www.elsevier.com/locate/foreco

Forest Ecology and Management 202 (2004) 369–377

Weevil resistance of progeny derived from putatively resistant

and susceptible interior spruce parents

Rene I. Alfaroa,*, Lara vanAkkera, Barry Jaquishb, John Kingc

aPacific Forestry Centre, Canadian Forest Service, 506 West Burnside Road, Victoria, BC, Canada V8Z 1M5bKalamalka Research Station, British Columbia Ministry of Forests, 3401 Reservoir Road, Vernon, BC, Canada V1B 2C7

cBritish Columbia Ministry of Forests, P.O. Box 9519, Stn Prov Govt, Victoria, BC, Canada V8W 9C2

Received 2 April 2004; received in revised form 29 July 2004; accepted 3 August 2004

Abstract

Controlled-cross progeny of interior spruce (Picea glauca (Moench) Voss � Picea engelmanni Parry ex Engelm.) parents

ranked as resistant or susceptible to the white pine weevil, Pissodes strobi (Peck) were screened for resistance to the same insect

by augmentation of the trial site with weevils. Progeny from two resistant parents (R � R progeny) sustained significantly fewer

weevil attacks (13% of the trees were attacked) in the year following the augmentation, than progeny from susceptible parents (S

� S progeny) (68% of the trees were attacked). Progeny obtained by crossing one resistant and one susceptible parent (R � S

progeny) sustained intermediate attack levels (47% were attacked). Characteristics of the bark resin canals of the crosses were

explored using microscopy techniques. Bark resin canal density was highest in R � R progeny, lowest in S � S progeny and

intermediate in R � S progeny. There was a negative correlation between the percentage of trees attacked in each cross and the

average density of the outer resin canals for each cross. A discriminant function was developed that distinguished between

resistant and susceptible progeny using bark characteristics. The function was characterized by positive coefficients for outer

resin canal density and inner resin canal size, and a negative coefficient for bark thickness. Thus, trees with thin bark, large inner

resin canals and dense outer resin canals are more likely to be resistant to P. strobi.

# 2004 Elsevier B.V. All rights reserved.

Keywords: Genetic resistance; Pest resistance; Pest management; Pissodes strobi

1. Introduction

The destruction of spruce leader growth by the

white pine weevil (Pissodes strobi Peck) causes

millions of dollars of losses to the British Columbia

* Corresponding author. Tel.: +1 250 363 0648;

fax: +1 250 363 0774.

E-mail address: ralfaro@pfc.forestry.ca (R.I. Alfaro).

0378-1127/$ – see front matter # 2004 Elsevier B.V. All rights reserved

doi:10.1016/j.foreco.2004.08.001

(BC) forest industry as a result of stem defects, which

decrease lumber quality (Alfaro, 1989; Alfaro, 1994)

and reduce yield by up to 40% (Alfaro et al., 1997b).

Pissodes strobi is native to North America, and is

distributed across the continent. In the east P. strobi

feeds primarily on white pine (Pinus strobus L.) and

Norway spruce (Picea abies (L.) Karst.) while in BC

its main hosts are Sitka spruce (Picea sitchensis

(Bong.) Carr.), white spruce (P. glauca (Moench)

.

R.I. Alfaro et al. / Forest Ecology and Management 202 (2004) 369–377370

Voss), Engelmann spruce (P. engelmanii Parry) and

interior spruce hybrids (P. glauca � P. engelmanni)

(Humble et al., 1994).

Damage results from weevil larvae consuming the

phloem beneath the bark of the uppermost tree shoot

or terminal leader. As the larvae mine downwards in

synchrony they girdle and usually kill the leader. If the

attack is successful, the weevil completes its lifecycle,

which usually results in the death of a minimum of 2

years of terminal growth (Silver, 1968). P. strobi is

considered the most damaging pest of spruce

regeneration in BC (Hall, 1994). On the BC coast,

weevil damage is so severe that it has severely limited

the use of Sitka spruce in reforestation programs. In

BC’s interior there are thousands of hectares of pure

interior spruce plantations that are currently suscep-

tible to weevil attack (BC Ministry of Forests, 2002).

