UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Warren’s collar weevil in lodgepole pine stands in the Kispiox Forest District 1994

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
ubc_1994-0373.pdf
ubc_1994-0373.pdf [ 1.78MB ]
ubc_1994-0373.pdf
Metadata
JSON: 1.0075249.json
JSON-LD: 1.0075249+ld.json
RDF/XML (Pretty): 1.0075249.xml
RDF/JSON: 1.0075249+rdf.json
Turtle: 1.0075249+rdf-turtle.txt
N-Triples: 1.0075249+rdf-ntriples.txt
Citation
1.0075249.ris

Full Text

WARREN’ S COLLAR WEEVIL IN LODGEPOLE P1IIE STMDS IN THE KISPIOX FOREST DISTRICT by GEOFFREY THOMAS BYFORD B.Sc., Simon Fraser University, 1988 A THESIS SUBMITTED iN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Faculty ofForestry, Department ofForest Sciences) We accept this thesis as conforming THE UNIVERSITY OF BRITISH COLUMBIA June 1994 © Geoffrey Thomas Byford, 1994 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 Fc-c ScitC6S The University of British Columbia Vancouver, Canada Date_______________________ ABSTRACT The distribution and abundance of Warren’s collar weevil, Hylobius warreni Wood (Coleoptera: Curculionidae), was examined in lodgepole pine, Pinus contorta var. latfolia Engelni, plantations in the Kispiox Forest District in north-central British Columbia. The effect of weevil feeding damage on height growth of dominant and co-dominant trees was also examined. The weevil was found distributed throughout the forest district. All 31 surveyed plantations, ranging in age from 5 years to 16 years, had evidence of larval feeding damage. Weevil-caused mortality ranged from 0-8.8%. The average percentage of sampled trees attacked was 29%. The percentage of trees attacked within a plantation was directly related to plantation age and average tree height. There was not a significant relationship between the thickness of the organic layer above mineral soil and the percentage of trees attacked in the first year of the study. Plantation density had no apparent effect on the percentage of trees attacked. The highest levels of weevil damage were found on circum-mesic, well drained sites in the ICHmc3 biogeoclimatic variant. Plantations established on sites which originally had a pine component appeared to be particularly susceptible to weevil damage. It appears that the collar weevil is not reducing plantations to below minimum stocking levels (<700 sph). The populations of the collar weevil found in 2 of the 3 intensively sampled plantations, ranging in age from 13-17 years old, were similar to those reported for naturally regenerated stands of similar age in Alberta. One plantation had populations considerably higher than reported elsewhere. This may have been due to a scarcity of lodgepole pine in the surrounding timber types. The percentage of stems attacked within a diameter class increased proportionally with increasing diameter class, and increasing basal area class. The percentage of the stem 11 circumference girdled increased with increasing root collar diameter and LFH layer depth. The percentage of lodgepole pine attacked within the 3 study sites ranged from 81 % to 92 %. These attack rates were higher than rates reported for naturally regenerated stands. This may have been due to lower tree densities found in planted versus naturally regenerated stands. The high incidence of root collar weevil within the Kispiox Forest District may be related to a man-caused increase in lodgepole pine within the area. This increase has occurred because of the conversion of mature mixed species coniferous stands to lodgepole pine plantations following clearcutting. Destructively sampled trees indicated that initial weevil attacks within plantations had occurred within the previous 6 years. The time of first attack was uniform within plots indicating that weevils were not dispersing from the stand margin. Results from stem analysis indicated that height growth of weevil damaged trees was not affected in the short term. The long term impacts of sub-lethal multiple weevil attacks and larval damage on height growth are not presently known. 111 TABLE OF CONTENTS PAGE Abstract ii Table of Contents iv List of Tables vi List of Figures vii Acknowledgements ix 1.0 Introduction 1 1.1 Warren’s Collar Weevil 2 1.2 Biology and Life History 4 1.3 Effects of Warren’s Collar Weevil at the tree level 5 1.4 Effects of Warren’s Collar Weevil at the stand level 7 1.5 Control 8 1.6 Susceptible sites 9 1.7 Progression of weevil infestations with stand age 10 1.8 Warren’s Collar Weevil in British Columbia 11 2.0 Methods 12 2.1 Study Area 12 2.2 Distribution and abundance ofWarren’s Collar Weevil 13 2.2.1 Selection of sampling method and plot type 13 2.2.2 Sampling intensity and maximum survey area 13 2.2.3 Plantation selection 15 2.2.4 Plot information 15 2.2.5 Plot summaries 17 2.2.6 Within plantation distribution 17 2.2.7 Historical information 19 2.2.8 Survey summary 19 Iv 2.3 Height growth and dispersal study 19 2.3.1 Study locations 19 2.3.2 Sampling methodology 21 2.3.3 Height growth study 22 3.0 Results 24 3.1 Distribution and abundance of Warren’s Collar Weevil 24 3.1.1 Selection of plot type 24 3.1.2 Plantation distribution 24 3.1.3 Plantation summaries 24 3.1.4 Weevil incidence in relation to site and stand factors 28 3.1.5 Within plantation distribution 33 3.1.6 Within district distribution 34 3.2 Height growth and dispersal study 34 3.2.1 Plot summaries 34 3.2.2 Pattern of attack 44 3.2.3 Stem analysis 44 4.0 Discussion 60 4.1 Distribution and abundance of Warren’s Collar Weevil 60 4.2 Height growth and dispersal study 67 5.0 Conclusions 72 6.0 Recommendations 74 Bibliography 76 V LIST OF TABLES PAGE Table 1. Summary data for plantations surveyed for Hylobius warreni in the Kispiox Forest District, Hazelton B.C., Summer 1989 26 Table 2. Number and percentage of sample trees attacked by Hylobius warreni, by host species, at Date Creek in the Kispiox Forest District, Summer 1990 35 Table 3. Population estimates ofHylobius warreni and characteristics of attacked lodgepole pine trees at Date Creek in the Kispiox Forest District, Summer 1990 35 Table 4. Number and percentage of sample trees attacked by Hylobius warreni, by host species, at Mosquito Flats in the Kispiox Forest District, Summer 1990 40 Table 5. Population estimates ofHylobius warreni and characteristics of attacked lodgepole pine trees at Mosquito Flats in the Kispiox Forest District, Summer 1990 40 Table 6. Number and percentage of sample trees attacked by Hylobius warreni, by host species, at Shandilla in the Kispiox Forest District, Summer 1990 45 Table 7. Population estimates ofHylobius warreni and characteristics of attacked lodgepole pine trees at Shandilla in the Kispiox Forest District, Summer 1990 45 Table 8. Timing of attack ofHylobius warreni on sample trees in a 13 year old lodgepole pine plantation at Date Creek in the Kispiox Forest District, Summer 1990 49 Table 9. Timing of attack ofHylobius warreni on sample trees in a 17 year old lodgepole pine plantation at Mosquito Flats in the Kispiox Forest District, Summer 1990 50 Table 10. Timing of attack ofHylobius warreni on sampled trees in a 17 year old lodgepole pine lantation at Shandilla in the Kispiox Forest District, Summer 1990 51 Table 11. Stem analysis of lodgepole pine trees attacked by Hylobius warreni at Date Creek in the Kispiox Forest District, Summer 1990 53 Table 12. Stem analysis of lodgepole pine trees attacked by Hylobius warreni at Mosquito Flats in the Kispiox Forest District, Summer 1990 56 Table 13. Stem analysis of lodgepole pine trees attacked by Hylobius warreni at Shandilla in the Kispiox Forest District, Summer 1990 58 vi LIST OF FIGURES PAGE Figure 1. Schematic representation of the layout of two plot types tested for assessing the incidence ofHylobius warreni damage in the Kispiox Forest District, Hazelton, B.C., Summer 1989 14 Figure 2. Summary plot map for a plantation surveyed for Hylobius warreni damage in the Kispiox Forest District, Hazelton, B.C., Summer 1989 18 Figure 3. Hylobius warreni study sites in the Kispiox Forest District, Summer 1990 20 Figure 4. The distribution of pine plantations surveyed for Hylobius warreni damage in the Kispiox Forest District, Hazelton, B.C., Summer 1989 25 Figure 5. Age distribution of pine plantations surveyed for Hylobius warreni damage, Kispiox Forest District, Summer 1990 27 Figure 6. Percentage of sample lodgepole pine with Hylobius warreni damage in relation to plantation age in the Kispiox Forest District, Summer 1990 29 Figure 7. Percentage of sample lodgepole pine with Hylobius warreni damage in relation to average tree height in the Kispiox Forest District, Summer 1990 30 Figure 8. Percentage of sample lodgepole pine with Hylobius warreni damage in relation to plantation density in the Kispiox Forest District, Summer 1990 31 Figure 9. Percentage of sample lodgepole pine with Hylobius warreni damage in relation to average organic matter layer depth in the Kispiox Forest District, Summer 1990 32 Figure 10. Diameter distribution of sampled lodgepole pine trees and proportion attacked by Hylobius warreni at Date Creek in the Kispiox Forest District, Summer 1990 36 Figure 11. Percentage of sample lodgepole pine with Hylobius warreni damage in relation to root collar diameter at Date Creek in the Kispiox Forest District, Summer 1990 37 Figure 12. Percentage of sample lodgepole pine with Hylobius warreni damage in relation to average organic matter layer depth at Date Creek in the Kispiox Forest District, Summer 1990 38 Figure 13. Diameter distribution of sampled lodgepole pine trees and proportion attacked by Hylobius warreni at Mosquito Flats in the Kispiox Forest District, Summer 1990 41 vi’ Figure 14. Percentage of sample lodgepole pine with Hylobius warreni damage in relation to root collar diameter at Mosquito Flats in the Kispiox Forest District, Summer 1990 42 Figure 15. Percentage of sample lodgepole pine with Hylobius warreni damage in relation to average organic matter layer depth at Mosquito Flats in the Kispiox Forest District, Summer 1990 43 Figure 16. Diameter distribution of sampled lodgepole pine trees and proportion attacked by Hylobius warreni at Shandilla in the Kispiox Forest District, Summer 1990 46 Figure 17. Percentage of sample lodgepole pine with Hylobius warreni damage in relation to root collar diameter at Shandilla in the Kispiox Forest District, Summer 1990 47 Figure 18. Percentage of sample lodgepole pine with Hylobius warreni damage in relation to average organic matter layer depth at Shandilla in the Kispiox Forest District, Summer 1990 48 Figure 19. Percentage of sample lodgepole pine with Hylobius warreni damage in relation to basal area class for three plantatations in the Kispiox Forest District, Summer 1990 52 Figure 20. Cumulative average tree height of lodgepole pine, in four Hylobius warreni girdling classes, at Date Creek in the Kispiox Forest District, Summer 1990 54 Figure 21. Cumulative average tree height of lodgepole pine, in four Hylobius warreni girdling classes, at Mosquito Flats in the Kispiox Forest District, Summer 1990 57 Figure 22. Cumulative average tree height of lodgepole pine, in four Hylobius warreni girdling classes, at Shandilla in the Kispiox Forest District, Summer 1990 59 VII’ ACKNOWLEDGEMENTS I thank the Prince Rupert Region of the B.C. Ministry of Forests whose financial support made this thesis possible; specifically I thank Mr. Tim Ebata, Regional Entomologist. I also thank the staff at the Kispiox Forest District for there support and services; Ms. Barb Costerton, Mr. Jeff Lemieux, Mr. Andrew Reviakin, Mr. Ian Williamson, and Mr. Ian Wilson for there assistance in data collection. I thank the Science Council of British Columbia for financial support through the Graduate Research Engineering and Technology Award Program (G.R.E.A.T.). I would like to thank Interior Reforestation Co.Ltd for their support during the preparation of this thesis and Mr. John Przeczek for providing comments on the manuscript. I thank the members of my committee: Dr. Peter Marshall, Dr. Tom Sullivan and Dr. Gordon Weetman. I would also like to thank Drs. Burton and Kozak. I am especially indebted to my Faculty Advisor, Dr. John McLean, whose patience and valuable advice have enabled me to complete this thesis. Finally, I would like to thank my family for support throughout my scholastic career, especially my wife Jackie for her constant encouragement and unfailing support. ix 1.0 INTRODUCTION The Kispiox Forest District lies within a coastal-interior transition zone in north-central British Columbia. The vegetation exhibits characteristics of a coastal maritime climate and an interior cordilleran climate (Haeussler et a!. 1985; Meidinger and Pojar 1991). As a result of this transitional climate, the forest cover consists ofboth coastal and interior tree species. The major tree species include lodgepole pine (Pinus contorta var. latfolia Engelm.), hybrid spruce (Picea x glauca x sitchensis x engelmannii), sub-alpine and amabilis fir (Abies lasiocarpa (Hook.) Nutt. and A. amabilis (Dougl.) Forbes), western hemlock (Tsuga heterophylla (Raff.) Sarg.), western redcedar (Thujaplicata Donn), trembling aspen (Populus tremuloides Michx), and paper birch (Betulapapyrfera Marsh.). The timber profile in the district has an overabundance of mature/over-mature western hemlock and true fir timber types (>140 years old). More than 80% of the timber inventory consists of these two species (Ministry ofForests 1981). Timber allocation within the district has concentrated on harvesting these timber types in an attempt to minimize losses expected from insects, disease and windfall. Reforestation of harvested areas is usually achieved by planting spruce and lodgepole pine. Lodgepole pine is planted on drier sites and spruce is planted on wetter sites. The future forest profile will consist of more spruce and pine than exists in the current inventory due to the replacement of hemlock and true fir forests with these two species. Approximately 12 000 ha of pine have been planted in the Kispiox Forest District as of 1988 (Ministry of Forests 1988). This figure is projected to increase as more areas are harvested. Associated with this increase in the pine component of the profile is a potential increase of insects and diseases which attack lodgepole pine. One such insect is Warren’s collar weevil (Hylobius warreni Wood .