches of old-growth and regeneration forest sites that provided a good simulation of commercial forests in this region. As wind was shown to be a very important factor with respect to dispersal direction by T. lineatum (Chapters 2 and 5), this valley provided a unique opportunity for studying population movement, because of the characteristic diurnal up-valley wind patterns that are expected during the summer (Barry

1981).

The objectives of the study reported in this chapter were; 1) make meteorological comparisons between an open and

forested site within the valley; 2) test if short distance

flight (5 - 50 m) of T. lineatum in an old-growth forest is upwind; 3) ascertain long distance flight capabilities of 97 the beetles; and 4) survey their dispersal rate

(distance/time).

6.2 Materials and Methods

Site Parameters

Mark-recapture experiments were carried out in the

Cedar Creek Valley, of the Coquitlam Lake Watershed

District, in the lower coastal mainland of British Columbia.

It was a typical U-shaped valley with an up-valley azimuth

ranging from 48 to 68° (Fig. 19). A 3 km area served as our working area. Within the study area, the valley bottom

elevation dropped from 600 to 500 m. The peaks on both

sides of the valley reached 1200 m in elevation. The forest cover was predominantly over-mature western red cedar, Thuja plicata Donn. and western hemlock, Tsuga heterophylla (Raf.)

Sarg. As a result of stand breakdown, understory vegetation was very dense, consisting mostly of sapling stage amabilis

fir, Abies amabilis (Dougl.) Forbes. Clearcut areas 35 to

50 ha in size, regenerated 8 to 10 years ago, occur in a patch-like distribution along the lower slopes of the valley

(Fig. 19).

The experimental work was concentrated along the north

side of the creek. Moving up the valley, along the road,

the first opening on the north side spans 700 m in length.

It is followed by 700 m of old growth, then another 700 m

opening, a 200 m old growth patch, and then a 900 m opening, EXPERIMENTS ® RELEASE SITES(A.B) i WEATHER STATION(C)

2

ROAD OLD REGENERATION GROWTH CREEK

Fig. 19. A perspective view of Cedar Creek Valley, Coquitlam Lake Watershed, British Columbia. Five mark-recapture experiments for studying T. lineatum dispersal were carried out in this valley from June-August, 1987. The up-valley azimuth ranges from 48 to 68°. The road moving from the 00 southwest to the northeast covers a distance of 4 km in this view. Slope peaks reach 1200 m in elevation, while creek elevation averages 550 m. where the road then crosses the creek on to the southern

side of the valley into another large opening (Fig. 19).

The most recent logging (1986/1987) occurred 1 km southeast

of the down-valley edge of our working area, and thus was

not considered a competing source of attraction for the marked ambrosia beetles. However, recent windthrow was present along the eastern edges of the openings, and in a

survey made before conducting the experiments, ambrosia beetle attacks on the down trees were noted. This will be discussed later.

Multiple-funnel traps baited with lineatin, the

aggregation pheromone of T. lineatum, were used to capture beetles following their release. Two Hereon® controlled

release dispensers, with a total release rate of 340 ug/24 b h16-, were placed in each trap. Release sites and trap placement were installed according to experimental

objectives. Five experiments were conducted to test various hypotheses concerning T. lineatum response and flight toward pheromone-baited traps in forested and open sites.

Experiment 1

Beetle recapture direction was evaluated at distances

of 25 and 50 m from the release point with respect to wind,

in a dense old-growth forest setting. In Chapter 5, although upwind flight was observed at 25 m, the 5 m traps

in place may have had an influence on the 25 m catches.

This experiment corrects for that, with 25 m as the closest 100 source of attraction. The two circular traplines were set up at release site A (Fig. 19). At 25 m, 8 traps were placed at angular intervals of 45°, starting at an azimuth of 22.5°. At 50 m, 16 traps were placed at angular

intervals of 22.5°, starting at an azimuth of 0°. Three releases were run using this design.

Experiment 2

The low proportion of marked beetles recaptured in experiment 1 prompted experiment 2, which was designed to maximize beetle catch. The 25 and 50 m traplines were kept

in place, with an additional 4 traps placed 5 m from release point A, at angular intervals of 90°, starting at an azimuth of 0°. Meanwhile, a first attempt at long distance flight recapture was made with traplines of 10 traps each, placed

700 m west (trapline W) and 700 m east (trapline E-B) of release site A (Fig. 19). For both these lines, 4 traps were placed above the road and 6 below it, with each trap approximately 2 0 - 35 m apart. Three releases were run using this design.

Experiment 3

For experiments 3, 4, and 5, the release site (B, Fig.

19) was moved to the edge of the forest, near the road and a regeneration opening. This site was chosen to better simulate a beetle emergence site, since fall dispersing beetles favor forest edges for overwintering (Dyer and 101

Kinghorn 1961; Lindgren and Borden 1983). This site was

also less dense vegetatively than site A, possibly enhancing

the recapture of marked beetles.

In experiment 3, increased emphasis was placed on

assessing long distance flight capabilities, in relation to

the characteristic up-valley winds, which occurred

throughout the summer. A circular trapline was deployed 2 5 m from the release point B, using the same azimuth settings

as in experiments 1 and 2. Linear traplines were deployed

at the west edge, center, and east edge of the adjacent

opening, 50 m (trapline E-A), 375 m (trapline E-B), and 700 m (trapline E-C), east of release site B, respectively.

Trapline W from experiment 2 was also used, and was 1.0 km west of release site B. Four releases utilizing this design were carried out.

Experiment 4

Beetle flight dispersal outside the forest was the

focus of this experiment. Traplines W, E-A, and E-C were used, as well as three additional ones placed further up- valley. Two traplines (E-D and E-E) were set up 1.1 and 1.9 km east of release site B, respectively (Fig. 19). Line E-F was established on the south side of the valley ca. 1.3 km

from release site B. Four traps were placed above the road

and 2 below it for lines E-D and E-E. Three traps were placed above and below the road for line E-F. Only two

releases of this design were run in this experiment. 102

Experiment 5

This experiment had the same design as experiment 4

except that trapline E-A was removed. I thought that by

removing these traps, long distance flight would be better

facilitated. Unfortunately only one release was carried out

due to the lack of beetles.

Beetle Preparation

Adult T. lineatum used in this study were collected

during their mass flight period of May and June, 1987, using

lineatin-baited multiple funnel traps placed adjacent to a

large log boom storage area at the mouth of the north arm of

the Fraser River, in Vancouver, B.C. Beetle collection and

storage procedures were similar to those first discussed in

Chapter 2.

The procedures used for marking beetles with

fluorescent powder were described in Chapter 5. Four

colored powders were used throughout the summer, with a different color used for each successive release.

Release-Recapture

The release-recapture procedures used for all 5

experiments were very similar. Beetles were usually

released between 1200 and 1300 h PDT or when the temperature

rose above 15°C, the low temperature flight threshold for T.

lineatum. Between 2000 to 7000 beetles were placed on a 1 103 m wooden tray, with a small frame ridge 0.2 m from the edge, effectively giving the beetles two ridges from which to take-off. The tray was set 0.5 m above the ground. An absorbent paper, lined on one side with cellophane, was placed underneath the tray to catch any beetles that fell off. Similar paper was used as a bedding for the inner square of the tray, where the beetles were placed.

In experiments 1-3, two beetle collections were made

from the circular traplines during the day of release. For all of the releases but the first one, collections were made

2 hours after release and at the end of the day, usually 4 hours after release. The first release was 6 hours long with the collection interval occurring at 3 hours. In the experiments that utilized linear traplines (2 - 5), beetles were collected at the end of the release day. For all experiments, beetles were collected again prior to the next release (Day 2 catches). Day 2 trapping periods ranged from

2 to 12 days.

The beetles caught in the traps were brought back to the lab for counting, and examination under an ultra-violet lamp to identify marked specimens. The number of marked males and females and unmarked beetles were tallied. The number of released beetles not flying was also recorded. 104

Weather Data

Weather was monitored at the release sites as well as in an adjacent regeneration opening. At the open site (C)

(Fig. 19), a portable weather station was set up to record air temperature, relative humidity, wind direction, and wind speed continuously throughout the summer. The station consisted of a Campbell Scientific (C.S.) model 201 combination temperature (-35 to 50°C )/relative humidity (10 to 95%) probe, fitted with a gill radiation shield; a

Weathermeasure model 2 02 0 micro response wind vane (±2° accuracy; low threshold of 0.8 km/h); and a model 2030 micro response anemometer (+ 0.24 km/h accuracy; low threshold of

0.8 km/h). These data collecting devices were latched on to a crossarm, placed 2 m above the ground on a stainless steel tower. The data were collected in a C.S. CR21 micrologger.

The data from the micrologger were stored on a C.S. RC 235 cassette recorder. Measurement intervals were set at 10 min to coincide with manual weather measurements made at the release site. Data were read from the cassette tapes using a C.S. C 2 0 cassette interface into University of British

Columbia computer main frame files. Problems with this system on different occasions resulted in no weather data from the clearing for 3 of the 13 mark-recapture releases.

At the release site, measurements were carried out only on the day beetles were released. Wind direction and speed data were collected at 10 min intervals, with the use of a 105 bee-smoker placed over a circular grid 1 m in radius and broken up into 16 angular intervals of 22.5° (360° circle).

The spout of the smoker was oriented vertically, 0.5 m above the forest floor. Wind speed was measured by timing the movement of smoke for a distance of 1 m. Wind parameters were measured 10 m east of the release tray in experiments 1

and 2, and 10 m north of the release tray for experiments 3

to 5. Temperature measurements were taken every 10 min from

a mercury thermometer set in the shade. Relative humidity was measured hourly with a sling psychrometer.

Data Analysis

The circular traplines in experiments 1 to 3 and the wind measurements in all the experiments allow for the presentation of the beetle catches and wind direction in the

form of circular histograms. The statistical procedure used

is described in detail by Batschelet (1981), and is

summarized in Chapter 5.

To test whether beetle flight direction differed from wind direction, a one-sample test, analogous to a one-sample t-test for data on a linear scale (Zar 1984), was conducted using wind direction as a pre-assigned angle to be tested against the beetle flight direction angles. A + d-value was determined for each test angle, serving as 95% confidence

intervals (Batschelet 1981). If the pre-assigned angle fell within the confidence limits of the flight angle, the angles were not considered to be significantly different. 106

Frequency distribution comparisons were made with interval 1

and 2 data along with full day (total) data, however since

differences between individual intervals and total data for

a day were minimal, with respect to direction, the total

data comparisons only are presented.

A two-way analysis of variance (ANOVA) was run (SAS

1985) to test for differences in the number of beetles

caught at the different distances for experiments 1 and 2.

A test for beetle catches at 25 m between experiments 1, 2,

and 3 was run using a one-way ANOVA (SAS 1985).

6.3 Results and Discussion

Weather

The high pressure system that usually settles into the west coast during July and August, and brings in warm stable

air masses, did not come until the end of August in 1987.

As a result, the weather along the coastal mountains varied

from day to day, making it difficult to complete as many beetle releases as desired; however, 13 were completed.

During beetle release days, the meteorological measurements and observations made within forested and open

settings allowed for between site comparisons. Since

beetles were released only on warm days, the meteorological

comparisons between the different settings are only

appropriate for dry days in which the temperature in the

forest rose above 15°C for at least a couple of hours. For 107 these days, some interesting patterns are apparent. The temperatures in site A and B (less dense) averaged 3 and 2°C less, respectively, than in the open setting, regardless of the specific average temperatures for the day. As expected, relative humidity (R.H.) was higher under the canopy than in the open. Forest sites A and B averaged 2 0 and 10% higher

R.H., respectively, than the open site (C). While R.H. decreased during the day at the open site, it decreased on all 6 measured occasions at site A and on 3 of 5 occasions at site B.

Wind in the open site was significantly directed up- valley on all measured occasions (Table 8). The range of mean directions the wind blew from was 198 - 236°, while the valley direction toward Lake Coquitlam at that site is 228°.

Average wind speed differed little between all measured days, ranging from 5.0 - 6.3 km/h (Table 8). The variation in wind speed observed among release days was fairly small, with the range of standard deviations calculated to be 1.1 -

1.6 km/h.

In contrast, at site A, although variable throughout the day, average wind direction tended to be downslope toward the creek (Table 8). Significant (non-random) wind direction was noted on 3 of 6 test days. However this might be somewhat deceiving and will be discussed in a later section. Wind speed was light, averaging 1.1 - 1.3 km/h, slightly less than that at the release site in the second- growth, even-aged forest reported in Chapter 5, and ca. 0.2 Table 8. Comparison of wind direction and speed between forested and open sites within the Cedar Creek valley (up-valley azimuth of 48 to 68°) in the Coquitlam Lake watershed, B.C., 1987.

Forest Open Site

X Uindspeed X Windspeed

2 3 ExpK. Rel. Julian Site N ~9 r + S.O. N 0F r + S.D. c w c Date (Obs.) (km/h) (Obs.) (km/h)

1 1 169 A 34 4 0.305 1.37 + 0.46 33 236 0.723* 6.05 + 1.44 2 175 24 333 0.417* 1.39 + 0.37 24 209 0.842* ...4 3 177 26 355 0.426* 1.33 + 0.58 26 226 0.703* 5.94 + 1.26

2 1 181 A 27 334 0.281 1.12 + 0.60 27 219 0.802* 5.33 + 1.41 2 191 26 350 0.428* 1.35 + 0.56 -- ...... 3 195 27 262 0.195 1.08 + 0.59 27 211 0.726* 6.32 + 1.30

3 1 202 B 25 220 0.763* 1.40 + 0.70 24 198 0.831* 6.12 + 1.26 2 205 7 201 0.515 0.65 + 0.46 -- ...... 3 210 27 277 0.724* 1.23 + 0.46 -- ...... 4 216 26 251 0.724* 1.88 + 0.75 25 234 0.672* 6.01 + 1.49

4 1 218 B 26 241 0.706* 1.43 + 0.36 26 207 0.810* 5.04 + 1.13 2 230 24 228 0.730* 1.42 + 0.56 24 201 0.768* 6.24 + 1.57

5 1 233 B 24 231 0.660* 1.54 + 0.60 24 206 0.800* 5.51 + 1.44

I Site A was dense and far from any openings; site B was less dense and within 25 m of the road and 35 m from the clearcut. 2 — 0 represents mean wind direction. 3 w — Stars represent a significant 0 (Rayleigh's Test for Randomness; p < 0.05). 4 w - No data. 109

times the speed observed in the open site in this study.

The range of standard deviations for the 6 releases was 0.4

- 0.6 km/h, very similar to the second-growth forest site.

At site B, which was close to the road and to another

large opening, and in addition had a less dense understory

than site A, wind consistently blew from down- to up-valley

(Table 8). Average wind direction for 7 release days ranged

from 201 - 277°, and on 6 of these days, the direction was

significant. Wind speed was similar to that at site A and

averaged from 0.7 - 1.9 km/h, with standard deviations of

0.4 - 0.8 km/h. Weighting individual wind directions by wind speed had a minimal effect on the computation of mean

wind direction (j?w) in the first mark-recapture study

(Chapter 5). Thus it was assumed to be a similar case for

release sites A and B in the valley, since the wind speed

and its variation were so similar at all three sites.