Silvicultural, chemical and biological controls have

been investigated but none has proven completely

successful (Alfaro et al., 1995). However, genetic

resistance is being considered as a promising tool for

weevil management (Alfaro et al., 2002). Over the last

10 years, the BC ministry of Forests and the Canadian

Forest Service have established and monitored a

number of progeny trials for the purpose of ranking

families for growth traits, estimating genetic para-

meters and screening spruce for genetic resistance to P.

strobi. The first interior spruce trial was established in

1973 and consisted of 173 wind-pollinated families

planted at three test sites in north central BC (Kiss and

Yeh, 1988; Kiss and Yanchuk, 1991; Xie and Yanchuk,

2002). Seventeen years after planting, weevil resistance

rankings (averaged over the three sites) were calculated,

based on the performance of each family in terms of

weevil damage. Weevil resistance in these trials was

shown to be heritable with a family heritability value of

0.7 (Kiss and Yanchuk, 1991; King et al., 1997). The

weevil resistance rankings developed in this study

allowed for the selection of parents and the production

of an F1 progeny generation.

A trial to investigate the resistance of the F1

generation was initiated at the BC Ministry of Forests’

Kalamalka Research Station in Vernon, BC, utilizing

progeny from three types of controlled crosses that

were developed based on the resistance rankings

identified by Kiss and Yanchuk (1991): (1) both

parents identified as putatively weevil resistant, (2)

one parent identified resistant and one susceptible, or

(3) both parents identified as weevil susceptible.

Although the F1 generation cannot be used to

determine if there is any segregation for genes of

large effect, it does at this time provide the most

advanced pedigreed population for investigating the

inheritance of resistance. As well as facilitating

screening of the F1 generation, the establishment of

this trial provided the opportunity to explore tree

characteristics that may act as mechanisms of

resistance to the weevil. To date, it is thought that

resistance is likely the result of a suite of mechanisms

whose composition varies from genotype to genotype

(Alfaro et al., 2002). The resin canals that occur in the

bark of many conifer species are thought to be one

such mechanism contributing to defense against insect

and pathogen attack by acting as physical and

chemical barriers to invasion. In spruce species, there

is a continuous ring of larger resin canals located

toward the inside of the cortex (inner resin canals)

while smaller canals are located, usually in pairs,

within the sterigmata ridges (outer resin canals) (Jou,

1971). This network of resin canals is effectively a

physical and chemical ‘‘mine field’’, which weevil

adults and young larvae avoid during feeding and

oviposition (Alfaro et al., 2002). When punctured,

resin canals ooze resin into the feeding and oviposition

punctures, deterring the feeding adult, or drowning the

eggs and young larvae. As a consequence, the

distribution of resin canals in the bark affects the

amount of food and space available for weevil feeding

and oviposition, and may influence host selection as

well as weevil survival. High bark resin canal density

is thought to contribute to the weevil resistance

observed in some families of interior and Sitka spruce

(Tomlin and Borden, 1994, 1997; Alfaro et al., 1997a).

This paper presents the results of the screening of

F1 interior spruce progeny for weevil resistance. We

also describe the bark resin canal characteristics of the

crosses and determine the feasibility of using these

traits to screen trees for weevil resistance.

2. Materials and methods

2.1. Trial establishment

The controlled-cross progeny trial was established

at the Kalamalka Research Station, Vernon, BC in

R.I. Alfaro et al. / Forest Ecology and Management 202 (2004) 369–377 371

Table 1

Parent identities and resistance classes for progeny obtained by

controlled crosses of putatively weevil resistant and susceptible

interior spruce

Cross Cross

resistance

Female

parent no.

Female

resistance

Male

parent no.