(Coleoptera: Curculionidae)). Ministry of Forests staff, Licensees and silviculture contractors have reported increasing incidence of this insect in lodgepole pine plantations within the Kispiox Forest District in recent years. These increases were of concern because of the abundance of young lodgepole pine 1 plantations within the district. Of particular concern was the potential reduction in stocking due to weevil caused mortality in young stands. As part of the Forest Resources Development Agreement (FRDA I: 1985199O)1 between the Province and the Federal government, funding was made available to assess the distribution and abundance ofWarren’s collar weevil in pine plantations within the district. The objective of the project was to provide a better understanding of the relationship between Warren’s collar weevil and lodgepole pine plantations. The specific objectives of the study were: 1) To determine the distribution and abundance of Warren’s collar weevil in 6 to 16 year old lodgepole pine plantations within the Kispiox Forest District. (2) To assess the impacts of weevil injury on the development of lodgepole pine stands to the free-growing stage. (3) To determine if weevil incidence is related to predictable site or stand factors. A hazard map could be produced, using these factors to identify potential problem areas. (4) To determine timing of the first and subsequent attacks on planted lodgepole pine. (5) To determine the course of infestations within pine plantations. (6) To determine the impact of weevil feeding damage on the height growth of lodgepole pine. 1.1 Warren’s Collar Weevil The larvae of Warren’s collar weevil feed in the root collar region and on the larger lateral roots of a number of coniferous tree hosts (Fumiss and Carolin 1977). Hosts include various species of pines, Pinus spp., spruces, Picea spp., true firs, Abies spp. and tamarack, Larix laricina (DuRoi) K.Koch (Warner 1966; Grant 1966). In western North America, the primary host species is lodgepole pine, while white spruce (Picea glauca (Moench) Voss) is an alternate and less preferred species (Reid 1952; Stark 1959; Cerezke 1969). Larval feeding results in partial or complete girdling of the major lateral roots and the root collar of the host tree (Reid 1952). Intensive Forest Management Program - Brushing and Weeding and Pest-control Sub-program 2 Damage to the root collar region of the tree may cause death either directly or indirectly. Complete girdling of the phloem and cambium tissue at the root collar results in direct mortality while indirect mortality may occur as a result of root decay flingi entering through the larval wound (Warren 1956a; Stark 1959). Much of the early work on the biology ofWarren’s collar weevil and its relationship with its host tree species was done in the 1950’s and early 1960’s in Manitoba, Saskatchewan and Alberta (Warren and Whitney 1951; Whitney 1952; Reid 1952; Warren 1956a-c, 1958; Stark 1959; Whitney 1961). During this period, the taxonomy of the genus was being clarified by a number of researchers (Wood 1957; Warren 1960; Finnegan 1961; Wilson eta!. 1966; Warner 1966). Warren’s collar weevil was first recognized as a distinct species by Wood (1957). Prior to this, it was identified as Hypomolyxpiceus (De 0.). Wood separated H. piceus into two distinct species: Hylobius warreni and Hylobiuspinicola (Couper). The primary basis for the separation was wing form. H. warreni has vestigial wings and is a flightless species while H. pinicola has fhnctional wings. The majority of our current knowledge regarding Warren’s collar weevil was obtained through the work of Dr. Herb Cerezke in Alberta during the 1960’s and early 1970’s. He published a number of papers including works on the basic biology of the weevil, its effect on host trees, and its population dynamics in relation to forest management practices (Cerezke 1967, 1969, 1970a-c, 1972, 1973a-b, 1974). 3 1.2 Biology and Life History 1.2.1 Description Warren’s collar weevil is a relatively large dark brown to blackish beetle with white to pale yellow scales or dots on the elytra (Wood 1957). They range in length from 11.7 mm to 15.1 mm, with the females being slightly larger than the males. The sub-globose eggs are a translucent white and range from 0.5-0.8 mm (Duncan 1986). The larvae are creamy white with a tan coloured head capsule. They pass through six larval instars and a brief pre-pupal stage (Stark 1959). The pupae are approximately 10 mm in length and appear much as the adults do. 1.2.2 Life History Female weevils deposit eggs in the root collar region of susceptible trees from late June to August (Stark 1959; Cerezke 1969). The egg is deposited in a protective niche which is excavated by the female and covered over with excrement (Warren and Whitney 1951; Cerezke 1969). The larvae emerge an average of 42 days after oviposition (Cerezke 1969). After emerging, first instar larvae feed randomly on the phloem layer of host trees. However, as the larvae reach the second and third instars feeding becomes more circumferentially oriented and often extends into the cambium (Cerezke 1969). Depending on the time of eclosion, weevil larvae may overwinter as first to fourth instars (Stark 1959; Cerezke 1969). During this time larvae are inactive, and do not commence feeding until the following spring. It is during this second year of larval feeding that the host tree sustains the majority of damage (Stark 1959; Cerezke 1969). Larvae overwinter a second time, usually as fifth or sixth instars. Larval wounding of the tree base results in the host tree exuding resin at the wounded area. The larvae use the pitch to form a protective feeding gallery or tunnel. These tunnels are composed of a matrix of bark particles, frass and resin, and act as a barrier for the developing larvae, protecting them from potential predators, parasites and dessication (Warren and Whitney 1951; Warren 1956a; Cerezke 1969). 4 In the spring of the third year larvae feed briefly before constructing a pre-pupal chamber a few centimetres from the base of the tree. This chamber consists of a matrix of resin, frass and bark particles (Stark 1959; Cerezke 1969). The larvae spend a short time in the pre-pupal stage before pupating in June (Cerezke 1969). Adults emerge after a 4 week pupation period (Cerezke 1969). Adult emergence occurs from late June through August (Warren and Whitney 1951; Stark 1959; Cerezke 1969). The adults can survive for up to four years and they are reproductively mature for at least three years (Cerezke 1970a, 1973). They are nocturnal and emerge from the dufl at dusk and either disperse to find suitable host trees or ascend a host to feed in the crown (Cerezke 1969, 1970a). They feed on needles and bark on the upper portion of the branches, and occasionally on the terminal buds (Cerezke 1969; Warren 1956a; Stark 1959). Adults may also feed on the bark of small roots (Warren and Whitney 1951). Adult feeding rarely causes significant damage but may cause the terminal shoots to become twisted and distorted (Cerezke 1969). Dispersal is ambulatory, as Warren’s collar weevil has lost the capability for flight (Wood 1957; Warren 1960; Cerezke 1969). 1.3 Effects of Warrren’s Collar Weevil at the Tree Level Warren’s collar weevil may affect its host tree in a number of ways. Larval feeding may result in direct tree mortality due to mechanical injury of the root collar (Warren 1 956c). Indirect mortality may result through windthrow of attacked trees whose roots have been weakened by larval injury (Cerezke 1969). Indirect mortality may also occur through attack by secondary coniferophagous insects, or fungi, invading a tree weakened by larval damage (Warren and Whitney 1951; Smerlis 1957; Whitney 1961). Weevil damage may also cause growth reductions and changes in the anatomical structure of attacked trees (Cerezke 1969, 1970c, 1972, 1974). 2 Organic layer above mineral soil 5 1.3.1 Mortality The most evident effect of Warren’s collar weevil larval feeding is direct mortality of the host tree. Generally, direct mortality occurs in trees less than 30 years old, and rarely occurs in mature trees (Warren 1956a; Cerezke 1969, 1970a-b, 1974). The reason for this is twofold. First, in younger trees larval feeding tends to be oriented circumferentially around the root collar, whereas in older, larger trees feeding tends to occur in patches and on the larger lateral roots (Cerezke 1970a). The feeding pattern in younger trees causes the flow of nutrients from the crown to the roots to be severed, resulting in root and tree death. In older trees there is often a strip of undamaged bole which continues to supply the roots with nourishment through the undamaged phloem tissue. Second, the more mature roots of older lodgepole pine may be more resistant to weevil feeding than those of younger trees (Cerezke 1970b). There is also evidence that older trees may repair tissue damage caused by root collar weevil larvae (Cerezke 1969). Larval wounding may pre-dispose trees to attack by secondary organisms. Warren and Whitney (1951) found that larval wounds acted as infection courts for the entrance of the root rot fungi Armillaria and Polyporus in white spruce. It is these secondary organisms which cause tree death. There is little evidence of any association between root rots, or other decay fungi, and Hylobius wounds in lodgepole pine (Stark 1959; Cerezke 1969). Mortality may also be caused by secondary bark beetle colonization of weevil-injured trees. The expenditure of energy by the trees to pitch-out weevils and repair larval wounds may reduce their ability to defend against other tree feeding organisms (Coulson and Witter 1984). Larval feeding may also weaken the trees root system to such a degree that it becomes susceptible to windthrow or snowpress (Duncan 1986). 1.3.2 Anatomical Effects Weevil injured trees have a number of ways of compensating for tissue injury. They may increase the radial growth of wood on both the girdled and non-girdled portions of the stem to compensate for loss of strength on the wounded portion. This may occur as a °budding” type of growth which seals off the damaged area and increases the area of conductive tissue (Cerezke 1974). The tree may also respond to weevil injury by producing traumatic vertical resin ducts in 6 the growth rings directly above the wounded portion of the stem (Cerezke 1972). This greatly increases the supply of resin to the injured area and may be an attempt by the tree to “pitch-out” the weevil larvae. Although the tree is rarely successful in pitching-out weevils, the flow of resin may be beneficial in coating the wound and sealing off the injured area from the spores of pathogenic fungi. A third mechanism by which the host compensates for weevil injury is the production of adventitious roots above the girdled area (Cerezke 1969). These roots compensate for the loss of conductive tissue below the wound and provide the crown of the tree with a partial supply of water and minerals. 1.3.3 Growth Losses Partial girdling at the root collar and on lateral roots may lead to significant reductions in both radial and height increment (Cerezke 1969, 1970c, 1974). Losses of 17.16% in mean radial growth and 11.50% in mean height growth were recorded near Robb, Alberta in the second and third years following 50% girdling of lodgepole pine stems (Cerezke 1970a). There were no significant differences in either radial or height growth between attacked and unattacked trees on an ancillary site near Grande Prairie (Cerezke 1 970c). Cerezke (1974) artificially girdled the root collars of lodgepole pine trees and found that there was a general decline in height growth with increased circumference of the root collar girdled. Height increment decreased gradually until 60- 80% of the root collar circumference had been girdled, after which there was a rapid decrease. However, losses in radial increment did not exceed 5% when compared with controls for a range of girdling classes between 0 and 90%. 1.4 Effects of Warren’s Collar Weevil at the Stand Level Warren’s collar weevil infestations have a number of consequences at the stand level. Mortality of individual or groups of trees may result in pockets ofNot Satisfactorily Restocked (NSR) areas within a stand (Herring and Coates 1981). Similar patches of windthrown or snowpressed trees may also occur (Duncan 1986). 7 The primary impact ofH. warreni at the stand level is growth loss. The cumulative loss of height and/or diameter growth of individual stems must lead to reduced volume production within a stand. Stand level volume loss attributable to root collar weevil damage has not been quantified; however, Cerezke (1969) states that losses to dominant and co-dominant trees may be considerable on good growing sites. 1.5 Control The strategies for controlling Warren’s collar weevil populations may be grouped into two broad classes: chemical controls and silvicultural controls. Chemical controls are implemented after stands have become infested, while silvicultural controls are designed to prevent weevils from invading or re-invading a stand. 1.5.1 Chemical Controls During the 1950’s, ethylene dichloride was effective in killing larvae of Warren’s collar weevil and benzene hexachloride (Lindane© ) was effective in preventing the re-invasion of previously attacked trees (Warren 1 956a). These chemicals were applied directly to the base of infested trees and were extremely toxic. Carbosulfan granular insecticides have been used to control a related Hylobius sp. in Europe (Castellano and Marsh 1987). These are also applied directly to the root collar of individual trees. The cost of this labour intensive activity limits their use to intensively managed, high value plantations, and precludes their use in more extensively managed timber production plantations. . The negative impact of such chemicals on non-target organisms must also be considered. 