Experiment 1

Beetle recapture averaged 6.4, 8.3, and 5.8% for releases 1, 2, and 3, respectively (Table 9), considerably less than from the first two experiments reported in Chapter

5. In the first 2 releases, ca. 80% of the beetles at both

25 and 50 m were recaptured during time interval 1 (Table

10). This changed for release 3, where ca. 1/2 of the beetles were caught during the first interval. Of all beetles recaptured, approximately 12% were recaptured during the next collection day following the day of release. This Table 9. Flight and recapture success of T. lineatum in the mark-recapture flight study at the Coquitlam Lake Watershed, B.C. (Day 1 and Day 2 data pooled), 1987.

Captured Beetle Totals Marked Unmarked Exp. Rel. No. % That % of Flown Released Flew NO. Beetles No.

1 1 2 000 81.5 105 6.4 3 569 2 2 340 81.5 158 8.3 2 490 3 2 175 86.0 109 5.8 2 092 Totals 6 515 83.0 372 6.9 8 151

2 1 2 400 88.3 179 8.4 1 238 2 3 990 53.6 234 10.9 4 110 3 2 000 85.2 187 10.9 847 Totals 8 390 71.1 600 10.1 6 195

3 1 2 400 48.8 83 7.1 1 820 2 5 000 76.8 501 13.1 828 3 5 000 61.3 192 6.3 1 322 4 4 600 67.8 178 5.3 287 Totals 17 000 65.8 954 8.5 4 057

4 1 4 000 73.9 85 2.9 1 120 2 7 000 53.5 92 2.5 46 Totals 11 000 60.9 177 2.6 1 166

5 1 4 900 81.6 17 0.4 191 Table 10. The number of marked T. Iineatum recaptured with respect to time in a dense old growth forest in the Coquitlam Lake Watershed, B.C., 1987.

No. Beetles Recaptured Exp. Rel. Release Time Traplines Site Interval1 5 m 25 m 50 m N 5? N%N

1 A 1 31 77.5 52 80.0 2 2 5.0 3 4.6 Day 2 7 17.5 10 15.4

2 A 1 50 80.6 79 82.3 2 8 12.9 15 15.6 Day 2 4 6.5 2 2.1

3 A 1 29 46.0 24 51.1 2 24 38.1 16 34.0 Day 2 10 15.9 7 14.9

2 1 A 1 57 74.0 15 51.7 21 39.6 2 16 20.8 14 48.3 20 37.7 Day 2 14 18.2 0 0.0 12 22.6

2 A 1 50 54.3 32 45.1 34 48.6 2 30 32.6 21 29.6 24 34.3 Day 2 12 13.0 18 25.4 12 17.1

3 A 1 32 52.5 33 53.2 32 50.0 2 12 19.7 17 27.4 20 31.3 Day 2 17 27.9 12 19.4 12 18.8

3 1 B 1 34 45.3 2 5 6.7 Day 2 36 48.0

2 B 1 340 80.9 2 -- -- Day 2 80 19.1

3 B 1 95 53.4 2 14 7.9 Day 2 69 38.8

4 B 1 124 75.2 2 16 9.7 Day 2 25 15.2

Time intervals: 1 = 1st half of day 1 ; 2 = final half of period; Day 2 = period until next collection (2 - 12 days later). Represents % of beetles caught at that distance. 112

indicates that at short distances, the majority of T.

lineatum that respond to nearby olfactory stimuli (ie. 25 -

50 m) do so within the first few hours of release.

The overall number of beetles recaptured did not differ

significantly between 25 and 50 m, with intertrap spacings

of 20 m for both traplines (Table 11). This is consistent with the data from Chapter 5, where an equal number of

beetles were captured at distances of 5 to 100 m from the

release site, indicating that while some beetles are

attracted to olfactory stimuli immediately, others fly

farther.

Directional data show that for all 3 releases, the mean

wind direction (0W) at Site A was from the north. R-values were significant for the latter 2 releases (Fig. 20).

Beetle flight direction (0^) a^ 25 m was non-directional for

all 3 releases, while at 50 m, 0^ was significantly directed

east to southeast for all 3 releases.

A one-sample test was used to compare 0^ at 50 m to £TW

in the forest for releases 2 and 3. For both releases, 0-^

differed significantly with #w (Table 12). Thus #b at both

25 and 50 m was not upwind. However, upon close examination

of wind direction histograms (Fig. 20) for releases 2 and 3,

it was noticed that the highest frequency of observations

occurred + 45° of the mean angle, with few observations

occurring at the mean. Therefore, despite a significant r- value, wind direction was quite variable, and must be

considered when interpreting these results. 113

Table 11. Mean percent of T. lineatum recaptured at varied distances in the Cedar Creek Valley, Coquitlam Lake Watershed, B.C., 1987.

Mean % of Release No. of Distance Flown Beetles Exp. Site Releases (m) Recaptured + S.E.

1 A 3 25 3.1 + 0.3 a2 B3

50 3.9 + 0.8 a

2 A 3 5 4.0 + 0.2 a

25 2.7 + 0.6 a B

50 3.1 + 0.4 a

3 B 4 25 7.1 + 1.3 A

1 Data subjected to arc sine transformation.

Distances within each experiment followed by different lower case letters are significantly different (SNK; p < 0.05).

Different capital letters, with respect to the number of beetles caught at 25 m for the 3 experiments, represents significant differences (SNK; p < 0.05). 114

Table 12. Comparison of T. lineatum flight and wind direction in an old growth forest in the Coquitlam Lake Watershed, B.C., 1987.

1 2 J Exp. Rel. Distance 0B + d 0W Significance (m) (p < 0.05)

1 25 NS NS 50 138 ± 38

2 25 NS 333 50 107 + 39 *

3 25 NS 355 50 103 + 37 *

1 5 91 ± 18 25 NS NS 50 NS

2 5 2 ± 28 NS 25 NS 350 50 13 ± 32 NS

3 5 94 ± 21 25 56 + 39 NS 50 61 ± 41

1 25 205 + 36 220 NS 2 25 193 + 17 NS 3 25 272 ± 15 277 NS 4 25 234 ± 15 251 *

Exps. 1 and 2 carried out at release site A and exp 3 at release site B.

0B ± d = mean flight direction of beetles (to the nearest degree) + 95% C.L.

0W = mean wind direction at the release site.

Difference between 0B and 0^: * = significant; NS = not significant. 115

CIRCULAR TRAPLINES WIND DIRECTION

EXP REL 5 m 25 m 50 m FOREST CLEAR CUT

1 1

2 1

3 1

Fig. 20. Directional data based on frequency distributions of T. lineatum flight and wind observations for experiments 1-3, in the Cedar Creek Valley, Coquitlam Lake Watershed, B.C., 1987. Line in circles represent mean direction. Numbers in circles represents the r-values. Those marked with a star are significant (p < 0.05). The number and associated line outside the circles are the scales for the frequency bars. 116

Experiment 2

Beetle recapture increased from an overall average of

6.9% in the first experiment to 10.1% in the second

experiment (Table 9), and must be attributed to the addition

of the four traps at 5 m. Beetle recapture with respect to time was fairly consistent for all 3 releases.

Approximately 1/2 of the beetles were recaptured during the

first half of the first day (time interval 1), with the

remaining catch being split between the second half of day 1

(time interval 2) and day 2 (Table 10). The overall number

of beetles recaptured did not differ significantly between

5, 25, and 50 m, with intertrap spacings of 8, 20, and 2 0 m, respectively (Table 11). Only 1 beetle was caught in each of the traplines W and E-B. Both catches occurred during the day of release.

A similar wind pattern was observed at the release site when compared to the previous experiment. However, only during the second release was wind significantly directed

(from the north) (Fig. 20). Beetle flight direction was

significantly directed for all three releases at 5 m, the third release at 25 m, and the second and third releases at

50 m. A one-sample test for comparing 0^ to Fw was made for the second release data since wind was significantly

directed for this release only. In this case, 0b was not

significantly different from 0^ at 5 and 50 m (Table 12).

Thus, upwind flight is indicated by these data. 117

Nevertheless, it should be noticed that the frequency of wind direction again surrounded the mean without showing many observations at the mean (Fig. 20).

In general it can be said that at release site A, wind

direction was extremely variable, and while the 10 minute

intervals may not have been fine enough to detect it in all

cases, wind shifts were constantly occurring. Under such

conditions, beetle flight direction would be expected to be

as variable as the wind itself, as these data seem to

indicate.

Experiment 3

Beetle recapture remained low at release site B and may be partially attributed to the reduced number of traps placed near the release site. Percent recapture averaged

8.5 through four releases, yet this was largely bolstered by

release 2 data (Table 9). The second release day stands out because 90 min following beetle release, a major rainstorm occurred. Wind speed measurements showed an average of only

0.7 km/h prior to the storm; the lowest measured over the

summer. Despite only a 90 minute flight time, the number of beetles recaptured during this release far surpassed the totals from all the other releases. Vite et al. (1964)

found that Dendroctonus frontalis Zimmerman activity was unusually high on cloudy days prior to the onset of a

rainstorm. Edson (1978) has similar results with the mountain pine beetle, Dendroctonus ponderosae Hopk. These 118

authors suggested that conditions within the canopy would be

more stable, trapping the pheromones in the forest. This

explanation may not be sufficient, however. Of the total

beetles caught in the traplines placed in the open settings

during the four releases (Fig. 21A), both of the beetles

caught in trapline W, as well as 80% of the beetles caught

on the eastern traplines (E-A, E-B, and E-C) were recovered

after the first 90 minute period. Unfortunately the weather

station malfunctioned at this time. Yet since wind

parameters at site B correlate well with open site wind

parameters (Table 8), it can be assumed that wind speed was

greatly reduced in the open. At low wind speeds, T.

lineatum prefer to fly (Rudinsky and Daterman 19 64) and

respond best to semiochemicals (Chapter 2). The combination

of increased flight capabilities along with a more apparent

or constant pheromone plume may explain the increase in

response. Other explanations for the increased response may

include changes in atmospheric pressure (Haufe 1954) or

atmospheric electricity (Edwards 1960), typical of meteorological conditions prior to a rainstorm. However

these factors were neither measured nor tested.

Beetle recapture during time interval 1 ranged from 45

to 81% of all beetles recaptured (Table 10). Interval 2

catches were relatively low, ranging from 7 to 10%, while

day 2 catches increased in importance relative to the two previous experiments, which varied from 15 to 48%. The mean

% of beetles recaptured at 25 m was significantly higher 119 than in experiments 1 and 2 (Table 11). However, high

catches for the second release were probably most

responsible for this. Nevertheless, increased catches at 25 m may have been partly due to the reduction in the amount of understory vegetation, as well as the absence of the traps at 5 m used in experiment 2.

Directional data at release site B showed that wind direction was consistently southwesterly and significantly directed for releases 1, 3, and 4 (Fig. 20). For these 3

releases, beetle flight was significantly directed. One-

sample comparisons showed that 0b did not differ from 0W during releases 1 and 3, while during release 4, the difference was only by 2° (Table 12). It is apparent that when wind is strongly directed as it was at site B, then beetle flight direction is definitely upwind at 25 m. These results support those from Chapter 5, where upwind flight at

25 m may have been influenced by the 5 m traps.

During this experiment, T. lineatum showed its ability to fly long distances, even when surrounded by traps at close range. On the days of release, 2 beetles were recaptured 1 km down-valley (line W) across a large forest setting (Fig. 21A). Aside from the 25 m traps, the highest number of beetles (52) were recaptured at the next closest trapline, E-A. Only 6 were caught in line E-B traps, located in the middle of the 700 m long regeneration opening. Ten beetles were caught 325 m further up-valley in line E-C traps. This pattern held for day 2 captures as 120

A TOTAL -2 52 6 10 MALE - 1 40 2 6 FEMALE =1 12 4 4

WEST E-A E-B E-C

TRAPLINES

® Beetle release site with 8 traps placed in a circular trapline 25 m from the point Number caught represents beetles caught in these traps.

21. Recapture of marked T. Lineatum in pheromone traps, for experiment 3, in the Cedar Creek Valley, Coquitlam Lake Uatershed, B.C., 1987. Traps were placed circularly around the release point and in traplines 50, 375, and 700 m east and 1000 m west of it. These catches are based on 4 releases where a total of 11 196 beetles flew. A) The total number of beetles recaptured on the day of release; B) The number of beetles recaptured during the next collection date (2 - 10 days later). 121 well. Four beetles were caught in line W (Fig. 21B). Line

E-A traps again recaptured the most beetles with a sharp

decrease in line E-B, and then an increase in line E-C.

Three points should be noted: 1) T. lineatum flew down- valley 1 km, despite strong up-valley winds throughout the

day. Although long distance flight was expected to be

strictly downwind (Chapter 5), it now appears that beetles

are capable flying against the wind at distances as great as

1 km. ; 2) at distances > 375 m, as many or more T. lineatum were caught in day 2 periods than during the day of release;

and 3) fewer beetles were caught in the middle of the

regeneration opening than at the edge 325 m further away

(line E-C traps). The latter point may be explained by the

effect of a large opening on wind direction. Kimmins (1987)

explained that laminar air flow over a cool forest canopy will become turbulent upon contact with a large area where the surface temperatures are warmer and more variable. This turbulence causes air to move upwards in the middle of the opening. Air movement will then become increasingly laminar

as it approaches the next forest canopy. This pattern of

air movement was also observed at our work site when numerous lightweight seeds of fireweed, Epilobium

angustifolium L., were observed in the air, and their dispersal by the wind was monitored over the open setting in which traplines E-A, E-B, and E-C were deployed. In the middle of this setting, seeds 10 - 30 m above the ground were moving in all directions, whereas seeds observed 122

further along the valley, near the next forest canopy (near

line E-C) were being taken up-valley in the strong winds.