Male

resistance

1 R � R 1 R 21 R

2 R � R 1 R 29 R

3 R � R 1 R 87 R

4 R � R 1 R 167 R

5 R � R 1 R 1645 R

6 R � R 21 R 29 R

7 R � R 21 R 161 R

8 R � R 21 R 167 R

9 R � R 21 R 1645 R

10 R � R 29 R 161 R

11 R � R 29 R 167 R

12 R � R 87 R 161 R

13 R � R 87 R 167 R

14 R � R 87 R 1645 R

15 R � R 167 R 161 R

16 R � R 29 R 87 R

17 R � S 1 R 79 S

18 R � S 1 R 165 S

19 R � S 1 R 72 S

20 R � S 1 R 117 S

21 R � S 161 R 79 S

22 R � S 161 R 165 S

23 R � S 161 R 128 S

24 R � S 161 R 72 S

25 R � S 161 R 117 S

26 R � S 87 R 165 S

27 R � S 87 R 117 S

28 R � S 21 R 79 S

29 R � S 21 R 165 S

30 R � S 21 R 128 S

31 R � S 21 R 72 S

32 R � S 21 R 117 S

33 R � S 167 R 79 S

34 R � S 167 R 72 S

35 R � S 29 R 79 S

36 R � S 1 R 128 S

37 S � S 79 S 128 S

38 S � S 79 S 72 S

39 S � S 79 S 117 S

40 S � S 165 S 72 S

41 S � S 98 S 128 S

42 S � S 165 S 128 S

May 1995. Nine female parents and 11 male parents

were used to generate a total of 3150 F1 progeny of 42

full sibling families (Table 1). The parents for the

controlled-crosses were selected based on weevil

resistance rankings determined by Kiss and Yanchuk

(1991). Their study was based on 173 trees, selected

for growth characteristics in 1973, from within the

Prince George Selection Unit of BC. The selected

parent trees were tested for weevil resistance in three

open-pollinated progeny trials that had endemic

weevil populations. Resistance rankings were based

on retrospective weevil damage assessments con-

ducted in 1989 by Kiss and Yanchuk (1991), when the

trees were 17 years old at Aleza Lake and 16 years old

in the other two sites. The percent of trees damaged by

weevils, was calculated for each family in each

replicate based on the history of weevil attack since

plantation establishment. Average damage over the

three sites was calculated for each family, and the 173

families were ranked based on performance in terms

of weevil damage (Kiss and Yanchuk, 1991).

In addition to the Prince George parents, one

further parent (male parent #1645, Ontario origin) was

selected as putatively resistant as a result of

observations of low incidence of weevil attack within

a clone bank at the Kalamalka Research Station.

Of the crosses selected for the Kalamalka con-

trolled-cross progeny trial, 16 were putative resistant

female � resistant male (crosses 1–16); 20 were

putative resistant female � susceptible male (crosses

17–36); and 6 of the crosses were susceptible female�susceptible male (crosses 37–42) (Table 1). The trial

was established in a randomized complete-block

design with three replicate blocks. Trees were planted

in 25-tree square plots at 1.25 m � 1.25 m spacing.

2.2. Screening controlled crosses for resistance

A natural population of P. strobi occurred in the

area, which caused sporadic attacks in the plantation.

By the summer of 1999, 4 years after trial establish-

ment, 13.6% of the trees had sustained damage by

these local weevils. To increase the rate of attack and

facilitate detection of differences among crosses, P.

strobi populations were augmented by placing three

weevils on each tree in early October 1999. To achieve

this, weevil infested leaders were clipped from spruce

trees surrounding the Kalamalka site, in late July

1999. The leaders were then placed in cages and

emerging adults were collected and maintained in

5 gal pails containing food (cut spruce branches) and

water until they were released. Each tree in the trial

was assessed annually from 2000 to 2003, to

determine resulting attack rates among the crosses.

R.I. Alfaro et al. / Forest Ecology and Management 202 (2004) 369–377372

Table 2

Acronyms used to represent resin canal variables to describe the

bark of interior spruce leaders and laterals

Acronym Variable description

SZin Size of inner resin canals in square microns

SZout Size of outer resin canals in square microns

NMSin Number of inner resin canals per square

millimetre of bark

NMSout Number of outer resin canals per square

millimetre of bark

NMSall Total number of resin canals per square

millimetre of bark

AOCin Percentage of bark area occupied by

inner resin canals

AOCout Percentage of bark area occupied by

outer resin canals

AOCall Percentage of bark area occupied

by all resin canals

BTHK Bark thickness in millimetres

An attack was considered successful when weevil

oviposition and larval feeding resulted in the death of a

portion of the terminal growth, while failed attacks

were evidenced by weevil oviposition and no

subsequent destruction of the leader.