1.5.2 Silvicultural Controls The first step in implementing silvicultural controls for Warren’s collar weevil is identifying high risk stands which are scheduled for harvest. After identification of susceptible stands, it is necessary to develop a prescription which will reduce weevil populations in the subsequent stand. The most effective means of reducing populations is through clearcut 8 harvesting and site preparation (Cerezke 1970a). Infested stands should be clearcut harvested leaving no residual susceptible trees, and site prepared either with scarification or prescribed burning (Cerezke 1969, 1970a). Clearcut harvesting reduces weevil populations by up to 88%, and site preparation results in a further reduction of the residual population by removing or reducing the duff layer, which is critical for larval survival (Cerezke 1973b). Weevil populations in the subsequent stand are reduced accordingly. There are also silvicultural controls which can be applied after weevils have invaded a stand. These involve screefing the area around the base of infested trees and pruning the lower branches (Warren 1956a; Wilson 1967). Screefing removes the LFH layer at the root collar of host trees thus greatly reducing the weevil habitat. Pruning of the lower branches increases the light and heat around the tree base. This dries out the LFH layer making it unsuitable for oviposition by adult females. As for chemical controls, pruning and screefing may be effective in higher value stands but the, high labour costs make them unsuitable on an operational basis for lower value timber production plantations. 1.6 Susceptible Sites The relationship between forest site type and weevil abundance was first recognized by Warren (1956a-b). He found that damage in spruce stands in Manitoba increased as the average moisture content of the humus layer increased. The highest population of Warren’s collar weevil was found on a very wet site characterized by a peat layer over a gleyed soil. In B.C. and Alberta, the weevil is found on drier sites. Stark (1959) found the weevil distributed throughout the forested regions of Alberta, predominantly in the boreal forest. Cerezke (1969, 1970a-b) further refined the weevils site requirements. He found the weevil primarily on moist, rich sites within the boreal region. These were sites with good drainage characteristics, strong soil development and an abundance of herbaceous, moss and shrub species. In Alberta, the weevil shows a definite preference for lodgepole pine stands (Cerezke 1970b). Stands at low elevations have a higher incidence of weevil damage than stands at higher elevations (> 1600m). 9 In central British Columbia, plantations with the highest incidence of weevil were found on coarse textured, well drained soils, while those found on wet sites with poor drainage were less seriously affected (Herring and Coates 1981). In northwestern B.C., Garbutt (1988) found that weevil damage was confined to trees within one biogeoclimatic variant, the ICHmc3 (i.e. Interior- Cedar-Hemlock, moist, cold- Lower Nass Basin) (Meidinger and Pojar 1991). Susceptible sites within this variant were circum-mesic sites with well drained soils. Warren’s collar weevil is found on sites containing a relatively thick duff layer composed of a mixture of moss and herb species (Cerezke 1970a). There was a strong relationship between the depth of this layer at the tree base and the distribution ofweevils within a stand. In general, the number of weevils per tree increased as the thickness of the duff increased (Cerezke 1969). This relationship was stronger as tree size increased. 1.7 Progression of Weevil Infestations with Stand Age Host trees become susceptible to weevil attack between the ages of 6 and 16 years, when they reach a size of 1-1.5 m tall and 5 cm in diameter at stump height (30 cm) (Warren and Whitney 1951; Reid 1952; Cerezke 1969, 1 970a). Warren’s collar weevil attacks healthy, vigorous trees and shows a preference for dominant and co-dominants, although intermediate and suppressed trees may also be attacked (Reid 1952). Once weevils become established in a stand they remain there until stand replacement (Warren and Whitney 1951; Reid 1952; Cerezke 1969). The initial infestation of a regenerating pine stand is believed to originate from surrounding older stands of a susceptible host species (Cerezke 1969). In 15-25 year old lodgepole pine stands in Alberta, Cerezke (1969,1 970b) found that infestations spread into pine regeneration at a rate of 10-15 rn/year from adjacent infested mature pine. Trees at the periphery of a plantation or naturally regenerating cutbiock are attacked first, and the infestation proceeds inward as the population increases. Cerezke (1969) found an increase in attack incidence within stands between the ages of 10 and 60 years. After 60 years, the infestation rate levelled off when up to 90% of trees within a stand had evidence of previous weevil injury. Generally, there is an 10 increase in the number of weevil larvae found on individual trees as stand age increases, however the total population of larvae within a stand during its life remains relatively constant. This is due to natural stand thinning processes which result in the same number ofweevils being concentrated on a fewer number of stems (Cerezke 1970c). The final stage in the cycle of a weevil population occurs when an infested stand is replaced.. This can occur either by natural means such as wildfire, or by artificial means such as harvesting. Following clearcutting of infested stands, large increases in weevil populations in surrounding stands have been recorded (Cerezke 1969, 1973b). It is believed that surviving adult weevils migrate from the clearcut into surrounding trees. The adjacent timber, whether mature or immature, provides suitable habitat to maintain populations of the weevil. This “reservoir” population then re-invades the regenerating stand once trees reach a susceptible size. 1.8 Warren’s Collar Weevil in British Columbia In British Columbia, outbreaks of Warren’s collar weevil have traditionally been sporadic and localized. Recently, however, populations of the weevil have been on the increase in lodgepole pine stands in the Prince George and Prince Rupert Forest Regions. The first report of the insect having adverse impacts on forest stands occurred in the Prince George Forest Region (Herring and Coates 1981). Warren’s collar weevil was found in 9 of 11 lodgepole pine plantations. The percentage of living trees attacked ranged from 1.2% to 5.0%, and the mortality rate ranged from less than 1% to 8.2%. They estimated an annual increase of between 1.2% and 5.0% in the percentage of trees attacked within a plantation. Warren’s collar weevil has also increased in the Prince Rupert Forest Region in recent years. Specifically, there have been increases in populations of the insect in lodgepole pine plantations near Hazekon (Garbutt 1988). An average of 76% of 4 to 20 year old lodgepole pine trees sampled in five plantations were infested with weevil larvae (Garbutt 1988). 11 2.0 METHODS The study was completed over two field seasons: from May 1989 to August 1989 and from May 1990 to August 1990. The first field season concentrated on assessing the distribution and abundance of Warren’s collar weevil within the Forest District. The second field season concentrated on assessing the epidemiology of Warren’s collar weevil and the potential impact it may have on the height growth of lodgepole pine. 2.1 Study Area The study was done in the Kispiox Forest District in north-central British Columbia and all survey areas were in the Interior Cedar-Hemlock biogeoclimatic zone, moist cool biogeoclimatic subzone (Meidinger and Pojar 1991). The climate is transitional between coastal and interior influences. It is characterized by warm moist summers, cool wet falls, and cold winters. The average precipitation ranges from 500 - 1200 mm per year. The average daily temperature ranges from -13.9°C for the coldest month to 22.5°C for the warmest month (Haeussler eta!. 1985). The majority of sites surveyed over the two year period were within the ICHmc3 variant of the subzone (Meidinger and Pojar 1991) (ICHg3 under the previous classification system (Haeussler et a!. 1985)). The major tree species include lodgepole pine, hybrid spruce, subalpine fir, amabilis fir, western hemlock, western redcedar, trembling aspen, and paper birch. As the name of the zone implies, hemlock and cedar are the true climatic climax species, however their distribution has been limited by a history of frequent natural fires and human disturbance. Because of this frequent disturbance, the variant is characterized by widespread seral stands of aspen, paper birch, hazelnut (Corylus cornuta Marsh.) and other shrubs (Haeussler eta!. 1985). 12 2.2 Distribution and Abundance of Warren’s Collar Weevil To assess the distribution and abundance of Warren’s collar weevil in the Kispiox Forest District, a representative sample of pine plantations within the district had to be surveyed. The target plantations to survey were 5-to 16-year-old pine stands within the ICHmc3 variant (Meidinger and Pojar 1991), as these plantations were considered to be the most susceptible to colonization by Warren’s collar weevil. Additionally, the target was to sample a minimum of 30 pine plantations over the 4 month period. 2.2.1 Selection of sampling method and plot type Before proceeding with the survey, an appropriate sampling method had to be chosen. Two plot types were selected for evaluation: a 2 m by 25 m strip plot adapted from Fletcher (1986) and a 3.99 m radius circular plot commonly used in B.C. Ministry of Forests silviculture surveys. Both plot types had areas of 50 m2. A 4 ha area in a Date Creek plantation (Figure 1), known tobe infested by Warren’s collar weevil (Garbutt 1988), was selected for comparison testing of the 2 plot types. Alternate transects of strip and circular plots were oriented in an East - West direction. Circular plot centres were placed at 25 m intervals while strip plots were oriented sequentially along a transect, with 25 m between transects. The percentage of sample trees with larval damage was determined for each plot type. Dominant and co-dominant lodgepole pine trees with either old or current weevil scarring were considered attacked. The percentage of sample trees attacked was pooled for all plots within each plot type and the two plot types were compared using a Z statistic (Zar 1984). 2.2.2 Sampling intensity and maximum survey area Logistical constraints necessitated a maximum sampling intensity of 1% (2 plots/ha) for surveyed plantations. This was to ensure that the sampling of the desired number of plantations 13 Figure 1. Schematic representation of the layout of two plot types tested for assessing the incidence ofHylobius warreni damage in the Kispiox Forest District, Hazelton, B.C., Summer 1989. I. Scale: 1:10000 Circular plots (3.99rn radius) I—I Strip plots (2rn X 25m) 14 was completed within the study timeframe (16 weeks). The maximum area surveyed within any plantation was 50 ha. For plantations greater than 50 ha, an arbitrary 50 ha area was delineated for sampling. Approximately eight weeks into the project it became apparent that due to time, and budget constraints, the minimum number of plantations could not be surveyed using a sampling intensity of 1 %. Therefore, the sampling intensity was reduced to 0.5% of the plantation area (1 plot/ha) for the final 15 plantations. 2.2.3 Plantation Selection Thirty plantations were selected randomly from a pooi of plantations meeting the following criteria: they had to be predominantly pine (> 50% of total stems), they had to be a minimum of 5 years old, and they had to meet the B.C. Ministry of Forests minimum stocking standards (a minimum of 700 well spaced stems per hectare). 2.2.4 Plot information The following information, was collected from each plot: i). The total # of trees within the plot, classified by species and age class, where 2 broad age classes were established: Layer 1 and layer 2 trees. Layer 1 trees included all species that had seeded in naturally since the time of planting. Layer 2 trees were planted lodgepole pine trees. Layer 2 trees were characterized by the average height and dbh values given for each plantation ii) The # of dead lodgepole pine (P1) and spruce (Sx) and the cause of death from visible symptoms. iii) The # of dying or chiorotic P1 and the cause of stress from visible symptoms. iv) The # of green healthy looking pine with root collar weevil damage. v) The # of well-spaced trees per plot based on stem form and spacing (an arbitrary minimum 2 rn distance from competing conifers). The maximum # of well-spaced trees/plot was 6. vi) The # of weevilled (evidence of old or new larval feeding damage) potential well 15 spaced stems (i.e., meet all requirements of v) and have evidence of weevil damage). vii) The average LFH depth (organic layer above mineral soil), calculated by taking 4 depth samples every 4th plot. The plot was divided into 4 equal quadrants and a measurement was taken from the centre of each. Sample depths were pooled and the average calculated for each plantation. LFH, as defined in this study, included all organic matter above mineral soil including the moss layer, if present. Trees were recorded as weevil damaged if they had evidence of new or old attacks. New attacks were distinguished from old attacks by the presence of fresh pitch exudation and/or the presence of larvae. The average percentage of the stem circumference girdled by weevil larvae and the average number of larvae present per tree for each plantation were also determined for those plantations surveyed at the 0.5% intensity. This was done by arbitrarily selecting the first tree per plot with larval feeding damage and excavating the root collar and larger lateral roots. The percentage of the stem circumference girdled was estimated to the nearest 10%. The total number of larvae per tree was also recorded. 