Experiment 4

In this experiment, the number of beetles recaptured decreased considerably with the removal of the 25 m traps

from experiment 3 (Table 9). However, more beetles were

recaptured in the 50 m traps (E-A) in experiment 4 (79 beetles/release) (Fig. 22) than in experiment 3 (20 beetles/release) (Fig. 21). This supports data from Chapter

5, where the frequency of long distance flight increased when attractants placed close to the release site were removed. Sixty percent of the beetles recaptured in the 50 m traps were caught during the day of release (Fig. 22). At the longer distances, all beetles were recaptured during the day 2 periods. Although three beetles were recaptured (2 in

line E-D traps and 1 in line E-F traps) during the day of release, they were from experiment 3. Day 2 periods were 3 and 5 days, respectively, for releases 1 and 2. Overall,

1 beetle was caught 1 km down-valley (W), 8 were caught 700 m up-valley (E-C), 9 were caught 1.1 km up-valley (E-D), 1 was caught 1.9 km up-valley (E-E), and 1 was caught across the valley (E-F), 1.3 km away from the site. 123

A TOTAL = 0 93 0 0 0 0 MALE = 0 73 0 0 0 0 FEMALE = 0 20 0 0 0 0

TRAPLINES

® Beetle Release Site

Fig. 22. Recapture of marked T. Lineatum in pheromone traps for experiment 4, in the Cedar Creek Valley, Coquitlam Lake Watershed, B.C., 1987. Traplines were set up 50, 700, 1100, 1900 m east and 1000 m west of the release point. Additionally, line E-F was placed on the other side of the valley, ca. 1300 m east of the release point. These catches are based on 2 releases where a total of 6 698 beetles flew. A) The number of beetles recaptured on the day of release; B) The number of beetles recaptured during the next collection date (3-5 days later). 124

Experiment 5

Only 17 beetles or 0.4% of the beetles that flew (Table

9) were recaptured following the one release. Removal of

trapline E-A was a large reason for this. During the two

releases from experiment 4, a total of 19 beetles were

recaptured in lines W, E-C, E-D, E-E, and E-F, whereas in

the only release in this experiment, 17 beetles were

recaptured in those same traplines. Thus, again more

beetles were flying a longer distance. On the day of

release, 2 beetles were trapped on line E-C and 1 made it to

line E-D. The rest of the beetles, 1 in line W, 6 in E-C, 5

in E-D, 1 in E-E, and 1 in E-F, were caught during the day 2

period.

Population Dispersal

In experiment 4, the total number of beetles recaptured

at distances > 700 m up-valley was 19, and at distances of >

1 km, the number caught was 14. This is 10.6 and 7.8%,

respectively, of the 180 marked beetles recaptured during the two releases (Table 9). If a population estimate could be made for the area, the number of beetles flying those distances could be estimated. Therefore population

estimates for the forested patch containing release sites A

and B (Fig. 19) were made based on the mark-recapture data

from experiments 1 and 2. These experiments were used because they were closest to the peak emergence time of T. 125

lineatum. and trap arrangement was designed to capture as

many beetles within as close an area as possible. Data from

traplines W and E-B in experiment 2 were not used in making

the estimates, because they extended fay beyond the forested

release area.

The Lincoln index for population estimation was used:

N = an/r where N = population estimate a = # of marked beetles that flew r = # of marked beetles recaptured n = total # of beetles captured

The variance was determined as s2 = a2n(n-r)/r3 and 75%

confidence limits (C.L.) as N + 2s (Southwood 1978). It is

assumed (although not truly known) that those beetles

recaptured at release site A come from an area equivalent to

the average flight range of the beetles up to the point of

attack. Total population estimates from such an area for

the 3 releases in experiment 1 are 118 000 + 12 000 beetles,

and for the 3 releases in the second experiment, 28 000 + 2

000 beetles. This is a total of ca. 146 000 beetles. It is

a conservative population estimate, since calculations were made on beetles that had flown over a 1 month period that probably followed the earlier peak emergence days of April

and May. Based on the assumption that uncaught beetles

behaved similarly to caught beetles, with respect to

distance, we can estimate that 15 400 beetles can fly to an

attractive host 700 m from their emergence site. It can

also be said that 11 400 beetles can fly to an attractive 126 host at least 1 km away. Such high numbers should be a

concern to those involved in harvest planning operations.

In conclusion, observed meteorological differences between the open site and forested sites A and B occurred as

expected. Forest cover reduces sunlight resulting in a

decrease in temperature with an increase in relative humidity. Despite strong, consistent down- to up-valley winds during sunny days, at the open site, wind direction at

release site A was quite variable, while at site B, with

less vegetation and closer to the valley influence (via the

road), wind direction was strongly directed from down- to up-valley.

T. lineatum flight direction at site A was mostly variable as was the wind direction. The dense understory may have had a negative effect on the number of beetles recaptured (Shore and McLean 1984), either directly or

indirectly due to its influence on the wind which in turn

influences the pheromone plume. Another factor that may have had an influence on beetle recapture was the presence of susceptible windthrown material at the eastern edges of the regeneration sites. No sample was made to see whether marked beetles were attracted to these sites.

Experiment 2 data showed that few beetles, emerging

from forests far away from any openings, would travel long distances to attractive material. However this should be tested further without attractants present near the release

site. 127

At release site B, where wind direction was generally unidirectional, beetle recapture was upwind at a distance of

25 m. T. lineatum showed its ability to fly relatively long distances in this experiment with most beetles caught in short-distance traps. More beetles were captured at a farther distance, at the edge of the opening than at a closer distance in the middle of it, indicating the strong effect of wind turbulence in an open setting on beetle response to attractive material.

A shift in response time was noted in this study. In experiments 1-3, the majority of beetles captured at 50 m or less occurred within the first 2 hours of release, yet at distances ranging from 350 - 700 m, as many beetles were caught during the days following the release as were caught during the day of release. In experiments 4 and 5 most recaptured beetles at distances > 1 km from the release site were caught after Day 1. This could have been due to the increased travel distance, the weakened state of the stored insects, or a combination of both.

Only 21 of the beetles released in experiments 4 and 5 were captured up-valley at distances > 1 km, with 3 beetles recaptured as far as 1.9 km. However, based on the population dispersal estimates above, a substantial number of beetles are capable of flying at least 1 km from sites with large numbers of overwintering beetles. Furniss and

Furniss (1972) have reported finding T. lineatum far from host material, near mountain tops along the west coast of 128 the United States. Dyer (1961) reported finding a marked beetle 4 km from its release site, one year later. This valley study represents more substantial evidence of the ability of T. lineatum to fly or be taken by the wind considerable distances.

Trap placement during this study was concentrated up- valley from the release sites in an unbalanced design, based on evidence from Chapter 5. Beetle recapture 1 km down- valley from the release site, against the wind, may warrant a re-evaluation of the hypothesis that long distance flight by T. lineatum is primarily downwind. This would require a more balanced design in future studies to better characterize flight patterns of T. lineatum in the absence of attractants close to the beetle release sites. 129

CHAPTER 7: POPULATION MOVEMENT PATTERNS OF TRYPODENDRON LINEATUM WITH RESPECT TO WIND AND VEGETATION WITHIN A VALLEY SETTING.

7.l Introduction

Beetle catch data from the first mark-recapture study, held in the University Endowment Lands (Chap. 5), suggested that long distance beetle flight would be consistently downwind from a release point or emergence site. This led to plot designs for recapturing beetles that favored downwind flight (up-valley in most cases) by T. lineatum in the first valley study (Chap. 6). The one trapline that was placed 1 km upwind (down-valley) from the release point, recaptured 8 beetles, which when compared to the numbers of beetles recaptured downwind, was unexpectedly high. Thus, in this chapter a second valley study is reported in which the objectives were: 1) compare T. lineatum recapture pattern to wind direction without any trapping bias and without attractants near the release sites that could potentially influence beetle flight direction; 2) as a follow-up to the first valley study, where T. lineatum recapture tended to be lowest in the center of open settings, beetle recapture patterns between open and forest settings were compared; and 3) an attempt was made to test whether T. lineatum tend to disperse along certain routes in the forest, ie., the operi area, or along forest edges such as rights-of-way or roads. 130

7.2 Materials and Methods

Site Parameters

Two mark-recapture experiments were carried out in the

Cedar Creek Valley of the Coquitlam Lake Watershed, in the lower coastal mainland of British Columbia, from June to

July, 1988. A detailed description of the working site is given in Chapter 6.

For both experiments, beetles were released simultaneously from two sites on the north side of Cedar

Creek (Fig. 23). Site A was located on the down-valley edge of a forest setting, and site B was located 650 m up-valley at the opposing edge of the same forest setting. This was thought to be superior to the one release site used in the first valley study (Chapter 6), because it allowed for several hypotheses of T. lineatum dispersal to be tested without biasing trap placement up- or down-valley. Also it allowed again for the release of beetles from the forest edges, which is where most T. lineatum are known to overwinter (Dyer and Kinghorn 1961). Beetles released from site A are hereafter designated as population A, while beetles released from site B are designated as population B.

Groups of four semiochemical-baited funnel traps were used as sources of attraction for recapturing marked beetles. Funnel traps within each group were set up at the corners of a square plot, 10 x 10 m in area, in an effort to 131

Fig. 23 Contour map of the Cedar Creek Valley in the Coquitlam Lake Watershed, with the designs of the first (A) and second (B) mark-recapture experiments, carried out in 1988. Dotted polygons represent regeneration openings surrounded by old-growth forest. The up-valley azimuth = 48 - 68°. 1321 produce a larger active space of semiochemical plume than produced by one trap, to increase the level of attraction of

T. lineatum to the designated sites in which traps were deployed (Lindgren et al. 1983). All four traps within each group were baited with lineatin, the aggregation pheromone of T. lineatum. Two Hereon® controlled release dispensers with a total release rate of 340 ug/24 were placed in each trap. Although ethanol, the host attractant for T. lineatum, is not considered effective in funnel traps

(Chapter 2, Salom and McLean 1988), it was included in two of the traps for each group. This was due in large part to a new, and potentially better dispenser, in the form of a

2.5 by 41 cm poly-vinyl chloride bag, filled with 15 ml of

95% ethanol. For each trap baited with ethanol, one dispenser was used, resulting in a release rate of 30 mg/24 h over a period of 4 5 days'2-.

Experiment 1

This experiment, which was replicated four times, was designed to evaluate long distance T. lineatum flight within the valley in the absence of attractants near the beetle release sites. Groups of funnel traps were set up in a symmetrical design up, down, and across the valley, with respect to release sites A and B (Fig. 23A). No funnel trap group was set any closer to a release site than 200 m. The first hypothesis tested was that in the absence of attractants near the release site, most of the released 133 beetles would be recaptured downwind. The plot design allowed for two ways to test this (Fig. 23A). First, the number of population A beetles recaptured in the seven trap groups (7 - 13), set up-valley from site A, could be compared to the number of population B beetles recaptured in an equal number of trap groups (1 - 7), set down-valley from site B. Secondly, the number of population A beetles recaptured in the five trap groups (1 - 5), set down-valley from site A, with the number of population B beetles recaptured in an equal number of trap groups (9 - 13), set up-valley from site B. The symmetrical placement of trap groups relative to sites A and B resulted in a mirror image that made these comparisons valid. A second hypothesis tested was that beetle recapture numbers would be different in an open setting when compared to a forested one, despite the unequal number of trap groups representing both treatments. Four groups (1, 2, 12, and 13) were set within the forest and nine groups (3 - 11) were set up in three separate regeneration openings; one up-valley, one down- valley, and one across the valley, from both release sites

(Fig. 23A).

Experiment 2

This experiment, which was also replicated four times, was designed to test what type of flying routes are preferred by dispersing T. lineatum. One approach was to compare the number of up- (population A) or down-valley 134

(population B) beetles trapped in the forest, 30 m above the

road (8), on the road (7), and in the forest, 100 m below

the road (6) (Fig. 23B). This is based on the hypothesis

that dispersing beetles use rights-of-way roads as

"flyways", making cut timber along these paths especially

susceptible to beetle attack (J.A. McLean, personal

communication)!5-. A second approach examined the route taken

by beetles flying toward traps farthest from the release

sites (1, 2, 11, and 12). Beetle catches were compared

between open trap groups (3 and 10) and forest trap groups

(4 and 9), placed in between the release sites and the long

distance trap groups (Fig. 23B). As in the first

experiment, long distance flight determinations, with

respect to wind were examined, without the presence of

semiochemicals near the beetle release sites. Again trap

groups were placed symmetrically up- and down-valley from

release sites A and B, and allowed for two tests. First,

the number of population A beetles recaptured in the eight

trap groups (5 - 12), set up-valley from site A, could be

compared to the number of population B beetles recaptured in

an equal number of trap groups (1 - 8), set down-valley from

site B. Secondly, the number of population A beetles

recaptured in the four trap groups (1 - 4), down-valley from

site A, was compared to the number of population B beetles

recaptured in an equal number of trap groups (9 - 12), up-

valley from site B.

|_5_ Dr. J.A. McLean, Professor, Department of Forest Sciences, University of British Columbia, Vancouver, B.C. 135

Mark-Recapture

Adult T. lineatum used in this study were collected during their mass flight period of April and May, 1988.

Beetles were captured using lineatin- and ethanol-baited funnel traps, placed adjacent to a large log boom storage area at the mouth of the north arm of the Fraser River in

Vancouver, B.C. Beetle collection and storage procedures were similar to those first discussed in Chapter 2.

The procedures used for marking beetles with fluorescent powder were described in Chapter 5. This included a description of a brief wind tunnel study showing no variation in response to semiochemicals and no increase in mortality one week following marking (Appendix 3). For the 1988 study, four colored powders were used throughout the study, with different colors used for population A and B for each release.

The release-recapture procedures for both experiments were very similar. On flight days, beetles were usually released from 1130 and 13 3 0 h PDT, or when the temperature rose above 15°C, the low temperature flight threshold for T. lineatum (Chapman and Kinghorn 1958). For each day of release, 8000 beetles were released simultaneously from both sites A (4000) and B (4000) . At each site, the beetles were placed on a 1 m wooden tray, with a small frame ridge 0.2 m from the edge, effectively giving the beetles two ridges from which to take-off. The traps were set 0.5 m above the 136 ground. An absorbent paper, lined on one side with cellophane, was placed underneath the tray to catch any beetles that fell off. Similar paper was used as a bedding from the inner square of the tray, where the beetles were placed.

Collections from all traps were generally initiated 4 -

5 h following beetle release. In a continuing effort to see how many marked beetles flew for more than 1 day within the valley, collections were again made during the morning prior to the next release.

The beetles caught in the traps were brought back to the lab for counting, and an ultraviolet lamp was used to identify marked specimens. The number of marked males and females were tallied, as well as the number of unmarked beetles collected. The number of released beetles not flying was also recorded.

Weather Parameters

Meteorological measurements were made at the release sites as well as in an adjacent opening. At the open site

(Fig. 23), a portable weather station (W.S.) was set up to record air temperature, relative humidity, wind direction, and wind speed continuously throughout the summer. Details of the weather equipment were described in Chapter 6. In an effort to increase the precision of these data, 5 min measurement intervals were used, compared with the 10 min 137 intervals used in the first two mark-recapture studies

(Chapters 5 and 6).

At each release site, measurements were carried out only while the experiments were being actually run.

Temperature measurements were made every 5 min with a mercury thermometer and relative humidity was measured hourly with a sling psychrometer. Wind direction and speed data were collected at 5 min intervals, with the use of a bee-smoker placed over a circular grid 1 m in radius and broken up into 16 angular intervals of 22.5° (360° circle).

The spout of the smoker was oriented vertically, 0.5 m above the forest floor. Wind speed was measured by timing the movement of a puff of smoke for a distance of 1 m. The circular grid was placed 10 m north of the release tray at both sites A and B.

Data Analysis

The wind measurements from all three sites allowed for the presentation of wind direction in the form of circular histograms. The statistical procedures used are described by Batschelet (1981), and are summarized in Chapter 5. For

the wind data in this study, mean wind directions (0W) were determined for the first half, second half, and overall measurement period of a release day, for all eight release days.