2.3. Bark resin canal study

In June 1999, lateral branch samples were collected

from each of the 42 crosses to measure cortical resin

canal density. Ten trees that had not been previously

attacked by the white pine weevil were selected for

sampling from each cross. From each of the 10 trees a

5 cm long branch sample was collected from the distal

end of a lateral branch in the uppermost whorl.

Laterals from this whorl develop from the same

meristem and in the same year as the leader. In order to

establish the relationship between leader and lateral

bark resin canal density, a 5 cm terminal leader sample

was also collected approximately 5 cm below the

terminal bud (the site where P. strobi normally

deposits its eggs) from one of the 10 trees selected

from each cross.

Samples were placed in glass vials and fixed in

formalin acetic acid (FAA) for approximately 48 h.

The FAA was then drained and replaced with 70%

ethanol. Cross-sections, ninety microns thick, were

made using a sliding microtome. The sections were

then stained with 0.01% aqueous safranin and

mounted between a glass slide and coverslip. A light

microscope was used to view the sections and

measurements were made using SigmaScan1 digital

image analysis system. The following measurements

were recorded for a representative quarter of a cross

section: number and size of inner and outer resin

canals, and bark thickness and area (Table 2). The

following variables were calculated for all samples:

density of inner, outer and all resin canals as expressed

by the number of resin canals per square millimetre of

bark, and as expressed by the percentage of bark area

occupied by resin canals.

2.4. Statistical analyses

All statistical analyses were conducted using

Statistica for Windows Release 4.5, StatSoft Inc. with

results considered significant at P < 0.05. Pearson

Chi-square analysis was performed on the weevil

attack data for the 2 years following augmentation of

the site with weevils (2000 and 2001) to determine

whether the probability of a tree being attacked by P.

strobi was independent of cross type. Analysis of

variance (ANOVA) and Tukey’s HSD (for unequal N)

multiple comparison test were used to detect

significant differences in resin canal density and size,

by progeny cross type. Arcsine square root transfor-

mation of the variable AOCout (the proportion of bark

area occupied by outer resin canals) was necessary to

meet test assumptions. Non-parametric tests, Kruskal–

Wallis ANOVAs and Kolmogorov–Smirnov two-

sample tests, were used to detect differences between

cross type in variables that did not meet the

assumptions of ANOVA (NMSout, SZin and SZout).

The relationships between resin canal characteristics

in a lateral branch and in a leader from the same tree

was estimated using regression analysis.

Pearson product–moment correlation was used to

explore relationships between weevil attack rates, i.e.

the percentage of trees in each cross that sustained

weevil attacks between 2000 and 2003, and average

1999 bark resin canal characteristics for each cross.

Trees that had been destructively sampled by

removing the leader, were not included in the mean.

Discriminant analysis was used to determine the

reliability of utilizing bark characteristics to predict

whether a given cross would be resistant or susceptible

to weevil attack, or whether it would fall into the

intermediate class. The following variables were

R.I. Alfaro et al. / Forest Ecology and Management 202 (2004) 369–377 373

selected for the analysis: size of inner and outer resin

canals (SZin and SZout), percent of bark area

occupied by inner and outer resin canals (AOCin

and AOCout), number of inner and outer resin canals

per square millimeter of bark (NMSin and NMSout)

and bark thickness (BTHK). Arcsine transformations

of the variables AOCin and AOCout, and log

transformations of NMSin and SZin were necessary

to meet test assumptions. Forward stepwise analysis

was used to select variables for inclusion in the model.

Fig. 2. Annual percent of trees with top-kill resulting from attack by

Pissodes strobi, in interior spruce progeny from controlled crosses of

putative weevil resistant and susceptible parents over a period of 5

years. Trial was screened for resistance by releasing three weevils

onto each tree. The vertical arrow indicates date of release. Codes for

the progeny were: R � R = resistant parent crossed with resistant

parent, R � S = resistant parent crossed with susceptible parent, S �S = susceptible parent crossed with susceptible parent.