16 2.2.5 Plot Summaries Each plantation was summarized in terms of the following attributes: I) Total trees/ha ii) Total well-spaced trees/ha iii) Total trees/ha with weevil damage iv) Total well-spaced trees/ha3with weevil damage v) % of weevil caused mortality to layer 2 pine vi) Total # of chiorotic and dead trees due to agents other than weevil vii) Average tree height viii) Average diameter at breast height (dbh), where trees were at least 3 m in height. Average tree height and dbh values for each plantation were determined by measuring a minimum of 30 dominant / co-dominant lodgepole pine trees per plantation. The age of the dominant/co-dominant trees was assumed to be the same as the plantation age. This was confirmed by sampling a few trees per plantation with an increment borer and also by conducting whorl counts. 2.2.6 Within Plantation Distribution A 1:10 000 or 1:20 000 scale plot map was produced for each plantation showing the distribution of plots with and without weevilled trees (Figure 2). This was done to determine if there was any apparent stratification of weevil damage within a plantation. This was then 3Well-spaced and weevilled well-spaced trees are not necessarily additive. A smaller non-weevilled tree would be tallied as a well-spaced tree even if it was within 2 m of a taller weevilled tree. The weevilled tree would then be tallied as a weevilled well-spaced tree. In the absence of weevil, only one well-spaced tree would have been tallied. 17 Kispiox Forest District 1989 Warren’s collar weevil survey summary Opening #: 93M032-005 Location: N. of Salmon River Date surveyed: May 1989 Scale: 1: 10000 Figure 2. Summary plot map for a plantation surveyed for Hylobius warreni damage in the Kispiox Forest District, Hazelton, B.C., Summer 1989. ‘I Total area (ha): 58 Area surveyed (ha): 58 $77 History symbol: N81 N I, I, II I, ‘I I’ I’ I’ LEGEND Po.C. A Point of commencement • Plots with weeviled trees o Plots without weeviled trees I’ I’ 18 compared with field information (i.e. plot notes) to determine the pattern of weevil attack within the sampled area. 2.2.7 Historical Information The B.C. Ministry of Forests history record system provided historical background information on surveyed plantations. This included information on original stand type, site preparation method, plantation establishment and, in some cases, ecosystem classification. This information was then used to determine if site history is usefhl in predicting weevil abundance within plantations. 2.2.8 Survey Summary After data compilation was completed for all plantations, simple linear regression (Zar 1984) was used to determine the relationship between percentage of sample pine attacked, and plantation age, tree height, tree density and LFH depth. Regression analysis was done using MIDAS (Fox and Guire 1976) on the University of British Columbia computing services network. 2.3 Height Growth and Dispersal Study 2.3.1 Study Locations The study was replicated at three locations within the Kispiox Forest District: Date Creek, Mosquito Flats, and Shandilla Creek (Figure 3). These locations had been surveyed in 1989 as part of the distribution study and were known to support weevil infestations. All plantations were surveyed from May through August 1990. All plantations were within the ICHmc3 biogeoclimatic variant (Meidinger and Pojar 1991). 19 _KISPIOX FOREST DISTRICT Figure 3. Hylobius warreni study sites in the Kispiox Forest District, Summer 1990 N 20 2.3.1.1 Date Creek This opening was clearcut logged in 1975 and was spot burned and planted in 1976. The original stand type was spruce-hemlock. The elevation is 457 m, the average slope is 20%, and it has a southerly aspect. Stand density was estimated at 5455 sph in 1989 (first year of study), however a large portion of this density was relatively small natural seedlings which had ingressed since planting. Lodgepole pine is the major species, while spruce is a minor component on wetter microsites. Natural regeneration included western redcedar, western hemlock and true fir. The 1989 survey found 45% of sampled trees had evidence of weevil feeding. 2.3.1.2 Mosquito Flats This opening was clearcut logged and broadcast burned in 1972, and planted in 1973. The original stand type was hemlock-pine. The plantation was manually weeded in 1985 and fertilized in the fall of 1989. The elevation is 487 m, the average slope is 10-20%, and the aspect west. Stand density was estimated at 1054 sph in 1989 (first year of study). Lodgepole pine is the major species, with spruce making up a minor component. Natural regeneration included western hemlock, western redcedar and some true fir. The 1989 survey found 84% of sampled trees had evidence of weevil feeding. 2.3.1.3 Shandilla Creek This opening was clearcut logged in 1968, burned in 1969, 1970 and 1972, and planted in 1973. The original stand type was hemlock-spruce. This plantation was fertilized in the fall of 1989. The elevation is 381 m, the average slope is 50%, and the aspect north. Stand density was estimated at 3095 sph in 1989 (first year of study). Lodgepole pine is the major species, with spruce making up a minor component. Natural regeneration includes western hemlock and western redcedar. The 1989 survey found 55% of sampled trees had evidence of weevil feeding. 2.3.2 Sampling Methodology A single rectangular plot 20 m wide and 150 m long (200 m long in the Date Creek plantation) extended perpendicularly from the stand margin into the plantation. The point of commencement for the plot was chosen for ease of location and proximity to mature timber 21 (preferably pine). Plot boundaries were located using a compass and 50 m nylon chain, and were marked with flagging tape. The four corners of the plot were marked with wooden stakes and the P.O.C. was marked with a metal tag. Within each plot all pine and spruce trees were permanently numbered and tagged and sampled for the following parameters: i) Organic layer depth (LFH) at the root collar of all sampled trees. Two measurements were taken on opposite sides of the tree and LFH depth was recorded as the mean of these two values. Measurement location was arbitrary. Notes were also recorded on the composition of the LFH layer at the tree base and within the plantation, in general. ii) Root collar diameter. iii) Number and age of weevil life stages present. iv) Percentage of root collar circumference girdled for attacked trees. Girdles were classed as new or old: new attacks had fresh resin and larvae were present, while old attacks were characterized by hard, whitish resin deposits. vi) Presence of other damaging organisms. 2.3.3 Height Growth Study 2.3.3.1 Tree Selection After all trees were sampled and damage assessments were completed, the trees were grouped into four girdling classes according to the amount of the root collar circumference girdled: 0%, 1-50%, 51-80% and 81-99% of the root collar girdled. The categories for the Date Creek plantation were 0%, 1-50%, 5 1-70%, and 71-99%. The girdling classes were changed for the former two plantations in mid-project because it was felt that growth losses may become more evident after 80% of the root collar has been girdled (Cerezke 1974). The five largest diameter trees in each class were selected for stem analysis. 22 2.3.3.2 Stem Analysis Each of the selected trees was felled and the internodal growth over the life of the tree was measured. Wooden discs were removed from the root collar and the years of weevil attack were assessed using a starch staining method developed by Cerezke (1972). When injury to the cambium of the tree occurs, the tree responds by producing traumatic resin ducts. These traumatic ducts show up as tangential bands in the annual rings in the years in which the damage occurs, and can be used to determine the time of the injury. Using this method and the presence of old weevil feeding scars, it was possible to determine when the tree was attacked. 2.3.3.3 Statistical Analysis Analysis of variance (one-way) was used to compare girdling classes for differences in root collar diameter, breast height diameter, and annual height increment. A Student-Newman Keuls test was used to test for significant differences among means for these parameters (Zar 1984). Annual height increments from 1979 to 1989 were tested for significant differences between girdling classes for the Mosquito Flats and Shandilla plantations. Height increments between 1980 and 1989 were tested for significant differences between classes in the Date Creek plantation. Pre-1979 height increments for Mosquito Flats and Shandilla, and pre-1980 height increments for Date Creek were excluded from analysis to ensure that growth was relatively linear from year to year. Reliability of height measurements below breast height was somewhat suspect. 23 3.0 RESULTS 3.1 Distribution and Abundance of Warren’s Collar Weevil 3.1.1 Selection of plot type There was no significant difference in the percentage of sampled trees with weevil damage between the 2 plot types (Z=0.176; p > 0.05). In strip plots, 46.3% of sampled stems had evidence of weevil feeding. In circular plots, 45.2% of sampled stems had evidence ofweevil feeding. 3.1.2 Plantation distribution Surveyed plantations were distributed throughout the Kispiox Forest District (Figure 4). Plantations ranged in age from 5 to 16 years (Table 1, Figure 5). The average age of surveyed plantations was 9 years. Plantation size ranged from 4 ha to 107 ha. The average plantation size was 58 ha. The average area surveyed was 40 ha. 3.1.3 Plantation summaries The average % of sampled pine with Warren’s collar weevil larval damage was 29 % (S.D.=26%; Range:0-100%) (Table 1). The average % of sampled spruce with larval damage was 1.4% (S.D.= 3.5%; Range: 0-15%). Weevil caused mortality was low in most plantations, averaging 1.5% (S.D.1.9%; Range:0-8.8%). The average percentage of the circumference of the root collar girdled of sampled trees for the final 15 plantations was 50% (S.D.=32%, n=107) and the average number of larvae per girdled tree was 0.33 (S.D.=0.49, n107). In most cases there were sufficient unattacked well spaced trees to maintain plantations at minimum stocking (20 of 31 plantations> 700 stemslha)(Table 1). Four of the eleven plantations which were below minimum stocking had less than 600 stems/ha. 24 FigUre 4. The distribution of pine plantations surveyed for Hylobius warreni damage in the Kispiox Forest District, Hazelton, B.C., Summer 1989 KISPIOX FOREST DISTRICT N A Number and location of surveyed planLations 25 Ta bl e 1. Su m m ar y da ta fo r p la nt at io ns su rv ey ed fo rH yl ob iu sw ar re n i i n th e K isp io x Fo re st D ist ric t, H az el to n B .C ., Su m m er 19 89 . Lo ca tio n O pe ni ng Pl an ta tio n Si te Ec os ys te m Or ig in al El ev ati on Sl op e A sp ec t So il LF H A re a Sa m pl in g % 1 % W ell W ee vi lle d 2 To tal Tr ee DB H # Ag e Pr ep ar at io n Cl as sif ic at io n St an d (m ) (96 ) Te xt ur e (cm ) (ha ) In te ns ity W ee vi lle d W ee vi l Sp ac ed W ell Tr ee s! H ei gh t (cm ) Ty pe (% ) M or tal ity St em s! ha Sp ac ed lh a ha (m ) Sh an di lla 93 M 00 1- 00 9 W .o fS ke en ax 93 M 00 1- 01 3 N as h V 93 M 01 1- 04 1 Su sk w a R. 93 M 02 4- 00 7 Su sk w a R. 93 M 02 4- 01 8 M os qu ito Fl at 93 M 02 4- 02 5 Ro bi ns on Lk . 93 M 03 3- 00 5 N as h V 93 M 01 1- 04 7 Se el y Lk . 93 M 02 2- 01 9 Su slc wa R. 93 M 02 4- 00 5 N .S al m on R. 93 M 03 2- 00 5 U tsu n Cr . 93 M 04 2- 00 7 D at e Cr . 93 M 04 2- 02 6 N. Kl in e Lk 93 M 05 1- 00 2 M ur de rC r. 93 M 05 1- 00 6 LO T 30 20 93 M 05 1- 01 4 Cu llo n Cr . 93 M 06 1- 01 6 Cu llo n Cr . 93 M 06 1 - 03 0 Iro ns id e Cr . 10 3P 06 0- 00 2 Iro ns id e Cr . 10 3P 07 0- 01 0 Iro ns id e Cr . 10 3P 07 0- 02 3 St en itt Cr . 93 M 05 2- 02 3 Sa lm on R. 93 M 03 2- 01 I Sa lm on R. 93 M 03 2- 01 4 N at lan Cr . 93 M 03 4- 00 5 5m .N as hY 93 M 01 1- 04 5 Lu no Cr . 93 M 01 4- 06 2 Sw an Rd . 93 M 03 2- 00 7 St er rit tC r. 93 M 05 2- 01 0 7 Si ste rs 10 3P 00 9- 01 1 N an ge es e R. 10 3P 07 9- 00 5 B ro ad ca st bu rn IC Hg 3.0 11 09 B ro ad ca st bu rn lC Hg 2.0 1 B ro ad ca st bu rn IC Hg 3. 01 !.0 9 B ro ad ca st bu rn IC Hg 2- 3 B ro ad ca st bu m lC Hg 3.0 1 B ro ad ca st bu m IC Hg 3.0 1 B ro ad ca st bu rn IC Hg 3. 03 a 38 1 50 N 45 7 20 N 54 9 10 SE 61 0 15 S 45 7 20 N 48 7 10 W 88 4 20 N 54 9 10 SE 45 0 20 5 76 2 10 S 30 5 0 F 53 4 10 V 45 7 20 S 39 0 0 F 58 0 15 SW 54 9 30 W 50 0 10 F 45 0 20 S 45 0 10 SW 45 0 15 SE 42 5 20 V 45 0 15 SW 30 0 0 F 30 0 15 NW 76 2 20 5 45 0 10 SW 40 0 20 W 27 5 15 V 45 0 5 F 46 0 15 N 55 0 0 F SL 7 50 - 7 50 SL 4 30 - 15 50 - 12 46 - 6 38 - 11 14 SL 8 33 - 5 50 Si L 8 43 Si L 7 58 SL IL S 7 50 - 5 4 - 8 50 - 7 42 SL 4 50 CL 7 50 IJC L 7 30 L 5 17 Si CL 7 50 SL 4 50 L 7 36 LS 5 28 Si CL 8 47 C 13 50 - 6 50 CI IS L 7 22 SL -L 6 4 CI LS 8 49 SL 8 39 LS 6 14 1 55 1.1 60 8 1 34 0. 5 61 4 1 75 1. 0 24 0 1 27 0. 3 72 6 1 71 1.1 42 8 1 84 1. 3 17 7 1 43 0 60 0 1 23 1. 2 10 30 0. 5 35 0.1 44 0 1 34 2. 2 82 5 1 42 1. 4 36 1 1 6 6. 3 10 24 8 45 2. 4 71 0 0. 5 19 0 71 2 0. 5 12 1. 6 81 2 0. 5 23 0. 3 64 8 0. 5 6 0. 7 68 8 0. 5 5 0 90 0 0. 5 lO O 2 82 9 0. 5 8 3. 6 89 6 0. 5 6 0 10 12 0. 5 4 1. 4 97 1 0. 5 22 8. 8 85 6 1 34 6. 0 76 3 1 3 0 88 8 1 12 0. 3 90 2 1 17 4. 0 82 3 0. 5 31 0 80 0 0. 5 3 1 77 8 0. 5 0 0 85 6 0. 5 8 0 81 4 49 2 30 95 7. 04 10 .1 21 4 28 94 6. 46 8.1 54 0 28 60 7. 32 10 .5 20 2 17 50 6. 37 9. 3 40 5 29 59 7. 45 9. 5 56 4 10 54 5. 07 10 .3 25 0 18 84 6. 89 10 .1 27 2 65 20 3. 60 4. 7 46 4 69 44 4. 09 4. 5 13 8 26 66 4. 16 5. 7 35 7 27 30 4. 08 4. 3 28 91 32 3.3 1 4. 4 54 55 - - 12 19 20 2. 57 - 76 34 68 2. 47 - 31 6 62 08 3. 86 5. 4 24 15 00 2. 08 17 25 50 2. 14 - 14 18 56 1. 64 - 28 26 88 1. 86 - 36 34 84 2.0 1 5 50 60 1. 97 - 0 24 71 2. 16 - 16 7 46 37 1. 67 - 2 29 84 0. 95 - 18 0 37 40 45 21 73 1. 74 - 40 0 28 50 1. 75 5 14 54 2. 00 - 0 39 59 1. 59 - 14 3 17 00 1. 