Non-parametric statistics were required for analyzing comparisons between the number of marked T. lineatum 138 recaptured in trap groups, based on the hypotheses tested.

Comparisons between up- and down-valley movement of populations A and B were made for both experiments only when wind direction throughout the measurement period of a release day was relatively similar between the weather station and sites A and B. Similarities in wind direction tended to occur when wind blew from down- to up-valley. On these occasions, the criteria used to allow for beetle flight direction comparisons required that wind direction was + 45° the valley azimuth from the site measured. The up-valley azimuth for the weather station and site A was ca.

48°, and for site B ca. 68°. The number of population A beetles recaptured up-valley were compared to the number of population B beetles recaptured down-valley. Also comparisons were made for each population in the opposite directions, respectively, with the data analyzed using the

Mann-Whitney U test (Zar 1984). The same statistical test was used for comparing the number of marked beetles recaptured in the forested and open settings.

A comparison among three possible routes taken by T. lineatum up- and down-valley in a forest for experiment 2 was analyzed using the Kruskal-Wallis, analysis of variance by rank test (Zar 1984). For longer distance flight, comparisons between an open and a forested route were made using the Mann-Whitney U test. 139

7.3 Results and Discussion

Weather

Simultaneous weather measurements at three different sites in the Cedar Creek Valley, provided an opportunity for more precise documentation of weather patterns than was previously obtained (Chapter 6). In 1987, the first experimental season in the valley, weather data were collected simultaneously at the open site and one forested site at a time (of the two forested sites used). There were notable differences in wind direction, wind speed, temperature, and relative humidity between the two forest settings during the summer. However, it was very difficult to know what changes if any were occurring in other parts of the forest at the same time, and if so, how they would affect T. 1ineatum flight. The study reported in this chapter, not only allows for the observance of simultaneous weather pattern differences in the forest, but also their effect on beetle flight, since marked beetles were simultaneously released from both sites.

As in the previous two mark-recapture studies, meteorological measurements were made only on days that beetles were released. These days were warm and sunny with temperatures rising above 15°C under the forest canopy.

Therefore, meteorological comparisons between the sites applies only for these specific conditions. 140

Both release sites A and B, located 25 m below a road and 650 m apart from each other (Fig. 23) were similar in vegetation cover and density. From eight measurement days through the experimental period of June and July, 1988, temperatures between the two sites, never differed from each other by more than 1°C (Table 13). In contrast, the temperature in the open site was generally 2 - 4°C warmer than the forested sites. This is similar to the results in

Chapter 6. Relative humidity was also similar at both sites and was usually higher than the open site (Table 13). Again this is consistent with the previous year's study (Chapter

6) .

Wind speed was very similar in both forested sites. At site A, the average wind speed ranged from 1.3 - 2.3 km/h, with standard deviations ranging from 0.4 - 0.8 km/h (Table

13). At site B, the average wind speed ranged between 1.2 -

2.0 km/h, with standard deviations ranging from 0.5 - 0.8 km/h. The average wind speeds and respective variations are virtually the same as those measured in 1987 from the forested sites in the valley (Chapter 6), and from the previous year in the second-growth forest (Chapter 5). This indicates that if the weighting of individual wind directions by wind speed had a minimal effect on the computation of average wind direction from the first mark- recapture study (Chapter 5), it should not be important in this study. As a result, the weighting of wind direction by wind speed was not calculated. 141

Table 13. Weather measurements made between 1130 and 1700 h PDT, from the Cedar Creek Valley in the Coquitlam Lake Watershed in British Columbia during 1988.

Exp. Rel. Julian Site X S.D. Temperature Relative Date Wind speed (km/h) (°C) Humidity (%) (km/h)

1 1 165 Open 4.41 1.74 19.6 40.9 A 1.25 0.39 16.8 61.8 B 1.55 0.45 16.3 53.8

2 170 Open 6.10 1.86 28.0 23.9 A 1.91 0.62 24.4 42.8 B 1.46 0.48 23.4 43.0

3 177 Open 4.40 1.44 20.4 39.4 A 1.47 0.38 18.2 52.8 B 1.22 0.47 18.0 56.5

4 195 Open 6.77 1.45 23.5 39.9 A 2.03 0.46 20.0 49.6 B 1.57 0.70 20.9 49.4

2 1 204 Open 5.04 1.81 27.8 24.1 A 1.68 0.76 23.9 42.4 B 2.01 0.55 24.3 39.2

2 207 Open 6.97 1.36 22.7 67.2 A 2.15 0.43 21.1 65.0 B 1.92 0.80 20.6 64.6

3 211 Open 5.80 1.92 28.8 35.2 A 1.93 0.77 25.7 46.8 B 1.63 0.60 26.0 47.0

4 215 Open 6.90 1.51 23.4 51.6 A 2.32 0.76 21.4 . 58.3 B 1.88 0.78 20.8 57.6 142

The wind direction data are presented individually as circular histograms (Figs. 24 - 27). The black bars radiating out of the circle circumference represent the frequency of measured values from that particular direction.

Within each circle, the dotted line represents the 0W and vector length for the first 2 h or first half of the measuring period for that day. The dashed line is the same for the latter 2 h or the latter half of the measuring period for that day. The solid line represents the overall

jzfw and vector length for the measuring period. If the lines are present, then the directions are considered significant based on the Rayleigh test for randomness (Batschelet 1981).

For all measuring periods, at each site, j0w was significant.

As in the first valley study, average wind direction at the open site was significantly directed up-valley on all test days (Figs. 24 - 27). Up-valley winds were also very consistent at release site A. The only difference was a large amount of variability during the first release of experiment 2 (Fig. 26A). For all but the first release of both experiments, up-valley wind predominated also at forested site B. In experiment 1, release 1, wind blew from

up- to down-valley with a J0W of 27°, during the first half of the measuring day (Fig. 24A). In experiment 2, release

1, the overall 0W for the day was 10° (Fig. 24B). In the first valley study in 1987, all measurements from site B

(same in both studies) showed 0W blew up-valley. This indicates the up-valley winds that typically occur diurnally 143

A. Experiment 1/Release 1

12 345 ( 7 I 9 10 111213 FUNNEL TRAP GROUPS

Fig. 24. Wind direction and T. lineatum flight recapture for experiment 1, releases 1 (A) and 2 (B) , represented by circular and xy axis histograms, respectively, placed relative to their location along the length of the valley. The x-axis represents the valley azimuth. Lines in the circles are the mean direction for the first 2 h (dotted), final 2 h (dashed), and all day (solid), following beetle release, with line length signifying the r-value for the vector. 144

A. Experiment 1/Release 3

e c s >

o c < U W cc in ^ lb 11 12 13 - MALE t FEMALE w B. Experiment 1/Release 4 ea 20 - Cs. O 16 - 3? •o cc 12 - w CQ 8 - Site A SlttB 4 -

0 I 1 I 1 I 1 ' I ' I 20 - 16 - 12 B s- 8 D CD 4

1 11 11 0 1 1" i • i • i Tt TTT TTT 9 10 11 12 13 FUNNEL TRAP GROUPS

Fig. 25. Wind direction and T. lineatum flight recapture for experiment 1, releases 3 (A) and 4 (B), represented by circular and xy axis histograms, respectively, placed relative to their location along the length of the valley. The x-axis represents the valley azimuth. Lines in the circles are the mean direction for the first 2 h (dotted), final 2 h (dashed), and all day (solid), following beetle release, with line length signifying the r-value for the vector. 145

A. Experiment 2/ReIease 1

40 -(

30

20 Weather sltt A | SlteB Station • 10

Q 0 „ n PI _ H R

40

30

20 o 10 D BB CO 0 W 3 4 5 6 7 8 9 10 11 12 - 1 2 MALE FEMALE B. Experiment 2/Release 2

fa 40 - O 30 W 20 -I SiuB D 10 Z 0 I 1 'I'1 I

40 -|

30 ? •o e 20 - o B 10 - BB

0 1 1 1 i i • I ' I nrT ,,Br, Br 1 2 3 4 f 5 6 7 8 9 10 11 12 FUNNEL TRAP GROUPS

Fig. 26. Wind direction and T. lineatum flight recapture for experiment 2, releases 1 (A) and 2 (B), represented by circular and xy axis histograms, respectively, placed relative to their location along the length of the valley. The x-axis represents the valley azimuth. Lines in the circles are the mean direction for the first 2 h (dotted), final 2 h (dashed), and all day (solid), following beetle release, with line length signifying the r-value for the vector. 146

A. Experiment 2/Release 3

40 -

30

20 Weather Station 10

40

30

20

10

i i'T T • V • T • T fU. 1 2 3 4 5 6 7 8 9 10 11 12 MALE B. Experiment 2/Release 4 FEMALE

40 - 30 I 20 - Weather Site A SlteB SUUOD 10 - I 40 30 20

10

0 •II ' I ' "l" • T ' "I ' I ' i I ' i ' 1 2 3 4 5 6 7 8 9 10 11 12 FUNNEL TRAP GROUPS

27. Wind direction and T. lineatum flight recapture for experiment 2, releases 3 (A) and 4 (B), represented by circular and xy axis histograms, respectively, placed relative to their location along the length of the valley. The x-axis represents the valley azimuth. Lines in the circles are the mean direction for the first 2 h (dotted), final 2 h (dashed), and all day (solid), following beetle release, with line length signifying the r-value for the vector. 147 throughout the summer, on warm, cloudless days, in mountainous terrain (Barry 1981), is not absolute. However, based on two years of measurements, we found this phenomenon to be consistent.

Experiment 1

The proportion of marked beetles recaptured through the four releases averaged ca. 1% (Table 14). The % recapture among releases were similar, ranging from 0.8 - 1.3%. Over

11% of the recaptured beetles were caught during the second collection date of the respective releases, which is a much lower value than was recorded in 1987. The number of unmarked beetles captured was initially very high, with a decreasing trend as the summer progressed, resulting in a total of 148 000 unmarked beetles captured.

Beetle recapture pattern is first related to wind direction. During the first release, while wind blew up- valley from the weather station and site A, it blew down- valley from site B, especially during the first 2 h following beetle release (Fig. 24). The beetle recapture data supports the hypothesis of downwind movement, especially for population B, where over 80% were recaptured downwind (down-valley), in trap groups 1-7 (Fig. 24A).

Population A beetles were recaptured up- and down-valley,

which may have resulted from the opposing J2fw's at sites A and B. 148

Table 14. Flight and recapture success of Trypodendron lineatum in the mark-recapture study conducted at Coquitlam Lake Watershed in British Columbia during 1988.

Captured Beetles Marked Unmarked No. No. % of Flown Exp. Release Population Released Flying % Flown Total No. Beetles No.

1 1 A 4 000 3 166 79.2 16 0.5 B 4 000 3 250 81.3 46 1.4 Total 8 000 6 416 80.2 62 1.0 66 178

2 A 4 000 3 486 87.2 24 0.7 B 4 000 3 463 86.6 29 0.8 Total 8 000 6 949 86.9 53, 0.8 43 537

3 A 4 000 3 544 88.6 22 0.6 B 4 000 3 428 85.7 71 2.1 Total 8 000 6 972 87.2 93 1.3 27 226

4 A 4 000 2 898 72.5 19 0.7 B 4 000 2 608 65.2 35 1.3 Total 8 000 5 506 68.8 54 1.0 11 025

Population A 16 000 13 094 81.8 81 0.6 Totals B 16 000 12 749 79.7 181 1.4 Total 32 000 25 843 80.1 262 1.0 147 966

2 1 A 4 000 2 790 69.8 84 3.0 B 4 000 3 188 79.7 64 2.0 Total 8 000 5 978 74.7 148 2.5 24 519

2 A 4 000 3 370 84.3 99 2.9 B 4 000 2 790 69.8 28 1.0 Total 8 000 6 160 77.0 127 2.1 5 209

3 A 4 000 2 790 69.8 101 3.6 B 4 000 2 442 61.1 57 2.3 Total 8 000 5 232 65.4 158 3.0 3 306

4 A 4 000 2 906 72.7 121 4.2 B 4 000 2 442 61.1 68 2.8 Total 8 000 5 348 66.9 189 3.5 2 276

Population A 16 000 11 856 74.1 405 3.4 Totals B 16 000 10 862 67.9 217 2.0 Total 32 000 22 718 71.0 622 2.7 35 310 149

For releases 2-4, ^w was predominantly down- to up- valley at all three measurement sites. In most cases, beetle flight direction for both populations A and B were downwind (Figs. 24B and 25A and B). Population B were recaptured up-valley (downwind) for all three releases, while population A beetles were recaptured similarly in two of the three releases. The notable exception is for release

2, where population A beetles were recaptured down-valley, apparently upwind. The reason behind this deviation in pattern cannot be readily explained. Overall analysis of up- and down-valley beetle recapture patterns show that when wind was significantly directed up-valley at all three measurement sites, a significantly higher number of population A and B beetles were recaptured up-valley

(downwind) from the release sites (Table 15).

Data from the 1987 valley study suggested that beetles either avoided flying or could not successfully orient themselves to semiochemical-baited traps in the center of an open setting, where wind patterns tend to be quite unsettled

(Fig. 21). This notion is extended to compare the number of beetles recaptured in the forest setting, where wind speeds are low, and the open setting, where wind speeds are considerably higher (Tables 8 and 13). Within the plot design used for this experiment, all four trap groups in the forest (1, 2, 11, and 12) were further away from the beetle release sites than all the trap groups in the open (3 - 11)

(Fig. 23A). Yet the mean number of beetles Table 15. Comparison between up and down-valley movement of adult Trypodendron Iineatum during the mark-recapture study conducted in the Cedar Creek Valley within the Coquitlam Lake Watershed in British Columbia during 1988.

X Number of Harked Beetles Caught/Release ^

Exp. Trap Population Trap Population Locations A S.D. Locations B S.D.

1 Up-valley 10.3 5.1 Down-valley 2.3 2.5 *3 (7,8,9,10,11,12,13)2 (1,2,3,4,5,7)

1 Down-valley 7.7 9.1 Up-valley 37.0 19.3 (1,2,3,4,5) (9,10,11.12,13)

2 Up-valley 87.0 11.5 Down-valley 13.3 7.4 (5,6,7,8.9,10,11,12) (1,2,3,4,5,6,7)

2 Down-valley 0.7 1.2 Up-valley 20.7 9.9 * (1.2,3,4) (9,10,11,12)

Data are from releases 2, 3, and 4 for both experiments, where up-valley wind prevailed at all sites during the release days.

Numbers in parentheses represent trap groups from Figure 23.

Significant differences within a row are indicated by a star (Mann-Whitney U test; p < 0.10); the most sensitive probability this test can provide given a sample size of 3. 151 recaptured/release in the forested trap groups was significantly greater than those recaptured in the open trap

groups (U4 4 = 16; p < 0.05), with a recapture ratio of 3:1

(forest:open). It is apparent that wind speed is probably one of the primary reasons for this, as indicated by the data in Chapters 2 and 6.