3. Results

3.1. Weevil attack rates

The addition of weevils to the site resulted in an

increase in annual attack rate from 5% in the year prior

to weevil augmentation (1999) to 37% of trees

sustaining top kills, in the year following the release

(2000). By 2003, over 50% of the trees in the trial had

sustained top kills. Twenty percent of R � R progeny

and 68 and 86% of R � S and S � S progeny

respectively, were damaged (Fig. 1).

Chi-square analysis indicated that incidence of

attack in 2000 (the year following augmentation) was

associated with cross type (Pearson Chi-square =

534.87, P < 0.00001). Thirteen percent of the progeny

Fig. 1. Percent of trees with top kills due to attack by Pissodes strobi

from the year 2000 until 2003, in progeny from three types of

controlled crosses derived from putative weevil resistant and sus-

ceptible parents. Codes for the progeny were: R � R = resistant

parent crossed with resistant parent, R � S = resistant parent crossed

with susceptible parent, S � S = susceptible parent crossed with

susceptible parent.

from R �R crosses sustained top kills while R� S and

S � S progeny sustained 47 and 68% attack rates

respectively (Fig. 2). However, there was considerable

variation in attack rate within cross type. Attack rates

ranged from one to 37% among R � R crosses, from

14 to 63% among R � S crosses and from 58 to 77%

among S � S crosses (Fig. 3). The highest proportion

of failed attacks, defined as attacks during which

weevils oviposited in the leader, but brood develop-

ment was unsuccessful and the leader was not killed,

occurred in R � R crosses (17%), while failed attacks

were least frequent in S � S crosses (6%). Twelve

percent of the R � S trees also sustained failed attacks.

The overall frequency of weevil attacks declined in

2001 (Fig. 2) but significant differences still occurred

between the cross types (Pearson Chi-square = 226.92,

P < 0.00001). However, the population collapsed to

pre-infestation levels in 2002 and 2003, and attack

rates were low in all crosses.

3.2. Bark resin canal study

Regression analysis to explore the relationship

between the resin canal characteristics in leaders and

laterals (uppermost whorl) revealed a positive and

significant relationship between all variables except

the size of the outer resin canals (SZout) (Table 3). The

R.I. Alfaro et al. / Forest Ecology and Management 202 (2004) 369–377374

Fig. 3. Percent of trees with top-kills due to weevil attack in individual crosses resulting from controlled-crosses of putative weevil resistant and

susceptible parents. Codes for the progeny were: R � R = resistant parent crossed with resistant parent, R � S = resistant parent crossed with

susceptible parent, S � S = susceptible parent crossed with susceptible parent.

relationship between leader and lateral resin canal

density as expressed by number of resin canals per

square millimeter of bark (NMSall) (r2 = 0.41, P <0.00001) is represented by the equation:

NMSallðleaderÞ ¼ 0:4664 þ 0:3514 NMSallðlateralÞ

This equation (and the r2 value) is very similar to the

regression equation reported by Alfaro et al. (1997a,b)

studying a white spruce at a trial near Clearwater, BC.

We concluded that observations on resin canal

densities of lateral samples from the top whorl

adequately represent the resin canal characteristics

of the leader, which is the preferred site for weevil

activity. As a consequence, future selections for

Table 3

Results of regression analyses to determine the relationship between bark

whorl, in interior spruce

Variable Adjusted r2 P-value

SZin 0.16 <0.007

SZout 0.09 <0.06

NMSin 0.19 <0.005

NMSout 0.43 <0.000004

NMSall 0.41 <0.00001

AOCin 0.18 <0.005

AOCout 0.19 <0.003

AOCall 0.24 <0.001

BTHK 0.28 <0.0003

resistant trees using bark resin canal traits can be

conducted by sampling top whorl laterals, thus

avoiding the destruction of the leader.

Kruskal–Wallis ANOVA indicated significant dif-

ferences between cross type in both inner and outer

resin canal size (H(2,N=419) = 11.99 and H(2,N=377) =

13.08, respectively, P < 0.01). The R � R progeny had

larger inner resin canals than did progeny of the other

two cross types (R � S and S � S); however, outer

resin canals from the R � R crosses were only

significantly larger than those of R � S progeny (P <0.005) (Table 4).