85 - a. Ex pr es se d a s 96 of w ee v ill ed la ye r 2 pi ne (c ha ra ct er iz ed by av er ag e he ig ht an d db h v al ue s). b. M ee ta ll re qu ire m en ts of a w all sp ac ed tr ee w ith th e ex ce pt io n of w ee v il da m ag e (i.e .g oo d fo rm an d sp ac in g). c. n = 3 tr ee s. 26 HS x HC w PI H Cw B HC w HP I HB 16 16 16 13 16 16 16 10 13 10 8 7 13 7 6 12 6 6 5 6 6 5 5 5 5 9 5 5 5 6 6 Sp ot bu rn Sp ot bu rn Sp ot bu m Sp ot bu m Sp ot bu m Sp ot bu m Sp ot bu m Sp ot bu m Sp ot bu m Sp ot bu m Sp ot bu m Sp ot bu m Sp ot bu m Sp ot bu m Sc ar ifi ed Sc ar ifi ed Sc ar ifi ed Sc ar ifi ed Ni l Ni l Ni l Ni l Ni l Ni l IC Hg 3.0 1 PI Sx IC Hg 3 PI Sx lC Hg 3. 01 PI H IC Hg 3. 08 P1 IC Hg 3.0 1 HC w IC Hg 3.0 1 Sx Hw IC Hg 3.0 11 02 !0 3 Sx PI IC Hg 3. Ol a HB IC Hg 3.0 1 PI At IC Hg 3.0 1 HP I IC Hg 2.0 1 HS x IC Hg 3. 09 b Sx Pl lC Hg 3. 01 10 3 Sx B IC Hg 3.0 11 03 PI Sx lC Hg 3. 01 10 3 H lC Hg 3. 01 10 9 Sx PI IC Hg 3. 09 PI Sx IC Hg 2. 04 HB lC Hg S. 01 10 2 PI Sx IC Hg 3. 01 l0 3 SP I IC Hg 3. 01 !0 3 PI At IC Hg 3.0 1 Sx Cw IC Hg 2.0 11 04 HS x IC Hg 3.0 1 SB 10 9 8 7 6 z 4 3 2 0 Figure 5. Age distribution of pine plantations surveyed for Hylobius warreni damage, Kispiox Forest District, Hazelton, B.C., Summer 1989. 5 6 7 8 9 10 11 12 13 14 15 16 PLANTATION AGE 27 3.1.4 Weevil incidence in relation to site and stand factors The incidence of root collar weevil damaged trees was greater in older plantations than younger plantations (Table 1, Figure 6). The average percent of trees attacked in plantations older than 10 years was 48%, but was only 14% in plantations younger than 10 years. The percentage of trees damaged by Warren’s collar weevil also increased with increasing average tree height (Figure 7). Average tree height increased with increasing plantation age (Tree height = 0.45 xPlantation age -0.51, R2 = 0.91, SE = 0.64, p< 0.05). There was no apparent relationship between the percentage of trees with weevil damage and the total number of stems/ha (sph) (Figure 8). Surveyed plantations ranged from 1054 sph to 9132 sph. There was not a good relationship between average LFH layer depth and the percentage of pine with root collar weevil damage (Figure 9). The LFH layer depth varied from 3-15 cm. Burning or scarification did not reduce the abundance of weevils (Table 1). The average percentage of trees with larval damage was 55.7% on broadcast burn blocks, 26% on spotburn blocks, 15.8% on scarified blocks and 13% on blocks with no site preparation treatment. The average age of plantations with each treatment type was 16 years for broadcast burned blocks, 8 years for spotburn blocks, 6 years for no treatment and 5 years for scarified blocks. Two Salmon River plantations which were scarified had high levels of weevil damage (Table 1; 93M032-01 1 and 93M032-014). 28 100% 90% . 80% . 70% % OF STEMS ATTACKED = -12.03 + 4.20 x PLANTATION AGE 60% R2 = 0.64, SE = 13.37, p<O.OOl 50% .z 40% • . • . 30% ‘F 0 o .z • .z A . 10% / . S.$ 0% I I I I I 0 2 4 6 8 10 12 14 16 18 20 PLANTATION AGE Figure 6. Percentage of sampled Lodgepole pine with Hylobius warreni damage in relation to plantation age in the Kispiox Forest District, Hazelton, B.C., Summer 1989. 29 100% 90% . 80% . 70% % OF STEMS ATTACKED = - 4.27 + 8.69 x AVERAGE TREE HEIGHT R2 = 0.61, SE = 14.5, p <0.001 uv,,o 50°/ . . 40% © .30% . . .. 20% . . . 10% . II • . S 0% I •i I I 0 1 2 3 4 5 6 7 8 9 10 AVERAGE TREE HEIGHT (m) Figure 7. Percentage of sampled Lodgepole pine with Hylobius warreni damage in relation to average tree height in the Kispiox Forest District, Hazelton, B.C., Summer 1989. 30 100% % OF STEMS ATTACKED = 32.75 - 1.83 x 10 x SPH 90% R2 = 0.02, SE = 22.64, p > 0.05 . 80% . 70% 60% . 50% . . 40% .. . 30% . .. 20% . . .. 10% . . . • • • 0% I 0 2000 4000 6000 8000 10000 STEMS PER HECTARE Figure 8. Percentage of sampled Lodgepole pine with Hylobius warreni damage in relation to plantation density in the Kispiox Forest District, Hazelton, B.C., Summer 1989. 31 100% % OF STEMS ATTACKED = 29.12 -0.37 x AVG. LFH DEPTH 90% R2 = 0.002, SE = 22.88, p > 0.05 80% . o .70/o ‘no,v’J/o . 50% . . 40% o . . 30% 20° 0 . . . . 10% . • a . . 0% I I I 0 2 4 6 8 10 12 14 16 18 20 AVERAGE LFH DEPTH (cm) Figure 9. Percentage of sampled Lodgepole pine with Hylobius warreni damage in relation to average organic matter layer depth in the Kispiox Forest District, Hazelton, B.C., Summer 1989. 32 Eighty-seven percent of sampled plantations were within the ICHg3 subzone. Of these 87%, 85% were within the ICHg3.01 ecosystem association. One hundred percent of plantations surveyed were within the ICHmc3 biogeoclimatic variant. There was a wide range ofweevil incidence within all sampled plantations irrespective of ecosystem association. In general, weevil damage was prevalent on circum-mesic well drained sites. Within the Kispiox District this includes the ICHg3. 01/03/08/and 09 ecosystem associations. Under the new classification system susceptible site series would include the ICHmc3/0 1/04/05. Plantations on sites which had a pine component prior to harvest, or were adjacent to standing mature pine, appeared to be most susceptible to weevil damage (Table 1). The average percentage of stems with weevil attacks in plantations established on sites which previously had a P1 component in the inventory was 31%; the average for plantations on sites with P1 as a minor component (or no component) in the previous stand was 20%. However, some plantations which had minor components of P1 prior to harvest had relatively high rates of weevil damage (Table 1). There was no indication that elevation, slope, or aspect had an impact on weevil abundance. The elevational range of the plantations varied from 300 m to 884 m (average:487 m; S.D. :137 m) (Table 1). The slopes on surveyed plantations ranged from 0-50% (average:15%; S.D. :9.9%). 3.1.5 Within plantation distribution. Weevil attack pattern was not stratified within a plantation. Weevil damaged trees were either randomly distributed within sampled plantations or the distribution was clumped in no particular pattern. 33 3.1.6 Within district distribution. Root collar weevil was found in all surveyed plantations within the Kispiox Forest District (Table 1). The Seven Sisters plantation (103P009-O1 1) had a 0 % infestation rate based on our survey data; however, weevil damaged trees were found outside of the plot boundaries. High weevil attack rates were found in all older pine plantations within the district. Plantations with particularly high levels of infestation were found in the Suslcwa River, Nash Y and Shandilla areas of the district. Salmon River plantations, which were considerably younger, were also heavily infested. The upper Kispiox drainage (Sterritt Cr., Ironside Cr., Cullen Cr.) had relatively low levels of weevil damage. 3.2 Height Growth and Dispersal Study 3.2.1 Plot summaries 3.2.1.1 Date Creek Seventy percent of all sampled trees had evidence of weevil damage (Table 2). Eighty-one percent of all sampled lodgepole pine had evidence of attack, while 37% of all sampled spruce had evidence of attack. All dead pine appeared to have died as a result of weevil feeding as indicated by root collars which were completely girdled. The total mortality attributed to Warren’s collar weevil was only 3% of sampled trees (Table 2). The average diameter of attacked pine was 7.9 cm (Table 3). The average number of larvae per tree was 1.2. The number of weevils per hectare, based on the average number of weevils per tree and the estimated stems per hectare, was 850. The percentage of trees attacked within a diameter class increased with increasing diameter class (Figure 10). There was a significant relationship between root collar diameter and percentage of stem girdled (Figure 11). The regression of percent of stem girdled on LFH depth was also significant (Figure 12). 34 Table 2. Number and percentage of sample trees attacked by Hylobius warreni, by host species, at Date Creek in the Kispiox Forest District, Summer 1990. Lodgepole pine Spruce Attacked Unattacked Dead Attacked Unattacked Total Trees 265 66 9 18 49 Stems per hectare 663 165 23 45 123 % of Species Total 78 19 3 37 63 Table 3. Population estimates ofHylobius warreni and characteristics of attacked lodgepole pine trees at Date Creek in the Kispiox Forest District, Summer 1990. Weevil Numbers per Tree RCDa _______________________________ LFII (cm) % GIRDLED 1ST YEARC 2ND YEARd 3RD yEARe TOTAL (cm) AVERAGE 7.9 34 0.5 0.3 0.5 1.2 7 S.D. 3.9 33 1.0 0.6 1.0 0.9 4 a. Root Collar Diameter b. % ofroot collar circumference girdled c. Larvae hatched from the current years eggs (1990) d. Larvae hatched from 1989 eggs. e. Larvae and pupae from 1988 eggs f. Average depth oforganic layer above mineral soil. 35 70 60 c,) 50 40 30 0 Z20 10 0 30 ROOT COLLAR DIAMETER (cm) Figure 10. Diameter distribution of sampled Lodgepole pine trees and proportion attacked by Hylobius warreni at Date Creek in the Kispiox Forest District, Summer 1990. 0 5 10 15 20 25 36 100% 80% 60% 40% 20% 0% - —— — — — — — I — I I I 0 2 4 6 8 10 12 14 16 18 20 ROOT COLLAR DIAMETER (cm) Figure 11. Percentage of sampled Lodgepole pine with Hylobius warreni damage in relation to root collar diameter at Date Creek in the Kispiox Forest District, Summer 1990. (% of stem girdled =11.00 + 3.36 x RCD, R2 = 0.14, SE 30.41, p <0.05). .. .. • • ••. . . • . •. • •• • . ••• •. •• . . • • •• • • . •. • •.• • lI•• .. i%i• . * •• lb.. • . I • . .. •• • • • . . S . • • ••l • •. .. * n330 •• ••••.• * . 37 z0% 0 2 4 6 8 10 12 14 16 18 20 LFH DEPTH (cm) . ... . • : •. •I• • •.• . . . . . . . . . . . . . • • •• I I . • •I I •• S •• . I . •I 100% 80% 60% 40% 20% •. I . I •. •1a •1. •I • .% .. SI I. •I . . . •i• I I • • 1.11 . —I . . I . . . I I . . . •• •• • • S . I . • IS. II • • • • : .% .: II . . . . I . •. I . I I • I n313 I 0• Figure 12. Percentage of sampled Lodgepole pine with Hylobius warreni damage in relation to average organic matter layer depth at Date Creek in the Kispiox Forest District, Summer 1990. (% of stem girdled = 18.67 + 3.06 x AVG LFH depth, R2 = 0.10, SE = 31.06, p <0.05) 38 3.2.1.2 Mosquito Fiats Eighty-four percent of all sampled trees had evidence of weevil feeding (Table 4). Eighty- eight percent of all sampled pine had evidence of weevil damage, while 11% of sampled spruce had been damaged by root collar weevil feeding. Sixty percent of the dead pine were killed by weevil girdling and 40% were killed by stem infections of western gall rust (Endocronarlium harknessii (J.P.Moore) Y.Hirat). However, total mortality within the plot was only 2% of sampled stems (Table 4). The average diameter of attacked lodgepole pine was 14.1 cm (Table 5). The average number of weevil larvae per tree was 1.4. The number of weevils per hectare, based on the average number of weevils per tree and the estimated stems per hectare, was 1100. The percentage of trees attacked within a diameter class increased with increasing diameter class (Figure 13). There was a significant regressions between percentage of the root collar girdled and both root collar diameter (Figure 14) and LFH depth (Figure 15). The Mosquito Flats plantation had a small outbreak of a sawfly (Neodiprion nanulus contortae Ross) (Rod Garbutt Pers. Comm.)4during the summer of 1990; however, damage was minor. 4Forest Insect and Disease Ranger, Prince Rupert Forest Region 39 Table 4. Number and percentage of sample trees attacked by Hylobius warreni, by host species, at Mosquito Flats in the Kispiox Forest District, Summer 1990. Lodgepole pine Spruce Attacked Unattacked Dead Attacked Unattacked Total Trees 235 34 5 1 9 Stems per hectare 783 113 17 3 30 % of Species Total 86 12 2 11 89 Table 5. Population estimates ofHylobius warreni and characteristics of attacked lodgepole pine trees at Mosquito Flats in the Kispiox Forest District, Summer 1990. Weevil Numbers per Tree RCD’ ______________________________ LFH (cm) % GIRDLEb ISTYEARC 2NDYEARd 3RDyEARe TOTAL (cm) AVERAGE 14.1 60 0.2 0.5 0.7 1.4 7 S.D. 4.2 30 0.5 1.0 1.1 0.9 4 a. Root Collar Diameter b. % of root collar circumference girdled c. Larvae hatched from the current years eggs (1990) d. Larvae hatched from 1989 eggs. e. Larvae and pupae from 1988 eggs f. Average depth oforganic layer above mineral soil. 40 70 60 50 40 j) 20 10 ROOT COLLAR DIAMETER (cm) Figure 13. Diameter distribution of sampled Lodgepole pine trees and proportion attacked by Hylobius warreni at Mosquito Flats in the Kispiox Forest District, Summer 1990. 0 0 5 10 15 20 25 30 41 100% 80% c-) z 60% 40/o j) C 20% 0% 0 5 10 15 20 25 30 ROOT COLLAR DIAMETER (cm) Figure 14. Percentage of sampled Lodgepole pine with Hylobius warreni damage in relation to root collar diameter at Mosquito Flats in the Kispiox Forest District, Summer 1990. (% of stem girdled = 21.10 + 2.27 x RCD, R2= 0.12, SE =30.11, p <0.05) . . *,ii• ••••••• . . . . . .. .. S •. • •.• •.• • • • •. •• • •• I • • • • • . $ • • • •• •• • • • • S. • • . . . . . • . .. . %..... . I..• • . I • •. I •I •. • _#•4h • • •I • : • • I . •I I • I n=256 I I 42 100% •.. . ••.•••• .... . . Figure 15. Percentage of sampled Lodgepole pine with Hylobius warreni damage in relation to average organic matter layer depth at Mosquito Flats in the Kispiox Forest District, Summer 1990. (% of stem girdled = 38.60 + 1.91 x LFH depth, R2 = 0.05, SE = 31.23, p < 0.05). . .. . .. . $ . . . w.. • .. • •. • •. S. •• •. •• • . • . . . . . . . • . • I . 80% 60% 40% 20% • 0% • . .. . . . I. . a. . . • •. . • . n=256 • 0 5 10 15 20 25 30 LFH DEPTH (cm) 43 3.2.1.3 Shandilla Ninety-two percent of all sampled trees had evidence of weevil feeding (Table 6). Ninety- one percent of all sampled pine had evidence of weevil feeding, while 25 % of sampled spruce were attacked. Weevil caused mortality was 0.6 %. The average diameter of attacked pine trees was 13.8 cm (Table 7). The average number of weevil larvae per tree was 2.6. The number ofweevils per hectare, based on the average number ofweevils per tree and the estimated stems per hectare, was 3680. The percentage of trees attacked within a girdling class increased with increasing diameter class (Figure 16). There was a significant regression between the percentage of the stem girdled and both root collar diameter (Figure 17) and LFH layer depth (Figure 18). 