One other consideration is the ability of T. lineatum to disperse from one side of the valley to the other. Trap groups 2, 6, 7, 8, and 13 were set across the creek from both release sites (Fig. 23A). Throughout the four releases, 38% of the 262 total recaptured beetles were collected from these traps which represented 38% of all the traps deployed in the valley. The data show that beetles are capable of flying across valleys, with no apparent barrier presented by the creek.

Experiment 2

A new plot design (Fig. 23B), with the objectives of not allowing the presence of any trap group within 200 m of the release sites, while providing insight into the flight routes taken by T. lineatum. resulted in an increase up to

2.7% of the proportion of marked beetles recaptured (Table

14). This can be attributed to trap groups 6, 7, and 8, placed in between both beetle release sites, which caught

431/583 (74%) of the marked beetles recaptured during this experiment (Figs. 26 and 27). Over 26% of the recaptured beetles were collected during the second collection date for 152 the respective releases. It is possible that the higher proportion of beetles recaptured during this period, when compared to experiment 1 data, may have been a result of the increasingly negative effects storage has on the beetles through time, resulting in beetles being slower dispersers.

Nijholt (1967) found that the fat content of T. lineatum decreased as storage time increased. The proportion of released beetles flying also decreased with time (Table 14), as further evidence to the weaker condition of the insects.

In support of this hypothesis, the proportion recaptured during the second collections tended to increase with time;

13, 15, 31, and 22% for release 1 through 4, respectively.

The number of unmarked beetles recaptured during the experiment dropped to 35 000 beetles (Table 14), as was expected from other seasonal catch records (Chapman and

Nijholt 1980; Rudinsky and Daterman 1964; Shore et al.

1986).

Beetle recapture beetle patterns may again be related

to wind. For all but the first release, #w blew predominantly down- to up-valley at all three measuring sites (Figs. 26 and 27). During the first release, the same

#w was observed at the weather station and site A (Fig.

2 6A). But again, as in experiment 1, release 1, J0W at site

B came from the opposite direction, up-valley. Population A beetles were recaptured predominantly up-valley (downwind) in trap groups 6, 7, and 8. In contrast, an equal number of population B beetles were caught up- and down-valley. This 153

pattern may have occurred due to the opposing #w''s measured from sites A and B. In releases 2-4, almost all population A beetles were recaptured up-valley (downwind) and a higher proportion of recaptured population B beetles were collected up-valley than down-valley (Figs 26B and 27A and B). Despite strong up-valley winds, some population B beetles were recaptured in traps 6, 7, and 8. One explanation for this may be that given a choice, some beetles may have preferred to fly upwind under gentler wind conditions in the forest, rather than fly downwind under higher wind speed conditions in the open. Analysis of up- and down-valley beetle recapture patterns show that when wind was significantly directed up-valley, at all three measurement sites, a significantly higher number of population A and B beetles were recaptured up-valley

(downwind) from their release sites (Table 15).

In an effort to see if beetles used a road as a flyway, a comparison among recaptured beetle numbers in trap groups

6, 7, and 8 for up-valley flying population A and down- valley flying population B beetles was made (Fig. 23B). No statistical differences between trap groups were found for both recaptured populations A and B (Table 16). A slightly higher number of beetles from both populations were recaptured in trap group 7 (on the road), yet this may have been due to the release sites being closer to this trap group than the other two. The number of up-valley A beetles recaptured in trap groups 6, 7, and 8 was considerably Table 16. Movement of T. lineatum through a forest, flying up-valley (population A) and down-valley (population B) during four separate releases in Experiment 2, within the Coquitlam Lake Watershed in British Columbia during 1988.1/2

X Number of X Number of Trap Groups3 Population A S.D. Population B S. D. Beetles Caught/ Beetles Caught/ Release Release

30 m Above The Road (8) 29.8 6.2 4.0 2. 7

On The Road (7) 37.0 12.7 9.5 6. 4

100 m Below The Road (6) 22 .8 10.2 4.3 1. 9

1 Beetle recapture data are from one release period to the next.

No significant differences were found between trap groups for both populations A and B (Kruskal-Wallis test; p > 0.05).

Number in parentheses represents trap group number from Figure 23B. 155 higher than the number of down-valley B beetles, adding support to the downwind flight hypothesis (Table 16).

In flight route comparisons between traps groups in the forested and open sites, the results through the four releases show that T. lineatum were recaptured in higher numbers in forested trap groups 4 and 9 than in the non- forested groups 3 and 10 (Table 17). Unfortunately, this does not necessarily indicate beetle flight route, as it may just indicate, as in experiment 1, that T. lineatum are more easily recaptured in the forest than in the open setting.

In conclusion, it was found that without the presence of close-range attractants, T. lineatum generally fly downwind in their search for host material and mates. In the Cedar Creek Valley, this was usually up-valley, as a result of the characteristic diurnal up-valley wind patterns in valley systems (Barry 1981). It is important to note however that during these ideal weather conditions, the wind sometimes deviated from an up-valley direction. This was noticed twice at site B, and may be a result of wind eddies forming at the edge of the forest setting. As explained by

Kimmins (1987), a laminar air flow over a cool forest will curl into the forest at the upwind edge of an open site.

This phenomenon may have been more pronounced on the two

occasions that site B 0W differed from the J0W at the other two measurement sites.

An important outcome of this study is that T. lineatum clearly either flies better in the forest, orients to hosts Table 17. Comparison between an open and forested route for long distance flying Trypodendron 1ineatum, during 4 releases in Experiment 2, within the Coquitlam Lake Watershed in British Columbia during 1988.

X Number of Marked Beetles Caught/Release x

Setting Down-valley Up-valley Traps S.D. Traps S.D.

Open (3)2 0.0 b3 0.0 (10) 4.0 b 3.1

Forested (4) 3.5 a 2.4 (9) 11.3 a 3.6

Beetles from populations A and B were combined, using recapture data from one release period to the next.

Number in parentheses represents trap group number from Figure 23B.

Different letters in a column indicates significant differences between settings (Mann-Whitney U test; p < 0.05). 157 better in the forest, or both, rather than in an open setting. In experiment 1, more beetles were recaptured in trap groups in forests than in open settings, of which the latter were actually closer to the release sites. In experiment 2, trap groups in forests consistently recaptured more beetles than in open sites. During the

first valley study in 1987, more beetles were recaptured at the edge of an opening than in the center, when the edge was

further away from the release sites (Chapter 6). Thus, we have extended this continuum into the forest. The relatively light wind speed characteristic under a forest canopy is most likely responsible for the increased number of beetles recaptured. Others have suggested that vertical silhouettes within the forest serve to stimulate beetle

flight (S. Bombosch personal communication)!6-. However, for a beetle that is initially photopositive upon spring emergence, and has evolved to attack both windthrown

(horizontal) and bark beetle killed timber (vertical), typically oriented horizontally, this hypothesis is less plausible. Additionally, in a wind tunnel study described in Appendix 1, no preference was shown for semiochemical- baited substrates oriented vertically or horizontally.

The recapture data also suggest that flight distances of 1.5 km for spring dispersing T. lineatum are possible.

In the first valley study, T. lineatum exhibited the ability to fly 1.9 km, however, the frequency of recapture was never

16 Dr. S. Bombosch, Institut fur Forstzoologie, Universitat Gottingen, Busgenweg 3, D-3400 Gottingen, W. Germany 158 higher than one, for a particular release. The frequencies

of recapture from this study at distances of similar magnitude were much higher. The reason for the differences between the two studies is that during the most recent study, the long distance traps were placed in the forest.

Also the use of trap groups, instead of individual traps in a line, may have been a more effective strategy in attracting and subsequently capturing beetles, by increasing the size and concentration of the semiochemical plumes produced.

The data from this study provide a basis for characterizing the movement of T. lineatum during the mass

flight period in the spring. In most woodlands operations along the west coast of Canada, the previous year's harvest

inventory, not yet removed from the forest before this period, is susceptible to attack by T. lineatum. The data from this and the previous chapter suggest that placement of inventory can play an important role in its ultimate attack.

Logs^probably safest when stored as far away from forest margins as possible, in the center of an open setting.

Eventually the beetles will find this material, but conditions in an open setting in mountainous terrain are less likely to be conducive for beetle attack.

Alternatively, logs near forest margins should be removed more swiftly than the logs in the center of the open settings. Knowledge of wind patterns, especially prevailing winds if they occur, on warm, sunny days in an area 159 containing inventory, can be used as a guide to determine which portion of the inventory would be most susceptible to attack based on its location relative to the surrounding forest.

The implementation of mass-trapping techniques used in log sorts and storage areas has been shown to successfully reduce populations of European T. lineatum within a forest stand during a three year period, despite a yearly replacement of susceptible logs on the site (Konig 1988).

McLean and Salom (unpublished data) attempted to reduce the attack of susceptible logs on a right-of-way in a commercial forest on Vancouver Island, using mass-trapping techniques, however they were unsuccessful. Based on the data from

Chapter 6 and this chapter, it is likely that the lack of success in using this strategy resulted from the placement of traps near susceptible logs, rather than at the sites where beetle populations were likely to overwinter. I feel mass-trapping does have potential as a pest management tool in protecting logs before they are transported from the forest. In areas where harvesting is continuous, inventory susceptible to spring attack and removed from the site after

June, should be sampled for T. lineatum attack. If attack is abundant, a large brood population most likely has flown to the nearby forests to overwinter (Dyer and Kinghorn

1961). It is in this forest that semiochemical-baited trap groups should be put in place the following spring. The traps should not only be placed along the forest margin, but 160 as far as 100 m into the forest. In the spring, many of the emerging beetles should be captured in the traps, preventing them from attacking susceptible host material at new harvest sites. While not all beetles will be captured this way, it is likely that this trapping can reduce the populations in the forest. Combined with short time intervals between harvest and transport of inventory, along with the continued mass-trapping at the log sorts, I believe that populations of T. lineatum can be reduced significantly at commercial woodlands divisions, resulting in reductions in degrade loss due to this species. Research should be continued in this area to see if the proposed approach is plausible and also to test modifications that can make this strategy work most successfully. 161

CHAPTER 8: CONCLUSIONS

Dispersal

The realization that dispersal is a critical aspect of

the life history and population dynamics of many insect

species resulted from important reviews on the subject

(Southwood 1962; Johnson 1969) and the implications of these views on pest management have been the subject of several

symposia since then (Berryman and Safranyik 1980;

Danthanarayana 1986; Delucchi and Baltensweiler 1979; Rabb and Kennedy 1979). The definitions used to describe and

contrast dispersal frequently differ. Dingle (1985)

considers dispersal to be random or haphazard movements which occur purely as a result of chance circumstances, whereas migration is defined as specialized behavior evolved

for the displacement of the individual in space. Kennedy

(1975) stated that dispersal can involve specialized behavior, but movements occur within the habitat. This contrasts with migration which represents movement between different habitat localities. Wellington (1980) defined dispersal as an adaptive response to habitat, often

involving a period of obligatory travel independent of previous or current population densities.

The definition provided by Wellington fits in well with the life history patterns of scolytid beetles which begin their dispersal stage when they emerge from the brood tree or log and end when they respond to host stimuli and/or 162

attractive pheromones (D. Wood 1982). The factors inducing

dispersal and its characteristics, such as behavioral

reversals in flight, host selection, perception of pheromones, and mass response to aggregation pheromones are

clearly species-specific (Borden 1984) and are very much

related to their habitat (Atkins 1966b). Species-specific variation of dispersal is substantial and can be illustrated by considering flight distance to host material. Wollerman

(1979) found that the maximum distance flown by Scolytus multistriatus (Marsham) was 150 m in a mark-recapture study.

In contrast, Botterwerg (1982) showed that Ips typoqraphus

L. are capable of flying 8 km. An even more dramatic example comes from Nilssen (1984) who found three scolytid

species travelled (likely wind blown) 171 km in northern

Finland.

Host selection, the next stage in host colonization strategies for scolytids is either a random process, in which beetles land indiscriminantly on hosts, or a directed process, in which they land preferentially on hosts vs. nonhosts (D. Wood 1982). Selection of hosts by pioneer beetles are facilitated by their perception of host volatiles (1° attractant), followed by production of aggregation pheromones (2° attractant), and their perception and mass response to them (Borden 1984). Atkins (1966b) suggested that 2° attraction for scolytids evolved initially

in conjunction with their utilization of temporary habitats, assuring maximum utilization of scattered host material. 163

The ability of aggressive bark beetles to attack living, healthy trees through mass-attacks probably arose later.

In terms of pest management, we must consider dispersal and host selection in the context of population strategies.

For T. lineatum this required a close examination of 1) the effective distance a population is capable of flying to cause economic devaluation of susceptible logs; 2) the

influence of wind, forest cover, and topography on overall distribution of dispersing beetles; 3) optimal environmental conditions required for T. lineatum response and orientation to olfactory stimuli; and 4) the importance and roles of the semiochemicals used by T. lineatum. and how such information relates to beetle trapping strategies. A summary that addresses these considerations based on an integration of previous research with the research from this report is presented below.

Spring Dispersal of Trypodendron lineatum

Overwintering beetles that emerge and fly in the spring when temperatures warm up above 15°C (Chapman and Kinghorn

1958; Rudinsky and Daterman 1964), will usually take-off toward the open sky (Graham 1959). This has been interpreted as a photopositive response. If no attractants are present, beetles will generally take-off in all directions under very low wind speed conditions (< 1 km/h), with increasing frequency of downwind take-off as wind speed increases (Fig. 4). If attractive susceptible host material 164'

is near the emergence sites (+ 25 m), some beetles will respond immediately and fly upwind to the source of attraction, while other non-responding beetles will continue on their way (Figs. 17 and 20). In a homogeneous forest

setting, beetles will fly in all directions, with the greatest dispersal distances recorded downwind from the emergence site (Fig. 18). Under higher wind speed conditions, where the forest is broken up and may be more

influenced by prevailing winds, beetle flight will tend to be downwind (Table 15), until perception and response to semiochemicals sets them on an upwind track to the attractive source. Along the windward side of a forest margin, beetles will fly predominantly downwind into the

forest (Figs. 26 and 27). However, from the leeward side of a prevailing wind, while many beetles will fly downwind into openings at high wind speeds, some beetles will also fly upwind through the forest (Figs. 26 and 27), where the wind speed is markedly lower (Table 13). In fact, beetle capture or log attack is most successful in the forest (Table 17), where wind speeds are low, less successful at the edge of an open setting where wind speeds are higher, and least successful in the center of the opening where wind speeds and turbulence are greatest (Fig. 21).

Trypodendron lineatum are adapted to find rare host material in natural forest situations. Their small size restricts them from flying against strong winds (Chapman

19 62), and therefore they must optimize their search for 165

host material by using wind to aid in their dispersal. This

was indicated by the increased proportion of downwind (Fig.