Outer resin canal density (expressed as either

NMSout or AOCout) was greatest in progeny from

resin canal characteristics of the leader and laterals from the same

Regression equation

SZin(leader) = 18840 + 1.360SZin(lateral)

Not statistically significant

NMSin(leader) = 0.3620 + 0.2628 NMSin(lateral)

NMSout(leader) = 0.4007 + 0.2893NMSout(lateral)

NMSall(leader) = 0.4664 + 0.3514NMSall(lateral)

AOCin(leader) = 1.792 + 0.05068AOCin(lateral)

AOCout(leader) = 0.4282 + 0.4947AOCout(lateral)

AOCall(leader) = 1.976 + 0.5752AOCall(lateral)

BTHK(leader) = 652.9 + 0.9353BTHK(lateral)

R.I. Alfaro et al. / Forest Ecology and Management 202 (2004) 369–377 375

Table 4

Size of inner (SZin) and outer (SZout) resin canals in lateral branch

cross sections of interior spruce progeny from parents that are

resistant (R) and susceptible (S) to the white pine weevil

Cross type SZin mean (mm2) n SZout mean (mm2) n

R � R 15123.01 a 160 3854.30 a 153

R � S 12804.28 b 199 2992.83 b 181

S � S 13190.32 b 60 3215.09 ab 43

Means within columns followed by the same letter are not sig-

nificantly different; Kolmogorov–Smirnov two sample tests (P <

0.01). R � R = resistant parent crossed with resistant parent, R � S =

resistant parent crossed with susceptible parent, S � S = susceptible

parent crossed with susceptible parent.

two resistant parents (R � R) and smallest in progeny

from two susceptible parents (S � S), while progeny

from resistant parents crossed with susceptible parents

(R � S) had intermediate resin canal densities

(Table 5). Differences in inner resin canal density

among progeny groups were dependent on the variable

used to express density. The number of inner resin

canals per square millimetre of bark (NMSin) was

highest in R � S progeny and lowest in S � S progeny

while NMSin of R � R progeny was not significantly

different from either of the other groups. Inner resin

canal density expressed as the proportion of bark area

occupied by inner resin canals was highest in R � R

progeny and lowest in S � S progeny. Progeny

obtained by crossing resistant and susceptible trees

Table 5

Resin canal densities in lateral branch cross-sections of interior

spruce progeny from parents that are putatively resistant (R) and

susceptible (S) to the white pine weevil

Cross type n Outer resin canals Inner resin canals

NMSouta,b AOCoutc NMSind AOCine

R � R 160 2.00 a 0.73 a 2.52 ab 3.54 a

R � S 199 1.50 b 0.43 b 2.60 b 3.11 b

S � S 60 0.90 c 0.31 c 2.21 a 2.76 b

Means within columns followed by the same letter are not sig-

nificantly different; Tukey’s HSD for unequal N (P > 0.05). R � R =

resistant parent crossed with resistant parent, R � S = resistant

parent crossed with susceptible parent, S � S = susceptible parent

crossed with susceptible parent.a NMSout: number of outer resin canals per square mm of bark.b Differences between cross types were detected by Kolmo-

gorov–Smirnov two sample tests.c AOCout: percentage of bark area occupied by outer resin

canals.d NMSin: number of inner resin canals per square mm of bark.e AOCin: percentage of bark area occupied by inner resin canals.

had AOCin values that were not significantly different

from S � S progeny.

3.3. Correlation between weevil attack rates and

bark resin canal traits

There were significant negative correlations

between the percentage of trees in each cross that

sustained top kills due to weevil attack, and the

average outer resin canal density for the cross. Outer

resin canal density as expressed by both number

(NMSout) and area (AOCout) were significantly (P <0.05) negatively correlated with weevil attack and had

Pearson r values of �0.38 and �0.42, respectively.

There was no significant relationship between inner

resin canal density and weevil attack. However, there

was a significant negative correlation between the size

of inner resin canals (SZin) and weevil attack (Pearson

r = �0.34, P < 0.05).