3.2.2 Pattern of Attack Destructively sampled trees had been attacked within the last 5-6 years (Tables 8-10). The earliest attacks on trees in Date Creek and Mosquito Flats occurred in 1984, while the earliest attacks in Shandilla occurred in 1987. The majority of attacked trees had both old and new attacks. The time of weevil attack was uniform throughout the plots (Tables 8-10). Trees at the periphery of the plot were attacked during the same years as those within the interior of the plot. The proportion of trees attacked within a basal area class increased with increasing basal area for all three surveyed plantations (Figure 19). 3.2.3 Stem Analysis 3.2.3.1 Date Creek Root collar diameter (RCD) and diameter at breast height (DBH) were significantly greater in the attacked girdling classes than in the unattacked girdling class (Table 11). Average tree height in all three damaged girdling classes was greater than the average total height of unattacked trees (Figure 20). Annual height increment in the 1-50 % girdling class during 1989 44 Table 6. Number and percentage of sample trees attacked by Hylobius warreni, by host species, at Shandilla in the Kispiox Forest District, Summer 1990. Lodgepole pine Spruce Attacked Unattacked Dead Attacked Unattacked Total Trees 424 35 3 1 3 Stems per hectare 1413 117 10 3 10 % of Species Total 92 7.5 0.5 33 67 Table 7. Population estimates ofHylobius warreni and characteristics of attacked lodgepole pine trees at Shandilla in the Kispiox Forest District, Summer 1990. Weevil Numbers per Tree RCD ______________________________ LFI1 (cm) % GIRDLEb 1ST YEARC 2ND YEARd 3) TOTAL (cm) AVERAGE 13.8 60 0.2 0.6 1.8 2.6 9 S.D. 3.2 30 0.5 0.8 2.7 1.3 4 a. Root Collar Diameter b. % ofroot collar circumference girdled c. Larvae hatched from the cuerent years eggs (1990) d. Larvae hatched from 1989 eggs. e. Larvae and pupae from 1988 eggs f. Average depth of organic layer above mineral soil. 45 160 990/ C HEALTHY TREES140 ATI’ACKED TREES 100% 120 100 95% 80 0 60 40 z 21% 83% 100% r 100% 20 0% 100% 0 0510 15 2025 30 ROOT COLLAR DIAMETER (cm) Figure 16. Diameter distribution of sampled Lodgepole pine trees and proportion attacked by Hylobius warreni at Shandilla in the Kispiox Forest District, Summer 1990. 46 100% . . . — • •. 80% z 60% 40% 20% 0% - 0 5 10 15 20 25 ROOT COLLAR DIAMETER (cm) Figure 17. Percentage of sampled Lodgepole pine with Hylobius warreni damage in relation to root collar diameter at Shandilla in the Kispiox Forest District, Summer 1990. (% of stem girdled = 13.44 + 2.99 x RCD, R2 = 0.15, SE = 28.76, p <0.05). . . . S. •. . . . . . . . . . .% . . . . • . •, •. S . . • .• I I •. I I I I n=459 I I I • I •I • I .“v. •• •. • I .•,I.. I • I • I. I • I I I a aa a 47 100% 80/a z 60% 10j) 40/o 0 20% 0% 0 5 10 15 20 25 LFH DEPTH (cm) Figure 18. Percentage of sampled Lodgepole pine with Hylobius warreni damage in relation to average organic matter layer depth at Shandilla in the Kispiox Forest District, Summer 1990. (% of stem girdled = 24.71 + 3.32 x LFH, R2= 0.15, SE = 28.83, p <0.05). . .* ... .. •. •• *.:: .• •. •• •• •.. •• • • .: •••• :.‘;;: • • • ?: •;: •.•• •. •..: .• • • • 0• / • • • • • .•,• •. • •I •I • • • a .:1. • •. I • •. - .0 • • . • . • •.I . •0 • *11 • • •. •• • •. ••.• : • .1. • n=459 • • . -9-- • . 48 Table 8. Timing of attack ofHylobius warreni on sample trees in a 13 year old lodgepole pine plantation at Date Creek in the Kispiox Forest District, Summer 1990. Tree # % Girdleda RCDb(cm) Year(s) of Attack 44 0 11.5 Nil 206 0 10.2 Nil 239 0 9.3 Nil 318 0 9.5 Nil 341 0 10.3 Nil 376 16 18.0 1987-1989 96 20 16.3 1986-1989 287 20 15.1 1987-1989 345 30 15.5 1987-1989 70 47 13.2 1985-1989 212 53 14.4 1987-1989 216 53 14.2 1987-1989 371 61 16.0 1985-1989 207 62 16.5 1984-1989 139 66 13.2 1985-1989 98 72 15.2 1985-1989 178 75 18.0 1986-1989 389 84 16.3 1984-1989 221 89 15.0 1984-1989 277 99 14.0 1987-1989 a. % of root collar circumference girdled. b. Root Collar Diameter 49 Table 9. Timing of attack ofHylobius warreni on sample trees in a 17 year old lodgepole pine plantation at Mosquito Flats in the Kispiox Forest District, Summer 1990. Tree # % Gird1ed’ RCDb(cm) Year(s) of Attack 83 0 12.6 Nil 107 0 10.9 Nil 160 0 10.6 Nil 179 0 20.4 Nil 20 8 20.5 1987-1988 198 17 21.1 1987-1989 273 18 20.2 1987-1989 39 22 19.5 1987-1989 182 48 20.6 1988-1989 254 57 20.9 1987-1989 48 62 23.3 1984-1989 281 63 21.5 1988-1989 284 65 21.8 1987-1989 44 69 21.6 1987-1989 32 82 18.2 1987-1989 127 84 17.7 1986-1989 277 85 21.2 1985-1989 280 85 20.2 1985-1989 46 98 22.4 1987-1989 a. % of root collar circumference girdled. b. Root Collar Diameter. 50 Table 10. Timing of attack ofHylobius warreni on sampled trees in a 17 year old lodgepole pine plantation at Shandilla in the Kispiox Forest District, Summer 1990. Tree # % Girdleda RCDb(cm) Year(s) of Attack 510 0 10.2 Nil 531 0 11.6 Nil 547 0 14.2 Nil 626 0 11.1 Nil 966 0 12.0 Nil 612 8 19.2 1989 591 21 17.6 Pre-1987 880 21 18.4 1987-1989 549 38 14.9 1987-1989 668 38 20.1 1987-1989 876 51 21.4 1988-1989 921 55 20.6 1987-1989 877 63 19.9 1987-1989 743 79 19.6 1987-1989 862 79 19.6 1987-1989 871 84 20.2 1987-1989 771 89 18.9 1987-1989 673 94 20.4 1986-1989 889 98 20.2 1987-1989 727 99 19.8 1987-1989 a. % of root collar circumference girdled. b. Root Collar Diameter. 51 c-) (I) I 100% 80% 60% 40% 20% 0% 0 5 20 44 79 123 177 241 314 398 491 594 707 TREE BASAL AREA CLASS (cm2) Figure 19. Percentage of sample lodgepole pine with Hylobius warreni damage in relation to basal area class for three plantations in the Kispiox Forest District, Summer 1990. 52 Ta bl e 11 . St em an al ys is o fl od ge po le pi ne tr ee s at ta ck ed by H yl ob iu sw ar re n i a tD at e Cr ee k in th e K isp io x Fo re st D ist ric t, Su m m er 19 90 . A ve ra ge In te rn od e L en gt h G ird le ’ R C D 2 D B H 3 C la ss (cm ) (cm ) 19 80 19 81 19 82 19 83 19 84 19 85 19 86 19 87 19 88 19 89 0% 1O .2 a 4 7. 7a 27 .4 a 37 .4 a 47 .2 a 54 .2 a 60 .O a 49 .8 a 48 .8 a 47 .2 a 44 .O a 55 .6 ab 1- 50 % 16 .2 b 11 .3 b 47 .O a 45 .2 a 59 .8 a 54 .2 a 59 .6 a 62 .2 a 51 .8 a 49 .8 a 42 .4 a 72 .O b 51 -7 0% 14 .91 , 1O .7b 49 .6 a 48 .O a 62 .8 a 61 .4 a 60 .4 a 57 .6 a 59 .O a 57 .8 a 49 .2 a 49 .2 a 71 -9 9% 15 .7 b 11 .2 b 47 .2 a 54 .6 a 44 .2 a 46 .8 a 66 .6 a 69 .2 a 59 .2 a 45 .2 a 45 .6 a 42 .6 a 1. % o fr o o tc o lla rc irc um fe re nc e gi rd led 2. Ro ot Co lla rD ia m et er 3. D ia m et er at Br ea st H ei gh t 4. M ea ns fo llo w ed by th e sa m e le tte rn o ts ig ni fic an tly di ffe re nt (S NK ; P <O .0 5) 53 1000 400 100 Figure 19. Cumulative average tree height of lodgepole pine, in four Hylobius warreni girdling classes, at Date Creek in the Kispiox Forest District, Summer 1990. 900 800 700 600 500 300 200 0 1976 1978 1980 1982 1984 1986 1988 YEAR 54 was significantly greater than the height increment in all other girdling classes (Table 11). However, there were no significant differences in annual height increments between the girdling classes in any other year. 3.2.3.2 Mosquito Flats Root collar diameters were significantly greater in all attacked girdling classes than in the unattacked girdling class (Table 12). The 1-50 % and 5 1-80 % girdling classes had significantly larger DBIfs than did the unattacked girdling class. Average tree height was greater in all three damaged classes than in the unattacked girdling class (Figure 21). Height increment was significantly larger for attacked trees in all girdling classes in 1979 when compared with unattacked trees (Table 12). There were also significant differences between unattacked trees and trees in both the 51-80 % and 81-99 % girdling classes in 1980. However, unattacked trees had shorter height increments in both years. 3.2.3.3 Shandilla Root collar diameters were significantly larger in all attacked girdling classes when compared with the unattacked class (Table 13). The 1-50 % girdling class had significantly smaller RCD’s than did the 51-80 % and 81-99 % girdling classes. Unattacked trees had significantly smaller DBHs than did attacked trees, while trees in the 1-50 % girdling class had significantly smaller DBH’s than those in the 51-80 % girdling class. The average heights of attacked trees in all three girdling classes were taller than trees in the 0% girdling class (Figure 22). Unattacked trees had significantly shorter height increments than those in the 5 1-80 % girdling class during 1979 and 1980 (Table 13). Unattacked trees had significantly shorter increments than trees in all other girdling classes in 1984. Height increments were significantly greater in the 1-50 % class than in the 5 1-80 % class during 1989. 55 Ta bl e 12 . St em an al ys is o f l od ge po le pi ne tr ee s at ta ck ed by H yl ob iu sw ar re n i a tM os qu ito Fl at s in th e K isp io x Fo re st D ist ric t, Su m m er 19 90 . A ve ra ge In te rn od e L en gt h (cm ) G ir dl e 1 R C D 2 D B H 3 C la ss (cm ) (cm ) 19 79 19 80 19 81 19 82 19 83 19 84 19 85 19 86 19 87 19 88 19 89 0% 13 .6 a 4 1O .2a 18 .5 a 27 .8 a 29 .8 a 61 .O a 53 .8 a 61 .O a 61 .3 a 50 .8 a 48 .O a 44 .3 a 55 .5 a 1- 50 % 20 .4 b 13 .5 b 50 .8 b 45 .8 ab 63 .O a 54 .6 a 58 .2 a 61 .2 a 59 .4a 65 .4 a 52 .O a 52 .4 a 52 .2 a 51 -8 0% 21 .8 b 14 .2 b 61 .8 b 52 .8 b 50 .8 a 57 .8 a 51 .8 a 47 .2 a 39 .2 a 45 .O a 45 .8 a 51 .8 a 39 .8 a 81 -9 9% 19 .9 b 12 .5 ab 57 .2 b 55 .6 b 46 .4 a 59 .8 a 51 .6 a 70 .4 a 38 .8 a 51 .8 a 47 .8 a 43 .8 a 39 .8 a 1. % o fr o o tc o lla rc irc um fe re nc e gi rd led 2. Ro ot Co lla rD ia m et er 3. D ia m et er at Br ea st H ei gh t 4. M ea ns fo llo w ed by th e sa m e le tte r n o ts ig ni fic an tly di ffe re nt (S NK ; P <O .0 5) 56 1000 400 300 200 100 YEAR Figure 21. Cumulative average tree height of lodgepole pine, in four Hylobius warreni girdling classes, at Mosquito Flats in the Kispiox Forest District, Summer 1990. 900 800 700 600 500 0 1973 1975 1977 1979 1981 1983 1985 1987 1989 57 Ta bl e 13 . St em an aly sis o fl od ge po le pi ne tr ee s at ta ck ed by H yl ob iu sw ar re n i a tS ha nd ill ai n th e K isp io x Fo re st D ist ric t, Su m m er 19 90 . A ve ra ge In te rn od e Le ng th (cm ) G ir dl e 1 R C D 2 D B H 3 Cl as s (cm ) (cm ) 19 79 19 80 19 81 19 82 19 83 19 84 19 85 19 86 19 87 19 88 19 89 0% 11 .8 a 4 8. 7a 1- 50 % 18 .O b 12 .6b 51 -8 0% 20 .2 c 15 .lc 81 -9 9% 19 .9c 13 .S bc 37 .O a 40 .6 a 41 .4 a 49 .8 a 51 .O a 56 .2 a 52 .4 a 49 .O a 58 .6 a 44 .3 a 52 .O a 44 .8 ab 51 .O ab 48 .6 a 53 .6 a 68 .8 a 71 .4 b 67 .8 a 52 .6 a 49 .2 a 56 .6 a 60 .O b 55 .4 b 61 .4 b 64 .8 a 71 .O b 70 .6 a 74 .8 b 64 .O a 56 .4 a 55 .8 a 55 .4 a 38 .4 a 46 .O ab 56 .4 ab 62 .8 a 67 .4 b 66 .O a 76 .8 b 56 .8 a 62 .4 a 57 .O a 50 .2 a 52 .O ab 1. % o fr oo tc ol lar cir cu m fe re nc e gi rd led 2. Ro ot Co lla rD iam ete r 3. Di am ete r a tB re as tH eig ht 4. M ea ns fo llo we d b yt he sa m e let ter no ts ig ni fic an tly di ffe re nt (SN K; P< Z0 .05 ) 58 1000 400 300 200 100 1975 1977 1979 1981 1983 1985 1987 1989 Figure 22. Cumulative average tree height of lodgepole pine, in four Hylobius warreni girdling classes, at Shandilla in the Kispiox Forest District, Summer 1990. 900 800 700 600 500 0 1973 YEAR 59 4.0 DISCUSSION 4.1 Distribution and Abundance of Warren’s Collar Weevil The 3.99 m radius circular plot was chosen for the district wide survey for a number of reasons. First, circular plots were easier to establish than were strip plots. One person could put in a circular plot, whereas two people would be needed to establish a strip plot, thus requiring twice the time. Ease of establishment was important for the following reasons. In order to fi.iffill the goal of surveying 30 plantations in a 16 week period, it was necessary to optimize labour resources. The circular plot enabled almost twice the area to be surveyed with the same size crew. Also, if the survey was to be cost effective in the long run (i.e. could be integrated into the present Ministry of Forests silviculture surveys format), it was advantageous to have a method that could be implemented by a crew of one. This would keep survey costs at a reasonable level. Second, circular plots are commonly used in other silviculture assessments such as regeneration and free-growing surveys. The data required to assess Warren’s collar weevil abundance could be easily collected at the same time as a free-growing survey. The survey method was quite effective in assessing weevil damage. In only one case, the Seven-Sisters plantation in the SW region of the district, did the plots fail to contain trees with weevil larval damage when there was evidence of it within the plantation. Third, the time required to establish a circular plot was generally less than the time needed to put in strip plots (this was dependent on tree density). For strip plots, data gathering was a continuous process, whereas in circular plots, more time was spent travelling between plots. 