5), direct (Fig. 6) flight observed, as wind speed increased

in a wind tunnel, and general downwind movement observed in

the field (Table 15). Our studies indicate that the beetles

can easily fly 1.5 km (Figs. 24 - 27), and can likely fly

further than 2 km as long as host material occurs near or

within a forest setting. They can fly these distances in

one day, yet they can also locate a source of attractant

after a week (Fig. 22). Flight from one side of a valley to

another is little problem for these beetles (Figs. 24 and

25) .

Primary attractants, especially ethanol, are important

for females searching out, responding to, and accepting host material (Moeck 1970; Figs. 13 - 15; Table 4). Once the

female finds and accepts the host, she produces a powerful

aggregation pheromone (Borden and Slater 1969; MacConnell et

al. 1977), lineatin, that attracts both males and females to mass attack the host. Males, while less influenced by

ethanol than females are (Table 4), do respond to it in a positive manner with respect to flight direction (Fig. 13) and speed of response (Fig. 15). Yet males appear to rely mostly on lineatin to find and land on suitable host material (Fig. 14; Table 4), facilitating the mating process with females. The significance of quicker responses by

responding males and females when ethanol was present (Fig. 166

15) probably results from a more complete chemical message

from host and mates.

Striking differences in response to host attractants have been observed between North American and European populations of T. lineatum. In Europe, host attractants,

ethanol and <9£-pinene, when combined with lineatin, have a

synergistic effect on the number of beetles captured (Borden

et al. 1982; Paiva and Kiesel 1985; Vite and Bakke 1979),

regardless of what type of trap is used (Borden et al. 1981;

McLean et al. 1987). In North America, host attractants are

important to a lesser degree; there is a stronger trap by attractant interaction affecting T. lineatum capture.

Beetles respond to host attractants in drainpipe traps

(Shore and McLean 1983; Fig. 14; Table 4), yet do not respond to host attractants in a funnel trap (Salom and

McLean 1988; Fig. 8). Vite and Bakke (1979) explained that the drainpipe trap requires an extra behavioral response by the beetles to enter the small holes that simulate gallery openings. Host volatiles enhance this response. This behavior is not necessary for the capture of beetles in

funnel traps (Lindgren 1983).

Upon perception of the host, beetles fly anemotactically to it in a steady flight in low wind conditions. Once near the host, the beetles seem to test either the chemical gradient, plume structure, or the wind, before landing on the log, and either attack it or mate with another beetle. Under still wind conditions, the beetles 167 will land in higher frequencies, than if wind is present

(Figs. 2 and 7). They appear to dive into the host, somehow landing upright on their legs (personal observation). It appears that different mechanisms of host finding are used by T. lineatum under varying wind speed conditions. In still wind conditions, beetles are arrested and orient to semiochemicals. In the presence of low wind speeds, beetles attracted to the semiochemicals respond to the odor plume using chemo-anemotaxis (Chapter 2). The ability of the beetles to use these different mechanisms to search out and attack hosts under continuously changing conditions in natural forest settings, allow the beetles to optimize their search for a rare host.

New Considerations in Pest Management

The results from my experiments can aid in improving management of this pest at the log sorts and storage areas, as well as in the commercial forest settings. Currently, mass-trapping programs deploy semiochemical-baited funnel traps along the forest margin surrounding open areas containing susceptible logs. The traps in most cases are placed no further than 10 m into the margins (Phero Tech.

Technical Bulletin No. 84001). I suggest placing traps up to 50 - 100 m into the forest, if enough forest area is present. Since beetles are captured at higher frequencies in still air conditions of a forest, than in open areas, there is an opportunity to attract beetles away from the 168

logs set in the open areas, instead of just trying to set up

a barrier which will attract beetles in the direction of the

logs.

In commercial forests, where log inventories could not

be removed from the settings before mass flight of T.

lineatum in the spring, logs should be moved from the edges

of the open setting and piled or stacked near its center.

This could reduce the amount of logs attacked because of the

difficulties the beetles seem to have in orienting to

olfactory stimuli in open areas. Field studies need to be

conducted to see if mass-trapping can work. Konig (1988)

has shown that beetle numbers can be reduced in the forest

using this technique. I suggest placement of semiochemical- baited funnel traps placed in the forest, 0 - 100 m deep

into the forest, around a logged setting that was heavily

attacked by T. lineatum the previous season. Population movement out of the forest to a new site can be impeded by

capturing these beetles the following spring. Given a

choice, emerging beetles will fly in the forest rather than

an open setting, therefore the chance of capturing them before they disperse long distances to new sites is good.

Placement of cut blocks in a commercial forest can be modified to minimize beetle attack of high quality logs.

The general trend in logging old-growth forests along the

B.C. coast is to cut stands in the lower valley elevations

first, followed by harvests at higher elevations. These

latter cut blocks are very susceptible to T. lineatum 169

attack, because up-valley winds will aid in dispersing beetles from old cut blocks to new ones. If new cut blocks

can be placed upwind (providing the presence of a prevailing wind) to old cut blocks, the ability of the population from the old cut block successfully finding the new site would be, in my opinion, severely restricted. In B.C., as second- growth forests come into use or where road networks are

already established, placement of cut blocks upwind to previous settings may be become more plausible. In the

United States, where second growth forestry now dominates, this approach can presently be considered on a larger scale.

While improvements have and can be made toward managing

T. lineatum populations, more information on dispersal, especially during the fall season, can provide a means to develop a comprehensive pest management strategy extending

from the forest settings into the log sort and storage areas. Such a program could save the British Columbia

Forest Industry millions of dollars each year. 170

REFERENCES

Annila, E., A. Bakke, B. Bejer-Peterson, B. Lekander. 1972. Flight period and brood emergence in Trypodendron lineatum (Oliv.) (Col., Scolytidae) in the Nordic countries. Comm. Inst. For. Fenn. 76:7-28.

Angerilli, N. and J.A. McLean. 1984. Windtunnel and field observations of western spruce budworm responses to pheromone-baited traps. J. Ent. Soc. B.C. 81:10-16.

Atkins, M.D. 19 66a. Laboratory studies on the behaviour of the Douglas-fir beetle, Dendroctonus pseudotsugae Hopk. Can. Ent. 98:953-991.

Atkins, M.D. 1966b. Behavioral variation among scolytids in relation to their habitat. Can. Ent. 98: 285-288.

Atkins, M.D. 1969. Lipid loss with flight in the Douglas- fir beetle. Can. Ent. 101:164-165.

Baker, T.C. and L.P.S. Kuenen. 1982. Pheromone source location in flying moths: A supplementary non- anemotactic mechanism. Science. 216:424-427.

Baker, T.C. and C.E. Linn, Jr. 1984. Wind tunnels in pheromone research, pp. 75-110. In Hummel, H.E. and T.A. Miller (Eds.). Techniques in Pheromone Research. Springer-Verlag, New York. 464 pp.

Banks, H.T. P.M. Kareiva, and L. Zia. 1988. Analyzing field studies of insect dispersal using two-dimensional transport equations. Environ. Entomol. 17:815-82 0.

Barry, R.G. 1981. Mountain Weather and Climate. Metheun, London. 313 pp.

Batschelet, E. 1981. Circular Statistics in Biology. Academic Press. 371 pp.

Bauer, J. and J.P. Vite. 1975. Host selection by Trypodendron lineatum. Naturwiss. 62: 539.

Bell, W.J. and R.T. Carde (Eds.). 1984. Chemical Ecology of Insects. Chapman & Hall, London. 524 pp.

Bennett, R.B. and J.H. Borden. 1971. Flight arrestment of tethered Dendroctonus pseudotsugae and Trypodendron lineatum (Coleoptera: Scolytidae) in response to olfactory stimuli. Ann. Entomol. Soc. Am. 64:1273- 1286. 171

Berryman, A.A. and L. Safranyik (Eds.)* 1980. Proc. IUFRO Conf., Dispersal of Forest Insects: Evaluation, Theory and Management Implications. Idaho. 278 pp.

Birch, M.C. and P.R. White. 1988. Responses of flying male Anobium punctatum (Coleoptera: Anobiidae) to female sex pheromone in a wind tunnel. J. Ins. Behav. 1:111-115.

Borden, J.H. 1977. Behavioral responses of Coleoptera to pheromones, allomones, and kairomones. pp. 169-198. In Shorey, H.H. and J.J. McKelevey, Jr. (Eds.). Chemical Control of Insect Behavior: Theory and Application. John Wiley and Sons, Inc.

Borden, J.H. 1984. Semiochemical-mediated aggregation and dispersion in the Coleoptera. pp. 123-149. In Lewis, T. (Ed.) Insect Communication. Academic Press, London.

Borden, J.H. 1985. Aggregation pheromones. pp. 257-285. In Behavior. Kerkut, G.A. (Ed.). Vol 9. In Comprehensive Insect Physiology, Biochemistry, and Pharmacology. Kerkut, G.A. and L.I. Gilbert (Eds.). Pergamon Press, Oxford.

Borden, J.H. 1988a. The striped ambrosia beetle, Trypodendron lineatum (Olivier) pp. 579-596. In Berryman, A.A. (Ed.). Dynamics of Forest Insect Populations. Plenum Publishing Co. 603 pp.

Borden, J.H. 1988b. Use of semiochemicals to manage coniferous tree pests in western Canada. In Ridgeway, R., R.M. Silverstein, and M. Inscoe (Eds.). Practical Applications of Insect Pheromones and Other Attractants. (in press) .

Borden, J.H. and J.A. McLean. 1981. Pheromone-based suppression of ambrosia beetles in industrial timber processing areas, pp. 133-154. In Mitchell, E.R. (ED.). Management of Insect Pests With Semiochemicals: Concepts and Practice. Plenum Press, New York. 514 pp.

Borden, J.H. and C.E. Slater. 1969. Sex pheromone of Trypodendron lineatum: production in the female hindgut- Malpighian tubule region. Ann. Entomol. Soc. Am. 62: 454-455.

Borden, J.H., L. Chong, J.A. McLean, K.N. Slessor, and K. Mori. 1976. Gnathotrichus sulcatus: synergistic enantiomers of the aggregation pheromone sulcatol. Science 192:894-896. 172

Borden, J.H., L. Chong, K.N. Slessoe, A.C. Oehlschlager, H.D. Pierce, Jr. and B.S. Lindgren. 1981. Allelochemic activity of aggregation pheromones between three sympatric species of ambrosia beetles (Coleoptera: Scolytidae). Can. Ent. 113:557-563.

Borden, J.H., J.R. Handley, B.D. Johnston, J.G. MacConnel, R.M. Silverstein, K.N. Slessor, A.A. Swigar, and D.T.W. Wong. 1979. Synthesis and field testing of 4,6,6- lineatin, the aggregation pheromone of Trypodendron lineatum (Coleoptera: Scolytidae). J. Chern. Ecol. 5:681-689.

Borden, J.H., J.R. Handley, J.A. McLean, R.M. Silverstein, L. Chong, K.N. Slessor, B.D. Johnston, H.R. Schuler. 1980. Enantiomer-based specificity in pheromone communication by two sympatric Gnathotrichus species ( Coleoptera: Scolytidae). J. Chern. Ecol. 6:445-456.

Borden, J.H., D.W.A. Hunt, D.R. Miller, and K.N. Slessor. 1986. Orientation in forest Coleoptera: an uncertain outcome of responses by individual beetles to variable stimuli. pp. 97-110. In Payne, T.L., M.C. Birch, and C.E.J. Kennedy (Eds.). Mechanisms in Insect Olfaction. Clarendon Press, Oxford. 3 64 pp.

Borden, J.H., C.J. King, S. Lindgren, L. Chong, D.R. Gray, A.C. Oehlschlager, K.N. Slessor, and H.D. Pierce, Jr. 1982. Variation in response of Trypodendron lineatum from two continents to semiochemicals and trap form. Environ. Entomol. 11:403-408.

Botterweg, P.F. 1982. Dispersal and flight behaviour of the spruce bark beetle Ips typographus in relation to sex, size, and fat content. Z. ang. Ent. 94:466-489.

Byers, J.A. 1988. Upwind flight orientation to pheromone in western pine beetle tested with rotating windvane traps. J. Chern. Ecol. 14:189-198.

Carde, R.T. 1984. Chemo-orientation in flying insects, pp. 111-124. In Bell, W.J. and R.T. Carde (Eds.). Chemical Ecology of Insects. Sinauer Assoc., Inc. Massachusetts.

Chapman, J.A. 1956. Flight muscle changes during adult life in a scolytid beetle. Nature 177:1183.

Chapman, J.A. 1962. Field studies on attack flight and log selection by the ambrosia beetle Trypodendron lineatum (Oliv.) (Coleoptera: Scolytidae). Can. Ent. 94:74-92.

Chapman, J.A. and J.M. Kinghorn. 1958. Studies of flight and attack activity of the ambrosia beetle, 173

Trypodendron lineatum (Oliv.), and other scolytids. Can. Ent. 90:362-372.

Chapman, J.A. and W.W. Nijholt. 1980. Time and attack flight of Ambrosia beetle Trypodendron lineatum (Coleoptera: Scolytidae) in relation to weather in coastal British Columbia. Can. For. Serv. Inf. Rept. BC-R-5. 2 3 pp.

Choudhury, J.H. and J.S. Kennedy. 1980. Light versus pheromone-bearing wind in control of flight direction by beetles, Scolytus multistriatus. Physiol. Entomol. 5:207-214.

Danthanarayana, W. (Ed.). 1986. Insect Flight: Dispersal and Migration. Springer-Verlag, Berlin. 289 pp.

Daust, D.K. 1985. The influence of site characteristics on logging techniques and ambrosia beetle damage in the Lake Cowichan Area. B.S.F. Thesis. Univ. of British Columbia. 69 pp.

Delucchi, V. and W. Baltensweiler (Eds.). 1979. Proc. IUFRO Conf., Dispersal of Forest Insects: Evaluation, Theory and Management Implications. Switzerland. 342 pp.

Dickens, J.C. 1981. Behavioural and elctrophysiological responses of the bark beetle, Ips typographus to potential pheromone components. Physiol. Entomol. 6:251-261.

Dingle, H. 1985. Migration, pp. 375-414. In Behaviour. Kerkut, G.A. (Ed.). Vol. 9. In Comprehensive Insect Physiology, Biochemistry, and Pharmacology. Kerkut, G.A. and L.I. Gilbert (Eds.). Pergamon Press, Oxford.

Dyer, E.D.A. 1961. Flight capability of ambrosia beetle (Trypodendron). Can. Dep. Agric. For. Biol. Div. Bi- Mon. Prog. Rep. 17(1):4.

Dyer, E.D.A. and J.A. Chapman. 1965. Flight and attack of the ambrosia beetle, Trypodendron lineatum (Oliv.) in relation to felling dates of logs. Can. Ent. 97:42- 57.

Dyer, E.D.A. and J.M. Kinghorn. 1961. Factors influencing the distribution of overwintering ambrosia beetles, Trypodendron lineatum (Oliv.). Can. Ent. 93:746-759.