3.4. Predicting weevil resistance

Stepwise discriminant analysis selected variables

NMSout, SZin, AOCout, and BTHK as providing

significant discrimination among cross types (Wilks’

Lambda 0.43, P < 0.0001) (Table 6). The first dis-

criminant function explained 88% of the discrimina-

tory power of the model while the second discriminant

function was not statistically significant. As expected,

the most clear and significant discrimination was

possible between resistant and susceptible cross types

(Fig. 4). Each of the selected variables contributed

significantly to the function as indicated by the values

of the standardized coefficients (Table 6). The

Table 6

Standardized coefficients for canonical variables, developed during

discriminant analysis to determine whether spruce bark resin canal

characteristics can be utilized to predict resistance to weevil attack

Variable Root 1 Root 2

NMSouta 0.4491 �1.2842

SZinb 0.7381 0.5535

BTHKc �0.7578 �0.1295

AOCoutd 0.5494 0.9775

a NMSout: number of outer resin canals per square mm of bark.b SZin: size of inner resin canals.c BTHK: bark thickness.d AOCout: percentage of bark area occupied by outer resin

canals.

R.I. Alfaro et al. / Forest Ecology and Management 202 (2004) 369–377376

Fig. 4. Scatter plot of discriminant functions depicting discrimina-

tion amongst interior spruce cross types. Codes for the cross types

were: R � R = resistant parent crossed with resistant parent, R � S =

resistant parent crossed with susceptible parent, S � S = susceptible

parent crossed with susceptible parent.

function was characterized by positive coefficients for

NMSout, SZin and AOCout and a negative coefficient

for BTHK. Thus trees with thin bark, large inner resin

canals and dense outer resin canals were more likely to

be from R � R crosses.

4. Discussion and conclusions

With the exceptions of crosses 5 and 9 weevil

resistance was upheld in the progeny derived from two

resistant parents (Fig. 3). Progeny from crosses of two

susceptible parents were most susceptible to weevil

attack. Attack rates in R � S progeny tended to be

intermediate, as would be expected under an additive

genetic model. These results support previous work

showing resistance to be a highly heritable trait (Kiss

and Yanchuk, 1991; King et al., 1997). The relatively

high attack rates in crosses 5 and 9 may have been due

to their common male parent, 1645 from Ontario. The

results suggest that parent 1645 is not as resistant to

weevil attack as the screened parents from Prince

George, BC.

Resin canals were significantly larger and denser in

crosses from R � R parents versus R � S or S � S

parents. These results confirm previous work by

Tomlin (1996) and Alfaro et al. (1997a) who found

that resin canals are an important mechanism in spruce

defense against weevil attack. However, bark resin

canals provide only a first line of defense against P.

strobi and other resistance mechanisms such as

production of traumatic resin and the presence of

sclereid cells have been found to complement resin

canal defenses (Alfaro et al., 2002).

Discriminant analysis provided a useful tool to

combine the various measurements of resin canal

densities into a meaningful discriminant function.

These results suggest that measurements of bark resin

canal characteristics may be useful for indirect

screening of interior spruce families for inclusion in

breeding programs.

The results presented here indicate that interior

spruce resistance is inherited in the F1 progeny from R

� R crosses, thus facilitating the use of genetic

resistance as an element of a P. strobi management

system.

Acknowledgements

The authors wish to acknowledge K. Lewis, G.

Brown and G. Phillips, for their contributions to this

study.

References

Alfaro, R.I., 1989. Stem defects in Sitka spruce induced by Sitka

spruce weevil, Pissodes strobi (Peck). In: Alfaro, R.I., Glover,

S.G. (Eds.), Proceedings of a Meeting of the IUFRO Working

Group on Insects Affecting Reforestation: Biology and Damage

held under the Auspices of the XVIII International Congress of

Entomology. Vancouver, British Columbia, Canada, 3–9 July

1988. Forestry Canada, Victoria, pp. 167–176.

Alfaro, R.I., 1994. The white pine weevil in British Columbia:

biology and damage. In: Alfaro, R.I., Kiss, G., Fraser, R.G.

(Eds.), Proceedings of a Symposium on the White Pine Weevil:

Biology, Damage and Management. Richmond, British Colum-

bia, 19–21 January 1994. FRDA Rep., vol. 226, Can. For. Serv.,

Victoria, BC, pp. 7–22.