60 Fourth, circular plots could be distributed more evenly throughout a plantation than strip plots. Because of the continuous nature of strip plots, the minimum coverage within a given hectare was 2% of the area. To achieve the desired sampling intensity of 1% of the area for the study, it would not be possible to survey every hectare within a plantation using the strip plot method. The circular plot ensured that there were at least two plots per hectare (or 1 plot/ha for a 0.5 % intensity survey). Weevil populations are contagious in distribution (Cerezke 1969), thus the circular plot survey may be more usefhl for detecting weevil attacked trees, since it covers more ground than does the strip survey. The lesser perimeter of the circular plot (25 m vs 54 m for strip plots) resulted in fewer judgement calls ofwhether a tree was in or out of the plot. This minimized potential error associated with judgement calls. It also reduced the time spent determining if a tree was within a plot or outside of it. The surveyed plantations were well distributed within the district, and provided a good overview of the distribution of Warren’s collar weevil. The wide range of ages of plantations surveyed provided some insight into the expected levels of Warren’s collar weevil in the development of plantations to the free-growing stage. For example, based on this survey, a 6 year old stand with 5 % weevil incidence might be expected to have a 45 % weevil incidence by age 15 years. Plantations within a given drainage were of similar ages. For example, plantations in the Upper Kispiox River were relatively young (5-7 years), while plantations in the Suskwa River drainage were among the oldest in the district (10-16 years). The average percentage of sampled lodgepole pine with larval damage was found to be greater in the Kispiox Forest District than values reported in Alberta. Near Robb, Alberta, Cerezke (1969) reported attack rates ofbetween 2.8% - 13.7% for 2.8 m -3.2 m tall lodgepole pine. 61 The average number of weevils per tree was greater than those previously reported for similar stands. Cerezke (1970c) found between 0.03 and 0.13 weevils per tree in 15 and 20 year old stands in Alberta. This was considerably less than the 0.33 weevils per tree found in this study. The higher weevils per tree found in this study may have been a result of differences in stand density and not population levels. In Alberta, Cerezke’s studies were done in natural lodgepole pine stands where tree densities were quite high ( > 2500 sph). This study concentrated on pine plantations in which planting densities were generally in the range of 1100 -1300 sph. Thus, in plantations, weevil numbers per tree may be greater than would be found in naturally regenerated pine plantations. The average mortality rates attributed to Warren’s collar weevil larval damage were within levels reported by Cerezke (1969). Some young plantations within the district had high levels of larval caused mortality. In particular, two 5 year old plantations in the Salmon River area had high mortality levels (6% and 8.8%). These plantations were established on a flat bench adjacent to the Salmon River. The high mortality levels may have been a result of the harvesting pattern in this area. The original stands were dominated by mature lodgepole pine which was attacked by mountain pine beetle (Dendroctonusponderosae Hopkins). These stands were salvage harvested and planted to lodgepole pine the following year. Residual mature lodgepole pine within the area exhibited evidence of root collar weevil. The weevil populations in the clearcut areas immediately following harvesting (i.e., the first two years post-harvest) would have increased due to decreased development times in the cut stumps (Cerezke 1973b). These weevils would have dispersed from the clearcut areas to find any available suitable host. In this case, they attacked residual pine adjacent to clearcut areas. These adjacent areas were subsequently logged (i.e. 3 and 4 years following the harvest of the surveyed plantations), further decreasing the pine component in the area. Weevil populations in these blocks following harvest would also have increased; however, 62 the residual populations would have had few places to inhabit because of the continued removal of the pine type. Because of a lack of larger, more suitable pine trees in the area, adult weevils would have then invaded the original pine plantations, which were then 4 or 5 years old. The high levels of mortality would have been a result of both the high weevil populations and the small size of the host. The free-growing status of surveyed plantations was being affected by root collar weevil. By definition, a free-growing, or acceptable well-spaced stem, must be free of insect damage. The survey indicated that potential well-spaced stems were being attacked by root weevil. The effect that this insect will have over the duration of the stand is not known. Mortality rarely occurs in attacked trees older than 30 years (Cerezke 1969). Perhaps free-growing assessments should be delayed in plantations known to support weevil infestations to ensure that the weevil is not preventing free-growing standards from being achieved. Another alternative would be to schedule reconnaissance surveys in plantations between 20 and 30 years old to assess the health and stocking of stands which have been declared free-growing. In the Kispiox Forest District, it can be expected that the percentage of trees attacked by root collar weevil will increase as stand age and average tree height within plantations increase. This will occur for two reasons: increasing suitability of the host trees, and relatively constant stand densities. The population ofweevils within naturally regenerated stands is constant over the life of that stand (Cerezke 1 970a). As the stand ages, natural thinning processes reduce tree densities. Those trees that remain will be the larger, more dominant trees which are better able to compete for light and available resources. It is this group of trees which is most susceptible to attack by Warren’s collar weevil. In plantations, trees are planted at target harvest densities, so natural mortality is much lower than in natural stands. Therefore, the percentage of stems 63 attacked in plantations will increase much more rapidly than in natural stands since the number of stems are already at culmination densities. In addition, diameter growth will be greater in plantations because resources are not being lost to trees that would be naturally thinned out. Thus planted trees may provide better weevil habitat at a younger age than naturally regenerated trees. The probability of a tree being attacked was related to its basal area. The proportion of trees attacked within a basal area class increased with increasing basal area. This may be related to the nutritional quality of the host tree. The suitability of the phloem as a nutritional source for larvae may increase with increasing tree size. It has also been suggested (G. Weetman pers. comm.) that the incidence of the weevil may be related to crown closure. The ability of a stand to support a weevil infestation may increase as it approaches 100% crown closure. The understorey shading associated with crown closure may be a critical factor in weevil survival. There was no apparent relationship between the percentage of trees with Warren’s collar weevil damage and the total number of stems/ha (Figure 5). Cerezke (1970c) reported evidence that excessively stocked stands provided poor weevil habitat. None of the plantations surveyed were excessively stocked with dominant/co-dominant pine, as is found in some natural stands. Therefore, I would not expect tree density to have a large effect on the percentage of trees attacked. Some surveyed plantations had higher tree densities (> 2500 sph). These were composed of two tree layers. Trees that had seeded in naturally since planting (layer 1 trees) were generally smaller trees and germinants. It was this layer that resulted in the high densities recorded. Planted trees (layer 2 trees) were larger and at lower densities (—- 1200 sph). These trees had the most influence on stand conditions, and they better reflect the conditions that may affect the susceptibility of that stand to support weevil populations. 5Professor, Faculty of Forestry, The University of British Columbia 64 The depth of the LFH layer did not appear to be important in determining if trees were attacked in either study year. Cerezke (1970c) found that the duff depth (LFH) gave an indication of the quality of habitat, with the thicker, moister duff depths being the most suitable for larval survival. The method used to measure LFH depths in the first year of the study did not measure duff depth directly at the tree base. It was thought that this may have been the reason why a good relationship was not found. During the second year of the study, measurements of the LFH were made directly at the tree base. Measuring the LFH depth in this manner did show a significant relationship between the percent of the stem circumference girdled and the LFH depth. Although significant the equations accounted for only a small portion of the variation in circumference girdled at all three study sites. A more important factor to consider when assessing weevil habitat may be the quality of the LFH layer. Cerezke (1969) found that there were similarities in the LFH layer at the base of attacked trees. He found that sites with a mixture of a moss and herb layer over the LFH layer were common in attacked trees. This was also the case in the Kispiox district. Those sites with a mixture of herbs and mosses and a moist LFH layer had higher attack incidences. The presence of slash and small logs at the base of trees was also common in attacked trees. Presumably, this woody debris provides moisture and protection for developing larvae. Pine plantations on circum-mesic, well-drained sites will be susceptible to infestation by Warren’s collar weevil. Generally, these are the more productive sites for growing lodgepole pine. Proposed openings on such sites should be examined carefully at the pre-harvest prescription phase to assess the probability that weevils will affect the next stand. Suitable prescriptions can then be developed to limit the impact of Warren’s collar weevil over the next rotation. The presence of pine adjacent to a plantation is important in determining its susceptibility to root collar weevil infestation (Cerezke 1989). The adjacent pine serves as a reservoir for root collar weevil populations. 65 Surveyed plantations which were site prepared either with burning or with scarification did not have lower percentages of weevil attacked trees than those plantations with no site preparation. In fact, openings which had a site preparation treatment had higher percentages of weevil attacked trees than those that did not. Site preparation is thought to reduce weevil survival by reducing the LFH layer, which is important to adult weevil survival (Cerezke 1973b, 1989). The benefits of site preparation may be being masked by the age of the plantations surveyed and the harvest scheduling. The average age for broadcast burned blocks was 16 years, 8 years for spotburn blocks, 6 years for no treatment, and 5 years for scarified blocks. Perhaps enough time had elapsed to allow for sufficient build-up of organic material in the burned blocks to support weevil re-invasion. Of the scarified blocks, only two had high weevil incidence. These were found in the Salmon River area, discussed above. The change in forest structure as a result of mountain pine beetle salvage may be more responsible for the high weevil populations than the failure of scarification to reduce the LFH layer. Site preparation may be an effective tool in delaying the onset of infestations. This may include broadcast burning and scarification. 66 4.2 Height Growth and Dispersal Study With the exception of the Shandilla plantation, weevil populations found in the three study areas were consistent with those previously reported (Cerezke 1 970c). In Alberta, weevil numbers per hectare in 15-20 year old stands were in the range of 375 to 920. The populations at Date Creek and Mosquito Flats were 850 weevils/ha and 1100 weevils/ha, respectively. Populations at Shandilla were considerably higher, with almost 3700 weevils/ha. The reason for the high populations may have been related to the original stand composition and the surrounding forest cover. The original stand was composed of approximately 15 -20 % lodgepole pine. The leading species on the site were hemlock and spruce. The surrounding forest cover consisted of mixed species stands dominated by western hemlock. The pine component in the area was negligible. The existing stand was an island of pine within a hemlock dominated forest. After harvesting of the original block, weevils may have dispersed to surrounding residual pine, and maintained the population until the planted pine was of susceptible size. The population then increased as a result of an increase in their preferred host. Dispersal from the pine plantation would be limited by the lack of lodgepole pine at the plantation periphery. When searching for hosts, adult weevils would be more successful if they remained within the plantation. This likely contributed to very high population levels within the Shandilla plantation. There was a significant relationship between the root collar diameter of attacked trees and the percentage of the stem circumference girdled in all three sampled plantations. Although the relationship was significant the equation accounted for only a small portion of the variability in the percent of the root collar circumference girdled. The percent of the stem girdled increased with 67 increasing root collar diameter; however, the usefi.ilness of this relationship is limited by the low R2 value associated with it. The destructively sampled trees in this study had been attacked within the previous 5-6 years. This would imply that weevil populations may have increased dramatically within the previous five years. For instance, 92% of all lodgepole pine at Shandilla had been attacked within the previous four years. Likewise, 81 % and 87 % of lodgepole pine trees at Date Creek and Mosquito Flats respectively, had been attacked within the previous six to seven years. These attack rates were much greater than those reported in Alberta. Cerezke (1969) estimated that 10 - 30 % of the trees in 16-25 year old natural pine stands had evidence ofweevil attack. Also, attack levels of greater than 90 % of a stand did not occur until they were at least 60 years old. His studies were done in natural stands in which high initial tree densities occurred. These stands will gradually thin by natural mortality, thus the increase in the percentage of trees attacked with relation to stand age is partially an artifact of decreasing stand density. In north central British Columbia, however, where plantations are planted at or near target densities, there is relatively little change in tree density during the life of the stand. Additionally, there are relatively few trees for weevils to attack compared to natural pine stands of similar age. The high incidence of root weevils within the Kispiox Forest District in general, may be a result of the conversion