Edson, L.J. 1978. Host colonization and the arrival sequence of the mountain pine beetle and its insectan associates. Ph.D. Dissertation, University of California, Berkeley. 196 pp. 174

Edwards, D.K. 1960. Effects of experimentally altered unipolar air-ion density upon the amount of activity of the blowfly, Calliphora vicina R.D. Can. J. Zool. 38:1079-1091.

Farkas, S.R. and H.H. Shorey. 1972. Chemical trail- following by flying insects: A mechanism for orientation to a distant odor source. Science. 178:67-68.

Farkas, S.R. and H.H. Shorey. 1974. Mechanisms of orientation to a distant pheromone source. In Birch, M.C. (Ed.). Pheromones. North-Holland Publ. Co. 495 pp.

Fares, Y., P.J.H. Sharpe, C.E. Magnuson. 1980. Pheromone dispersion in forests. J. Theor. Biol. 84:335-359.

Fisher, R.C., G.H. Thompson, and W.E. Webb. 1953. Ambrosia beetles in forest and sawmill: their biology, economic importance and control. Part 1. For. Abstr. 14:381- 389.

Fockler, C.E. and J.H. Borden. 1972. Sexual behavior and seasonal mating activity of Trypodendron lineatum (Coleoptera: Scolytidae). Can. Ent. 104:1841-1853.

Fox, D.J. and K.E. Guire. 1976. Documentation for Midas. Stat. Res. Lab., University of Michigan. 2 03 pp.

Francia, F.C. and K. Graham. 1967. Aspects of orientation behavior in the ambrosia beetle Trypodendron lineatum (Olivier). Can. J. Zool. 45:985-1002.

Francke-Grossman, H. 1963. Some new aspects in forest entomology. Ann. Rev. Entomol. 8:415-438.

Funk, A. 1965. The symbiotic fung of certain ambrosia beetles in British Columbia. Can J. Bot. 43:929-932.

Furniss, M.M. and R.L. Furniss. 1972. Scolytids (Coleoptera) on snow-fields above timberline in Oregon and Washington. Can. Ent. 104:1471-1478.

Gara, R.I. 1963. Studies on the flight behavior of Ips confusus (Lec.) (Coleoptera: Scolytidae) in response to attractive material. Contrib. Boyce Thompson Inst. 21:51-66.

Graham, K. 1959. Release by flight exercise of the chemotropic response from photopositive domination in a scolytid beetle. Nature. 184:284-85. 175

Gray, D.R. and J.H. Borden. 1985. Ambrosia beetle attack on logs before and after processing through a dryland sorting area. For. Chron. 61:299-302.

Haufe, W.O. 1954. The effects of atmospheric pressure on the flight responses of Aedes aegypti (L.) Bull. Entomol. Res. 45:507-526.

Helland, I.S., J.M. Hoff, and O. Anderbrant. 1984. Attraction of bark beetles (Coleoptera: Scolytidae) to a pheromone trap: experiment and mathematical models. J. Chem. Ecol. 10:723-752.

Henson, W.R. 1962. Laboratory studies on the adult behavior of Conopthorus coniperda (Col.: Scolyt.). III. Flight. Ann. Entomol. Soc. Am. 55:524-530.

Johnson, C.G. 1969. Migration and Dispersal of Insects by Flight. Metheun, London. 763 pp.

Kennedy, J.S. 1975. Insect dispersal. pp. 103-119. In Pimental, D. (Ed.). Insects, Science and Society. Academic Press, New York.

Kennedy, J.S. 1978. The concepts of olfactory arrestment and attraction. Physiol. Entomol. 3:91-98.

Kennedy, J.S. 1986. Some current issues in orientation to odour sources. pp. 11-25. In Payne, T.L., M.C. Birch, and C.E.J. Kennedy (Eds.). Mechanisms in Insect Olfaction. Clarendon Press, Oxford.

Kimmins, J.P. 1987. Forest Ecology. MacMillan Publishing Co. New York. 531 pp.

Kinghorn, J.M. and J.A. Chapman. 1959. The overwintering of the ambrosia beetle Trypodendron lineatum (Oliv.) For. Sci. 5:81-92.

Kleinbaum, D.G. and L.L. Kupper. 1978. Applied Regression Analysis and Other Multivariable Methods. Duxbury Press, California. 486 pp.

Konig, E. 1988. Mass-trapping of Trypodendron lineatum 01. (Col., Scolytidae): effect of infestation of cut timber temporarily left in the forest. Z. ang. Ent. 106:262- 265.

Kydonieus, A.F. and M. Beroza. 1981. The hereon dispenser formulation and recent test results. pp. 445-454. In Mitchell, E.R. (Ed.). Management of Insect Pests With Semiochemicals: Concepts and Practice. Plenum Press, New York. 514 pp. 176

Lindgren, B.S. 1983. A multiple funnel trap for scolytid beetles (Coleoptera). Can. Ent. 115:299-302.

Lindgren, B.S. and J.H. Borden. 1983. Survey and mass trapping of ambrosia beetles (Coleoptera: Scolytidae) in timber processing areas on Vancouver Island. Can. J. For. Res. 13:481-493.

Lindgren, B.S., J.H. Borden, L.Chong, L.M. Friskie, D.B. Orr. 1983. Factors influencing the efficiency of pheromone-baited traps for three species of ambrosia beetles (Coleoptera: Scolytidae). Can. Ent. 115:303- 313.

Linton, D.A., L. Safranyik, L.H. McMullen, and R. Betts. 1987. Field techniques for rearing and marking mountain pine beetle for use in dispersal studies. J. Ent. Soc. B.C. 84:53-57.

MacConnell, J.G., J.H. Borden, R.M. Silverstein, and E. Stokkink. 1977. Isolation and tentative identification of lineatin, a pheromone from the frass of Trypodendron lineatum (Coleoptera: Scolytidae). J. Chern. Ecol. 5:549-561.

McLean, J.A. 1985. Ambrosia beetles: a multimillion dollar degrade problem of sawlogs in coastal British Columbia. For. Chron. 61:295-298.

McLean, J.H. and J.H. Borden. 1977. Attack by Gnathothrichus sulcatus (Coleoptera: Scolytidae) on stumps and felled trees baited with sulcatol and ethanol. Can. Ent. 109:675-686.

McLean, J.A., A. Bakke, H. Niemeyer. 1987. An evaluation of three traps and two lures for the ambrosia beetle Trypodendron lineatum (Oliv.) (Coleoptera: Scolytidae) in Canada, Norway, and West Germany. Can. Ent. 119:273-280.

McMullen, L.H. and M.D. Atkins. 1962. On the flight and host selection of the Douglas-fir beetle, Dendroctonus pseudotsucfae Hopk. (Coleoptera: Scolytidae) . Can. Ent. 94: 1309-1325.

Mathers, W.G. 1935. Time of felling in relation to injury from ambrosia beetles, or pinworms. B.C. Lumberman. 19:14.

Mizell, R.F. Ill, J.L. Frazier, and T.E. Nebeker. 1984. Response of the clerid predator Thanasimus dubius (F.), to bark beetle pheromones and tree volatiles in a wind tunnel. J. Chem. Ecol. 10:177-187. 1.77

Moeck, H.A. 1970. Ethanol as the primary attractant for the ambrosia beetle Trypodendron lineatum (Coleoptera: Scolytidae). Can. Ent. 102:985-995.

Moeck, H.A. 1971. Field test of ethanol as a scolytid attractant. Can. Dept. Fish, and For. Bi-mon. Res. Notes. 27 (2):11-12

Nijholt, W.W. 1967. Moisture and fat content during the adult life of the ambrosia beetle, Trypodendron lineatum (Oliv.) J. Ento Soc. B.C. 64:51-55.

Nijholt, W.W. 1979. The striped ambrosia beetle, Trypodendron lineatum (Olivier): an annotated bibliography. Can. For. Serv. Inf. Rept. BC-X-121.

Nijholt, W.W. and J. Schonherr. 1976. Chemical response behavior of scolytids in West Germany and western Canada. Can. For. Serv. Bi-mon. Res. Notes. 32:31-32.

Nilssen, A.C. 1984. Long-range aerial dispersal of bark beetles and bark weevils (Coleoptera, Scolytidae and Curculionidae) in northern Finland. Ann. Ent. Fenn. 50:37-42.

Paiva, M.R. and K. Kiesel. 1985. Field responses of Trypodendron spp. (Col., Scolytidae) to different concentrations of lineatin and -pinene. Z. ang. Ent. 99:442-448.

Payne, T.L. M.C. Birch, and C.E.J. Kennedy. (Eds.). 1986. Mechanisms in Insect Olfaction. Clarendon Press, Oxford. 3 64 pp.

Pedgley, D. 1982. Windborne Pest and Diseases: Meteorology of airborne organisms. E. Horwood Ltd. England. 250 pp.

Phelan, P.L. and J.R. Miller. 1982. Post-landing behavior of alate Myzus persicae as altered by (E)-B—farnesene and three carboxylic acids. Ent. Exp. Appl. 32:46-53.

Pitman, G.B. and J.P. Vite. 1969. Aggregation behavior of Dendroctonus ponderosae (Coleoptera: Scolytidae) in response to chemical messengers. Can. Ent. 101:143- 149.

Pojar, J. 1983. Forest Ecology. pp. 221-318. In Watts, S.B. (Ed.). Forestry Handbook for British Columbia. For. Undergrad. Soc, Vancouver, B.C. 611 pp.

Prebble, M.L. and K. Graham. 1957. Studies of attack by ambrosia beetles in softwood logs on Vancouver Island, British Columbia. For. Sci. 3:9.0-112. 178

Preiss, R. and E. Kramer. 1986. Pheromone induced anemotaxis in simulated free flight. In Payne, T.L., M.C. Birch, and C.E.J. Kennedy (Eds.). Mechanisms in Insect Olfaction. Clarendon Press, Oxford. 364 pp.

Rabb, R.L. and G.G. Kennedy (Eds.). 1979. Movement of Highly Mobile Insects: Concepts and Methodology in Research. North Carolina St. Univ. Pree. 456 pp.

Rudinsky, J.A. and G.E. Daterman. 1964. Field studies on flight patterns and olfactory responses of ambrosia beetles on Douglas-fir forests of western Oregon. Can. Ent. 96:1339-1352.

Salom, S.M. and J.A. McLean. 1988. Semiochemicals for capturing the ambrosia beetle, Trypodendron lineatum, in multiple funnel traps in British Columbia. J. Ent. Soc. B.C. 85:34-39.

Sanders, C.J. 1985. Flight speed of male spruce budworm moths in a wind tunnel at different wind speeds and at different distances from a pheromone source. Physiol. Entomol. 10:83-88.

SAS Institute. 1985. SAS user's guide: Statistics. SAS Institute, Cary N.C. 956 pp.

Schlyter, F., J. Lofqvist, and J.A. Byers. 1987a. The behavioral sequence in attraction to pheromone sources in the bark beetle Ips typocrraphus. Physiol. Entomol. 12:185-196.

Schlyter, F., J.A. Byers, and J. Lofqvist. 1987b. Attraction to pheromone sources of different quantity, quality, and spacing: Density-regulation mechanisms in bark beetle Ips typoqraphus. J. Chern. Ecol. 13:1503- 1523.

Schmitz, R.F. 1979. Dispersal of pine engraver beetles in second growth ponderosa pine forests, pp. 41-50. In Berryman, A.a. and L. Safranyik (Eds.). Proc. IUFRO Conf., Dispersal of Forest Insects: Evaluation, Theory and Management Implications. Idaho

Shepherd, R.F. 1966. Factors influencing the orientation and rates of activity of Dendroctonus ponderosae Hopkins (Coleoptera: Scolytidae). Can. Ent. 98:507- 518.

Shore, T.L. and J.A. McLean. 1983. A further evaluation of the interactions between the pheromones and two host kairomones of the ambrosia beetles Trypodendron 179

lineatum and Gnathotrichus sulcatus (Coleoptera: Scolytidae). Can. Ent. 115:1-5.

Shore, T.L. and J.A. McLean. 1984. The effect of height of the pheromone-baited traps on catches of the ambrosia beetle, Trypodendron lineatum. J. Ent. Soc. B.C. 81:17-18.

Shore, T.L. and J.A. McLean. 1985. A survey for the ambrosia beetles Trypodendron lineatum and Gnathotrichus retusus (Coleoptera: Scolytidae) in a sawmill using pheromone-baited traps. Can. Ent. 117:49-55.

Shore, T.L. and J. A. McLean. 1988. The use of mark- recapture to evaluate a pheromone-based mass trapping program for ambrosia beetles in a sawmill. Can. J. For. Res. 18:1113-1117.

Shore, T.L., J.A. McLean, and J.C. Zanuncio. 1986. Reproduction and survival of the ambrosia beetle Trypodendron lineatum (Coleoptera: Scolytidae) in Douglas-fir and western hemlock logs. Can. Ent. 119:131-139.

Shorey, H.H. 1973. Behavioral responses to insect pheromones. Ann. Rev. Entomol. 18:349-380.

Southwood, T.R.E. 1962. Migration of terrestial arthropods in relation to habitat. Biol. Rev. 37:171-214.

Southwood, T.R.E. 1978. Ecological Methods. Chapman and Hall, London. 1978. 524 pp.

Stinner, R.E., C.S. Barfield, J.L. Stimac, and L. Dohse. 1983. Dispersal and movement of insect pests. Ann. Rev. Entomol. 1983. 28:319-335.

Tilden, P.E., W.D. Bedard, K.Q. Lindahl, Jr. and D.L. Wood. 1983. Trapping Dendroctonus brevicomis: changes in attractant release rate, dispersion of attractant, and silhouette. J. Chern. Ecol. 9:311-321.

Thompson, G.A. 1985. Vegetation classification of the Endowment Lands. Greater Vancouver Regional District Parks Department, Burnaby, B.C. Tech Paper No. 4. 115 pp.

Vite, J.P. and A. Bakke. 1979. Synergism between chemical and physical stimuli in host colonization by an ambrosia beetle. Naturwiss. 66:528-529.

Vite, J.P., R.I. Gafa, and H.D. von Scheller. 1964. Field observations on the response to attractants of bark 180

beetles infesting southern pines. Contrib. Boyce Thompson Inst. 22:461-470.

Wellington, W.G. 1980. Dispersal and population change. pp. 11-24. In Berryman, A.A. and L. Safranyik (Eds.). Proc. IUFRO Conf., Dispersal of Forest Insects: Evaluation, Theory and Management Implications. Idaho.

Wellington, W.G. 1983. Biometeorology of dispersal. Bull. Entomol. Soc. Amer. 29:24-29.

Wollerman, E.M. 1979. Dispersion and invasion by Scolytus multistriatus in response to pheromone. Environ. Entomol. 8:1-5.

Wood, D.L. 1982. The role of pheromones, kairomones, and allomones in the host selection and colonization behavior of bark beetles. Ann. Rev. Entomol. 27:411- 446.

Wood, S.L. 1982. The Bark and Ambrosia Beetles of North and Central America: A Taxonomic Monograph. Great Basin Naturalist Memoirs, Number 6. Brigham Young University. 1359 pp.