Alfaro, R.I., Borden, J.H., Fraser, R.G., Yanchuk, A., 1995. The

white pine weevil in British Columbia: basis for an integrated

pest management system. Forestry Chronicle 71 (1), 66–73.

Alfaro, R.I., He, F., Tomlin, E., Kiss, G., 1997a. White spruce

resistance to white pine weevil related to bark resin canal

density. Can. J. Bot. 75, 568–573.

Alfaro, R.I., Brown, R., Mitchell, K., Polsson, K., McDonald, R.,

1997b. SWAT: a decision support system for spruce weevil

management (pp. 31–41). In: Shore, T.L., McLean, D.A.

(Eds.), Decision Support Systems for Forest Pest Management.

FRDA Rep., vol. 260, Victoria, BC, Canada, 72 pp.

R.I. Alfaro et al. / Forest Ecology and Management 202 (2004) 369–377 377

Alfaro, R.I., Borden, J.H., King, J.N., Tomlin, E.S., McIntosh, R.L.,

Bohlmann, J., 2002. Mechanisms of resistance in conifers

against shoot infesting insects (pp. 101–126). In: Wagner,

M.R., Clancy, K.M., Lieutier, F., Paine, T.D. (Eds.), Mechan-

isms and Deployment of Resistance in Trees to Insects. Kluwer

Academic Press, Dordnecht, The Netherlands, 332 pp.

BC Ministry of Forests, 2002. Forest Practices Branch. Integrated

Silviculture Information System (ISIS). Information current to

January 2002.

Hall, P.M., 1994. Ministry of Forests perspectives on spruce regen-

eration in British Columbia. In: Alfaro, R.I., Kiss, G., Fraser,

R.G. (Eds.), Proceedings of a Symposium on the White Pine

Weevil: Biology, Damage and Management. Richmond, British

Columbia, 19–21 January 1994. FRDA Rep., vol. 226, Can. For.

Serv., Victoria, BC, pp. 1–6.

Humble, L.M., Humphreys, N., Van Sickle, G.A., 1994. Distribution

and hosts of the white pine weevil Pissodes strobi (Peck), in

Canada. In: Alfaro, R.I., Kiss, G., Fraser, R.G. (Eds.), Pro-

ceedings of a Symposium on the White Pine Weevil: Biology,

Damage and Management. Richmond, British Columbia, 19–21

January 1994. FRDA Rep., vol. 226, Can. For. Serv., Victoria,

BC, pp. 68–75.

Jou, S.M., 1971. The resin canal system in Sitka spruce Picea

sitchensis (Bong.) Carr. M.Sc. Thesis, University of Washington,

Seattle, WA, 88 pp.

King, J.N., Yanchuk, A.D., Kiss, G.K., Alfaro, R.I., 1997. Genetic

and phenotypic relationships between weevil (Pissodes strobi)

resistance and height growth in spruce populations of British

Columbia. Can. J. For. Res. 27, 732–739.

Kiss, G.K., Yanchuk, A.D., 1991. Preliminary evaluation of genetic

variation of weevil resistance in interior spruce in British

Columbia. Can. J. For. Res. 21, 230–234.

Kiss, G.K., Yeh, F.C., 1988. Heritability estimates for height for

young interior spruce in British Columbia. Can. J. For. Res. 18,

158–162.

Silver, G.T., 1968. Studies on the Sitka spruce weevil, Pissodes

sitchensis, in British Columbia. Can. Ent. 100, 93–110.

Tomlin, E.S., 1996. Resistance of Sitka spruce to the white pine

weevil. Ph.D. Thesis, Simon Fraser University, Burnaby,

Canada, 231 pp.

Tomlin, E.S., Borden, J.H., 1994. Relationship between leader

morphology and resistance or susceptibility of Sitka spruce to

the white pine weevil. Can. J. For. Res. 24, 810–816.

Tomlin, E.S., Borden, J.H., 1997. Thin bark and high density of outer

resin ducts: Interrelated resistance traits in Sitka spruce against

the white pine weevil (Coleoptera: Circulionidae). J. Econ.

Entomol. 90 (1), 235–239.

Xie, C.-Y., Yanchuk, A.D., 2002. Genetic parameters of height and

diameter of interior spruce in British Columbia. Forest Genet. 9

(1), 1–10.