of natural mixed species stands to pine plantations. The short period over which the natural pine stands within the district were harvested could also have contributed to the current high population levels. Weevil populations were likely present as small endemic infestations in natural stands prior to these stands being converted to managed forests. Reliance on lodgepole pine for regenerating circum-mesic sites within the district has resulted in a large increase in available weevil habitat. Contributing to this increase was the removal of much of the mature pine during mountain pine beetle salvage operations. Removing the majority of pine in 68 certain areas, the Salmon River area for example, forced the weevil to disperse into young pine plantations. Had remnant pine stands been harvested over a longer time period, the population pressure within young plantations may have been reduced. The time of weevil attack found on destructively sampled trees was uniform throughout the plots. Trees within the interior of the plot were attacked during the same time period as those trees at the periphery of the plot located at the stand margin. Cerezke (1969) estimated that weevil dispersal rate from the stand margin into naturally regenerated stands would be 10-15 m per year. To support this, destructively sampled trees at the stand periphery should indicate younger ages of attack than those sampled toward the interior of the plantation. This was not the case for the Kispiox. Weevils may be capable of longer distance dispersal than previously thought. Another possibility is that weevils remain in the clearcut and do not disperse to surrounding mature trees. Weevil larvae may complete development in cut stumps and the resultant adults could survive for up to four years. During this time, the adults would feed on residual conifer species until trees reached a size suitable for females to oviposit. Since it requires two years for newly hatched larvae to complete development within the cut stumps (Cerezke 1973b), the time from development of an adult to the time of its death would be six years after harvest. Thus planted trees within the cutblock would just be susceptible to attack prior to the fourth year of an adult weevil’s lifespan. The pattern of weevil infestations was not clarified in this study. Destructive sampling indicated that trees in the attacked girdling classes were significantly greater in size than were trees in the unattacked girdling class for all three study sites. This illustrated the weevil’s preference for the larger dominant and co-dominant trees within a stand (Cerezke 1 970a). Adult weevils oviposit at the root collars of the larger trees, presumably because they provide a better food source for the developing larvae. The mechanism by which 69 adults select the larger trees within the stand is not known. Adults may choose trees based on visual cues such as stem thickness or tree height. This method of host selection has been demonstrated for the white pine weevil, Pissodes strobi Peck (VanderSar and Borden 1977). Alternatively, adults may choose trees based on olfactory cues. A related species, Hylobius pales (Herbst), a pest of plantation pines, has been shown to be attracted by a combination of turpentine and ethanol (Phillips et. at. 1988). Adults of Warren’s collar weevil may respond to these compounds which are released by host trees. Weevil response may be dependent on the release rates of these compounds: i.e. larger trees emit larger quantities of such compounds than do smaller trees, thus attracting more weevils. The results from the stem analysis indicated that there were no short-term impacts on lodgepole pine height growth as a result of weevil injury. Previous studies had found that girdling damage to the stem or root collar of lodgepole pine may result in growth losses. Cerezke (1 970c) found reductions in height growth in pine damaged by root weevil larvae. Height growth was reduced by 11.5 % in the second year and 16.4 % in the third year following 50 % of the stem circumference girdled. In a partial girdling experiment, height growth declined gradually until 60 % of the stem was girdled. After 60 % of the stem was girdled, height growth declined rapidly (Cerezke 1974). Sullivan and Sullivan (1986) found that semi-girdling by the snowshoe hare (Lepus americanus Erxleben) significantly reduced height growth in lodgepole pine. Trees less than 6.0 cm at d.b.h. were most affected. Sullivan and Vyse (1987) reported conflicting results on the impact of semi-girdling damage by red squirrels (Tamiasciurus hudsonicus Erxleben) on height growth of lodgepole pine. Height growth was significantly affected in one stand, but it was not affected in another stand. However, they warned that growth impacts may not become evident until several years following the damage. Additionally, further feeding damage may adversely affect already damaged trees and thus increase the probability of future height growth 70 reductions. This may reflect the situation found in this study. As trees are further damaged by subsequent weevil attacks, it is likely that growth reductions will occur. The average d.b.h. values for sample trees in this study were between 7.9 cm and 14.1 cm. These values are greater than the 6.0 cm values reported by Sullivan and Sullivan (1986). It is possible that the larger trees are more capable of repairing or overcoming weevil injury, although there was no evidence to indicate this in this study. Further, this study may not have found significant differences in height growth because of the obvious differences between girdled and control trees. The Kispiox Forest District will likely continue to support higher weevil populations than those experienced in the past. The reliance on the use of lodgepole pine for reforesting natural mixed species stands will provide increasing habitat for Warren’s collar weevil. The challenge is to keep populations of this insect at levels which are tolerable from a timber management perspective. Some management strategies which have been suggested include delaying planting until 2-3 years post-harvest, mixed species planting, planting at higher densities, and site preparation (Cerezke 1989). Management regimes which more closely mimic natural stands may reduce the impact of the weevil. These would include regeneration of mixed species historically found in the region. 71 5.0 CONCLUSIONS 1. Warren’s collar weevil is found throughout the Kispiox Forest District. 2. Warren’s collar weevil is not reducing plantations to below minimum required stocking levels. Weevil-caused mortality was generally quite low (< 5%) with a few exceptions. The free- growing status of lodgepole pine plantations is being affected by Warren’s collar weevil. The presence of larvae and/or larval damage to well-spaced trees prevents them from being declared free-growing under current Ministry of Forests guidelines. 3. This study did not identify specific site factors which could be utilized to predict or identify areas which have the potential to support Warren’s collar weevil infestations. A combination of general site conditions can identify potential problem areas. Pine plantations established on circum-mesic sites which previously had, or were adjacent to, timber types with a pine component appear to be most susceptible to colonization by Warren’s collar weevil. 4. Destructively sampled lodgepole pine trees with Warren’s collar weevil damage have been attacked over the previous 5 - 6 years. Trees have been attacked more than once during this time period. 5. Destructively sampled lodgepole pine trees with Warren’s collar weevil damage were attacked during the same time period. There was no apparent progression of weevil infestation from the stand margin to the interior of the stand as reported in Alberta. 72 6. Warren’s collar weevil has not yet had an effect on the height growth of attacked lodgepole pine trees. This finding may be confounded by the weevils apparent preference for larger diameter, dominant and co-dominant trees within a stand. 73 6.0 RECOMMENDATIONS 1. Warren’s collar weevil should continue to be monitored in pine plantations within the Kispiox Forest District. The most suitable means of doing this is through the silviculture surveys system currently in use. The impact of the weevil on a site specific basis should be monitored either in conjunction with stocking surveys or through forest health surveys. 2. Warren’s collar weevil should be monitored on a regional and district basis to better understand the impact of the weevil on a forest level basis. 3. The acceptability of well-spaced trees with Warren’s collar weevil damage as free-growing stems should be resolved. The high incidence of damage within existing lodgepole pine plantations poses a potential liability to both the Crown and Licencees in the event that attacked stems cannot be declared free-growing. 4. The impact of Warren’s collar weevil on the growth of attacked stems should be further studied. Longer term, more tightly controlled studies should be implemented to fi.irther determine the impact of larval feeding damage on the growth of attacked trees. 5. Operational trials that reduce populations of Warren’s collar weevil in infested stands should be implemented in conjunction with harvesting activities. The best opportunity to limit damage by Warren’s collar weevil is by limiting their re-invasion into plantations. The best means of achieving this should be examined. 74 6. Implement a detailed study of post-harvest survival of Warren’s collar weevil in a stand, or stands, with a chronic infestation. 75 BIBLIOGRAPHY Castellano, A., and M. Marsh. 1989. Control of the bark feeding weevil (Hylobius abietis) using a controlled release of the insecticide carbosulfan. Incitec Ltd., Queensland Australia. Promotional publication. 8 pp. Cerezke, H.F. 1967. A method for rearing the root weevil Hylobius warreni Wood (Coleoptera: Curculionidae). Can. Entomol. 99:1087-1090. Cerezke, H.F. 1969. The distribution and abundance of the root weevil, Hylobius warreni Wood, in relation to pine stand conditions in Alberta. Ph.D. thesis, University of British Columbia, xvii + 221 pp. Cerezke, H.F. 1970a. A method for estimating abundance of the weevil, Hylobius warreni Wood, and its damage in lodgepole pine stands. For. Chron. 46:392-396. Cerezke, H.F. 1970b. Survey report of the weevil, Hylobius warreni Wood, in the foothills of Alberta. Can. For. Serv., For. Res. Lab., Edmonton, AB. Tnt. Rep. A-38. 40 pp. Cerezke, H.F. 1970c. Biology and control of Warren’s collar weevil, Hylobius warreni Wood, in Alberta. Can. For. Serv., For. Res. Lab., Edmonton, AB. Tnt. Rep. A-27. 28 pp. Cerezke, H.F. 1972. Effects of weevil feeding on resin duct density and radial increment in lodgepole pine. Can. J. For. Res. 2:11-15. Cerezke, H.F. 1973 a. Bark thickness and bark resin cavities on young lodgepole pine in relation to Hylobius warreni Wood (Coleoptera: Curculionidae). Can. 3. For. Res. 3:599-601. Cerezke, H.F. 1973b. Survival of the weevil, Hylobius warreni Wood, in lodgepole pine stumps. Can. J. For. Res. 3:367-372. Cerezke, H.F. 1974. Effects of partial girdling on growth in lodgepole pine with application to the weevil Hylobius warreni Wood. Can. 3. For. Res. 4:312-320. Cerezke, H.F. 1989. Proceedings of root collar weevil meeting, August 1989, Hazelton, B.C. B.C. Ministry of Forests, Prince Rupert Forest Region, Memorandum. 14 pp. Coulson, R.N., and J.A. Witter. 1984. Forest Entomology. Wiley-Interscience. New York. 669 pp. Duncan, R.W. 1986. Terminal and root-collar weevils of lodgepole pine in British Columbia. Can. For. Serv., Pac. For. Ctr., For. Pest Leaflet 73. 6 pp. Finnegan, R.J. 1961. A field key to the North American species ofHylobius (Curculionidae). Can. Entomol. 93:501-502. 76 VanderSar, T.J.D., and J.H. Borden. 1977. Visual orientation ofPissodes strobi Peck (Coleoptera:Curculionidae) in relation to host selection behaviour. Can. J. Zool. 55:2042- 2049. Warner, R.E. 1966. A review of the Hylobius ofNorth America with a new species injurious to slash pine (Coleoptera: Curculionidae). The Coleopterists’ Bull. 20:65-81. Warren, G.L. 1956a. Root injury to conifers in Canada by species ofHylobius and Hypomolyx (Coleoptera: Curculionidae). For. Chron. 32:7-10. Warren, G.L. 1956b. The effect of some site factors on the abundance ofHypomolyxpiceus (Coleoptera: Curculionidae). Ecology. 37:132-139. Warren, G.L. 1956c. ph and the incidence of attack ofHypomolyxpiceus (DeG.). Can. Dept. Agric., Sci. Ser., For. Biol. Div., Bi-Monthly Prog. Rep. 12:2-3 Warren, G.L. 1958. A method of rearing bark and cambium-feeding beetles with particular reference to Hylobius warreni Wood (Coleoptera: Curculionidae) Can. Entomol. 90:425- 428. Warren, G.L. 1960. External anatomy of the adult ofHylobius warreni Wood (Coleoptera: Curculiomdae) and comparison with H. pinicola (Couper). Can. Entomol. 92:321-341. Warren, G.L., and R.D. Whitney. 1951. Spruce root borer (Hypomolyx sp.), root wounds, and root diseases of white spruce. Can. Dept. Agric., Sci. Ser., For. Biol. Div., Bi-Monthly Prog. Rep. 7:2-3 Whitney, R.D. 1952. Relationship between entry of root-rotting fungi and root wounding by Hypomolyx and other factors in white spruce. Can. Dept. Agric., Sci. Ser., For. Biol. Div., Bi-Monthly Prog. Rep. 8:2 Whitney, R.D. 1961. Root wounds and associated root rots of white spruce. For. Chron. 37:401-411. Wilson, L.F. 1967. Effects of pruning and ground treatments on populations of the pine root weevil. J. Econ. Entomol. 60:823-827. Wilson, L.F., C.D. Waddell, and I. Millers. 1966. A way to distinguish adult Hylobius weevils in the field. Can. Entomol. 98:1118-1119. Wood, S.L. 1957. The North American allies ofHylobiuspiceus (DeGeer) (Coleoptera: Curculionidãe). Can. Entomol. 89: 37-43. Zar, J.H. 1984. Biostatistical Analysis. 2nd Ed. Prentice-Hall Inc. 718 pp. 78

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
Germany 3 14
United States 2 0
China 2 1
City Views Downloads
Unknown 3 16
Beijing 2 0
Sunnyvale 1 0
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items