Zar, J.H. 1984. Biostatistical Analysis. Prentice-Hall Inc., New Jersey. 718 pp. 181

APPENDIX 1: RESPONSE OP TRYPODENDRON LINEATUM (OLIVIER) TO DIFFERENT SEMIOCHEMICAL-BAITED SUBSTRATES IN A WIND TUNNEL.

A study was designed to determine the optimal substrate, and its orientation, for use in conducting detailed behavioral observations of Trypodendron lineatum

(Olivier) in a wind tunnel. The substrates tested included a funnel trap, a western hemlock log, and a simulated log.

The dimensions of the latter two substrates are given in

Chapter 2. Both the true and simulated log were placed in horizontal and vertical positions to see if either position was preferred by the beetles.

Experimental procedures such as beetle preparation, release tray, distance to substrates, and semiochemicals are similar to those described in Chapter 2, study 1. One difference is that a wind speed of 0.3 m/s was used here.

The experiment was randomized complete block design, with the six experimental days serving as the blocks.

Separate releases for males and females of 50 beetles each were carried out for each treatment per block. Following release, beetles were given 20 min to respond. The data of interest for the funnel trap treatment was the number of beetles captured, and from the log treatments, the number of beetles landing on them. The data are presented as percentages of the number released, and were subjected to an arc sine transformation prior to analysis (SAS 1985). 182

No significant response differences were found between

= 01 > the substrate treatments for both males (F4 2n i* ' P

= 0.05) and females (F4 2Q 3.34; p > 0.05) (Table 18). No

differences in responses were found between sex (F^ 53 =

1.06; p > 0.05) either.

Based on these results, the easiest substrate to use, the simulated log, was employed in the detailed behavior wind tunnel studies in Chapter 2. This substrate was lightweight and its light coloration allowed for easy observation of beetles as they landed on or approached the substrate. The horizontal orientation of the substrate was chosen, because it simulated windthrown timber and logs normally attacked in the natural setting by ambrosia beetles. 183

Table 18. Mean recapture percentage of Trypodendron lineatum to different substrates baited with semiochemicals in a wind tunnel. Six replicates of 50 beetles, for a total of 300 beetles were released for each treatment by sex category.

Treatment Mean % of Beetles Recaptured + S.E.

Males Females

Funnel Trap 11.7 + 2.5 12.7 + 2.5

Simulated Log 12.7 + 2.5 14.0 + 1.8 (Horizontal)

Hemlock Log 14.0 + 2.0 15. 0 + 1.8 (Horizontal)

Simulated Log 14.7 + 2.8 17.0 + 3.0 (Vertical)

Hemlock Log 13.3 + 1.6 15.0 + 2.7 (Vertical)

No significant differences were found between treatments and sex (Analysis of Variance; p > 0.05). 184

Appendix 2: RESPONSE OF FLOWN AND UNFLOWN TRYPODENDRON LINEATUM TO SEMIOCHEMICAL-BAITED FUNNEL TRAPS IN A WIND TUNNEL.

A.2.1 Introduction

Physiological readiness is an essential prerequisite for dispersing scolytid beetles responding to attractive host material (Borden 1977). For the striped ambrosia beetle, Trypodendron lineatum (Olivier), Graham (1959) found that flight exercise for spring emerging beetles was a necessary prerequisite for switching from an initial photopositive dominated behavior to the chemotropic behavior used in finding host material. Flight exercise has since been documented as part of the dispersal behavior of other scolytid beetles, such as Conopthorus coniperda (Schw.)

(Henson 1962), Dendroctonus ponderosae Hopkins (Shepherd

1966) , Dendroctonus pseudotsucrae Hopkins (Atkins 1969) and

Scolytus multistriatus (Marsham) (Choudhury and Kennedy

1980). The amount of flight exercise needed varies between species as Bennett and Borden (1971) found that it lasts an average of 30 minutes for T. lineatum and 90 minutes for D. pseudotsucrae for tethered beetles on a flight mill.

For the studies presented in this thesis, previously flown beetles (captured in pheromone-baited funnel traps) have been used, due to the difficulty of rearing enough beetles to carry out the experiments satisfactorily. This, however, has posed a problem with respect to interpretation of experimental results. Pre-flown beetles may not 185 accurately reflect behavior of a naturally emerging population in the forest. As a result, a study was carried out to investigate differences in flight response of flown and unflown T. lineatum to semiochemical-baited funnel traps in a wind tunnel. In addition, the effect of storing flown beetles was investigated to see if these beetles needed additional flight exercise before responding to semiochemical-baited traps.

A.2.2 Materials and Methods

Unflown beetles

Western hemlock, Tsuqa heterophylla (Raf.) Sarg. logs, freshly attacked by T. lineatum at the University of British

Columbia Research Forest in Maple Ridge, B.C., were cut and then placed on top of duff material in a 5.0 x 4.3 x 1.2 m canvas tent in June of 1986. The logs remained in the tent until the following winter. Beetles that emerged from the logs in the fall could not escape the tent and were forced to enter the duff material covering the ground. The duff was collected in the winter and stored outside in plastic bags. As temperatures began to warm up in March, 1987, the bags were moved into a walk-in cooler and held at 4 °C.

To extract the beetles, the stored duff material was pulled out of the holding bags and placed at the bottom of sealed 60 x 38 x 27 cm cardboard seedling boxes. The layer of duff placed in the boxes was limited to ca. 8 cm in 186 thickness. Each box was fitted at one end with a 3 00 ml mason jar and a small section of a cloth towel inside of it.

This allowed light into the box, inducing warmed, photopositive beetles to make their way into the jar. Any flight that would occur in the box was considered minimal.

Opening of some boxes showed that beetles walked on the sides of the boxes and probably many walked into the jars.

However, since I did not observe beetle activity in detail at this point, it is impossible to say how much activity did occur.

A total of 21 - 3 0 boxes were used at a time. They were placed in a boiler room on the U.B.C. campus where temperature reached ca. 2 5 °C. Emergence took 2 4 days to complete as 5 re-filled (with duff) sets of boxes were used.

Jars were checked twice daily and emerging beetles were placed in the walk-in cooler stored at 4°C under a photoperiod of 14:10 h (L:D). Details of the storage procedures used are described in Chapter 2. A total of 1100 beetles were collected.

Flown beetles

Collections of flown T. lineatum were made at the beginning of their mass flight period in April, 1987, using lineatin-baited multiple funnel traps, placed adjacent to a large log boom storage area at Foreshore Park, University of

British Columbia Endowment Lands, Vancouver, B.C. Beetles 187 were collected daily and placed in a walk-in cooler under the same conditions as mentioned above.

Flown twice beetles

Flown beetles were taken from the cooler and placed in a 4 L pickle jar. The jar was completely covered with black tar paper except for the top opening which was covered tightly with a fine mesh Lumite^- screen. This provided the only light source for the beetles causing them to spend most of their time in the jar trying to fly to and through the screen. Flight exercise periods of 3 0 minutes were tried initially, however close observation showed that the beetles were in poor physical condition afterwards. Thus, a 15 minute flight exercise period was used in the experiment.

Although Bennett and Borden (1971) determined 30 minutes as the average flight exercise time needed for T. lineatum to respond to olfactory stimuli, no description of different time intervals tested is given. Thus, one must assume this was an arbitrary period of time used for testing. They did report, however, that an insect with 15 minutes of flight exercise exhibited flight arrestment following response to a second exposure to olfactory stimuli.

Test procedures

The study was conducted in the wind tunnel described in

Chapter 2. Two Lindgren 8-funnel traps were placed in the

|7 Chicopee Manufacturing Co., Cornelia, GA.. 30531. 188 upwind section of the tunnel, 0.5 m from the upwind screen.

Both traps were baited with slow release lineatin

(aggregation pheromone) lures. The release rates of the lures are based on lure size (ca. 7.0 ug/day/cm2 jl2-. Two standard lures of (25.8 cm2 each) were split into two, with one half of a lure placed in the top and the other half in the bottom of each trap. Therefore, the total release rate of lineatin in the wind tunnel was ca. 360 ug/24 h. Ethanol

(host attractant) was also used in this study and was released at a rate of 75 mg/24 h from a small plastic vial hung 0.5 m from the ceiling of the tunnel and placed in between both funnel traps.

For all groups of beetles, a walking test was given prior to the experiment to make sure the beetles used were in good physical condition. Beetles were released from a tray 1.5 m downwind from the traps. The tray has been previously described in Chapter 2.

A 3 x 2 x 2 factorial experiment was set up as a completely randomized design. Unflown, flown, and flown twice treatments were divided by sex and by response to unbaited (control) and baited traps. Thus, a total of 12 treatments were tested. The number of beetles released for each treatment ranged from 10 - 24. Lower numbers as well as less replications were used for the control treatments because of the small number of unflown beetles available for the experiment. For all treatments, beetles were given a period of 15 minutes prior to release to warm up from their 189

cold storage period. The air speed in the tunnel was held

at 0.15 m/s (Chapter 2), and the temperature in the tunnel

averaged ca. 23 + 2°C. Each experimental run was set for 10

minutes with the number of beetles caught in the trap

serving as the measurement of interest. The arc sine

transformed data were analyzed as a General Linear Model

(SAS 1985).

A.2.3 Results and Discussion

A three-way interaction between flight treatments, sex,

= 31 and bait was not statistically significant (F2 ^.26 °- ' P

> 0.05). Two-way interactions between flight and sex

F = 08 ( 2,126 °- ' P > 0.05), bait and sex (F1^126 = 2.63; p >

= 88 > 0.05), and bait and flight (F2 126 °* ' P 0-05), also

were not significant. The mean percent of beetles caught in

baited traps was significantly higher, ranging from 6.4 -

15.6 %, than in unbaited traps, ranging from 0.8 - 3.3 %

(Fl,126 = 86-5'' P < 0.05)) (Table 19). In the baited

treatments, response was higher for males than females

(Fl 12 6 = 7-21' P ^ 0.05). For the unbaited group, there were some cases where the observed number of beetles caught

occurred in only one of the replications, resulting in mean values and standard errors that are equal to each other.

Flight exercise treatments did not differ significantly

from each other (F2 126 = 2.88; p > 0.05). However, a

higher percentage of unflown beetles were captured than Table 19. Response of flown and unflown Trypodendron lineatum to semiochemical-baited funnel traps in a wind tunnel.

No. of Total Total Mean % Trt. Sex Groups No. Beetles No. Beetles Caught Per 1 (Reps) Released Caught Group + S.E.

Unbaited (Control)

Unflown M 6 60 2 3.33 + 3.332

F 6 102 1 0.83 + 0.83

Flown M 6 120 1 0.83 + 0.83

F 6 75 1 1.11 + 1.11

Flown 2x M 6 115 2 1.67 + 1.67

F 6 74 2 2.44 + 1.61

Baited

Unflown M 17 300 46 15.64 + 1.92

F 17 315 38 12.09 + 1.77

Flown M 17 337 41 12.03 + 2.34

F 17 343 22 6.43 + 1.18

Flown 2x M 17 341 44 12.80 + 1.76

F 17 335 27 7.98 + 1.66

Data subjected to arc sine transformation.

S.E. equal to mean percentages is a result of observing caught beetles in only one of the replications. 191

flown and twice flown beetles. The most apparent

differences for the baited treatments, was the response of

the unflown males and females (15.6 and 12.1%, respectively)

compared to a response by flown males and females (12.0 and

6.4%, respectively) (Table 19).

The hypothesis that a higher percentage of flown

beetles compared to unflown beetles would respond to

semiochemical-baited funnel traps was rejected for both

sexes. A second hypothesis tested was to see if storing

spring emerged flown beetles at cold temperatures (4 — 2 °C)

for a period of 4 weeks required the beetles to go through

another period of flight exercise before responding to

semiochemicals. This hypothesis was also rejected.

While, the results from this study contrast with the

flight exercise hypothesis first proposed by Graham (1959)

for T. lineatum, and subsequently supported by Francia and

Graham (1967) and Bennett and Borden (1971), a major

weakness in the study reported in this appendix was not

being able to monitor beetle activity prior to collecting

them in the emergence jars. The beetles may have had enough metabolic activity to stimulate their response to

semiochemicals. In fact the higher capture percentage of

the unflown beetles suggests that the flown beetle may have

been in a weaker state due to their longer exposure to warmer temperatures. Nijholt (1967) has shown that fat

content in adult T. lineatum decreases with activity over

time. 192

These results indicate that the amount of metabolic activity needed for T. lineatum response to semiochemicals should not alter our interpretation of experimental results where previously flown beetles have been used. 193

APPENDIX 3: INFLUENCE OF DUSTING T. LINEATUM WITH FLUORESCENT POWDER ON THEIR RESPONSE TO SEMIOCHEMICALS IN A WIND TUNNEL.

To study dispersal of Trypodendron lineatum (Olivier),

mark-recapture experiments were designed using fluorescent

powder as the marking agent. This technique was previously

used by Shore and McLean (1988) at a sawmill, but no

information was gathered to see what effect, if any, the

mark had on beetle behavior. The objective of this study

was to see if marking beetles with fluorescent powder, or

fluorescent powder and gum arabic, as suggested by Southwood

(1978), would reduce the response of T. lineatum to

semiochemicals.

The beetles used in this study were from the same group

of beetles collected in 1986 for the studies described in

Chapters 2 and 5. Beetles were marked using the vacuum

duster technique (Linton et al. 1987). The treatments used

were pure fluorescent powder, a mixture of 1/2 powder and

1/2 gum arabic, and unmarked beetles. One gram of dust was

used to lightly mark beetles for each group of 200 beetles.

Following marking, beetles were returned to the walk-in

cooler (4°C) where they had been stored since their capture.

Response of unmarked and marked beetles to a

semiochemical-baited (ethanol and lineatin) funnel trap was

tested in the wind tunnel described in Chapter 2. Beetles were tested 1, 3, 5, and 7 days following marking. Between

34 and 50 beetles were released from a tray 2 m upwind from 194 the traps at an airflow of 0.3 m/s. The beetles were given

45 min to respond, and the number caught in the traps were subsequently tallied.

A randomized complete block design was used. Days served as blocks, and all treatments were tested within each block. Males and females were tested separately. The data were analyzed by analysis of variance (SAS 1985).

Male and female responses did not differ significantly

(Fl,19 = °-87' P > 0.05), therefore their data were pooled.

No differences were found in the number of beetles caught

for any of the treatments (F2 18 = 0.74; p > 0.05), although slightly more marked beetles were captured than control beetles (Table 20). Thus, it was concluded that lightly marking T. lineatum with fluorescent powder does not adversely affect their flight response to semiochemical- baited funnel traps. 195

Table 20. The effect of marking treatments on T. lineatum response to a semiochemical-baited funnel trap in a wind tunnel.1

Treatment No. of Beetles Mean % Captured Released + S.E.2

Fluorescent 374 12.0 + 1.7 Powder

Fluorescent Powder & Gum 371 11.5 ± 1.6 Arabic

Unmarked Beetles (Control) 361 9.7 + 2.3

1 Male and female data were pooled together. 2 . Data were subjected to an arcsine transformation.