._ , a . . 1. .w.n..Aa.rya.-u. .Jixfl‘!3¢r J-

i
”Lei!“ III-ii . .
II..- blu . . I [Isl]. I

  

Vt

. . J; ("nu-hunt .- a...n ‘HAFVN II 1*” .115 .ia.$d,.fl.\fl.-wr .5. .lllx; . .m 4/1 it'll-Li...“ «l1 1 .

ABSTRACT

DEVELOPMENTAL AND GENETIC VARIATION IN
THE CORTICAL TERPENES OF SPECIES OF
PINUS AND PICEA

BY
Robert Douglas Westfall
The objectives of this study were: to compare

correlations among cortical monoterpenes in six species

including western white pine (Pinus monticola), eastern

 

white pine (Pinus strobus), limber pine (Pinus flexilis),

 

 

Scotch pine (Pinus sylvestris), red spruce (Picea rubens),

 

 

and white spruce (Picea glauca); to compare and evaluate

 

simple correlation, partial correlation, and factor analy-
ses, using the monoterpene data from the six species; to
determine the within-tree variation in cortical monoterpenes

of white spruce, blue spruce (Picea pungens), white x blue

 

spruce and eastern white pine; to measure the qualitative
and quantitative variation in cortical resin acids of white
spruce, blue spruce, and white x blue spruce.

Consistent relationships in the six species as
shown mainly by the factor analyses were as follows:

positive correlations between a-pinene and camphene, and

Robert Douglas Westfall

among 3-carene, y-terpinene, and terpinolene; negative
correlations between a-pinene and 3-carene, d-pinene and
terpinolene, B-pinene and 3-carene, and B-pinene and ter-
pinolene. Relationships specific to red and white spruces
were a positive correlation between B-pinene and B-phelland-
rene and a negative correlation between B-pinene and limo-
nene. The possible physiological significance of the
correlations is discussed and a tentative biosynthetic
scheme is proposed.

Within-tree variation in monoterpenes was signifi-
cant among three internodes in white spruce saplings. The
greatest variation was in a-pinene, B-pinene, and 3-carene.
However, a- and B-pinene were shown to vary according to
the concentration of 3-carene in the tree. Within-tree
patterns were similar in one-year-old seedling progenies
of white spruce, blue spruce, and white x blue spruce
families. The differences were between the upper and lower
stem positions in the seedlings. In six-year-old eastern
white pine seedlings, variation among three internodes was
again significant in a— and B-pinene and 3-carene, but
a- and B-pinene again varied with the 3—carene concentra—
tion. Factor analyses of the within-tree relationships
among the monoterpenes showed correlations consistent with
those found in between-tree analyses. But in white spruce

and eastern white pine, some of the correlations differed

Robert Douglas Westfall

according to the monoterpene contents of the trees analy—
sed.

The major identified resin acids in seedlings of
white, blue, and white x blue spruce were sandarac0pimarate,
leVOpimarate or palustrate, iSOpimarate, dehydroabietate or
strobate, abietate, and neoabietate. These compounds dif-
fered little between trees or species. Three unidentified
compounds were high in blue spruce, low in white spruce,
and intermediate in the hybrid. They were eluted close
together early in the separation of each sample. Samples
of the upper stem position in the seedlings generally
showed higher amounts of abietic acid and neoabietic acid
than in the lower stem. The range in concentration of the
total resin acid fraction is between 50 and 80 percent of
the oleoresin.

The results of this research have applications to
other areas of study including insect or disease relations,
hormonal balances in trees, and chemical systematics. In
addition, monoterpene and resin acid content may affect
such physical properties of the oleoresin as viscosity and
rate of crystalization which, in turn, may also be involved

in tree resistance to pests.

DEVELOPMENTAL AND GENETIC VARIATION IN
THE CORTICAL TERPENES OF SPECIES OF

PINUS AND PICEA

BY

Robert Douglas Westfall

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Forestry

1972

" c.
«H

ACKNOWLEDGMENTS

I wish to extend my sincere gratitude to Dr. James
w. Hanover (Chairman), who has given a considerable amount
of his time to guiding this study to its completion. I am
also indebted to the other members of the Guidance Commit-
tee--Drs. M. W. Adams, W. L. Myers, and C. J. Pollard--for
their constructive criticism and assistance during the
course of this study.

Appreciation is also due to Mr. R. Comstock for his
technical assistance and to my colleagues--Mr. G. Buchert,
Mr. D. Reicosky, and Mr. B. Rottink--for their comments and
assistance in this study. Finally, I wish to thank my wife,
Sandy, for her encouragement and assistance throughout my
graduate program.

Financial support for this study was provided by

the McIntire-Stennis CooPerative Forestry Research Program.

.ii

LIST OF

LIST OF

TABLE OF CONTENTS

TABLES O O O O O O O O I O O O O O O O O O O

FIGURES O O O O O O O O O I O O O O O O I 0

INTRODUCTION 0 O O O O O O O O O O O O I O O I I O 0

Chapter
I.

II.

Anatomy 0 O O O C I O O O O O O O O O O O O 0

Function of the Oleoresin . . . . . . . . . .
Biosynthesis of Terpenes . . . . . . . . . .
Genetics of Terpenes . . . . . . . . . . . .
Objectives of the Research . . . . . . . . .

CORRELATION ANALYSIS OF CORTICAL MONOTERPENES
IN SPECIES OF PINUS AND PICEA . . . . . . . .

Materials and Methods . . . . . . . . . . . .
Results and Discussion . . . . . . . . . . .

Comparison of the Methods of

Correlation Analysis . . . . . . . . . .
Biochemical Interpretation of

the Results . . . . . . . . . . . . . .

DEVELOPMENTAL VARIATION IN THE CORTICAL
MONOTERPENES OF PICEA GLAUCA, P. PUNGENS,
P. GLAUCA X PUNGENS, AND PINUS STROBUS . . .

 

 

Materials and Methods . . . . . . . . . . . .
Results and Discussion . . . . . . . . . . .

Study 1 - White Spruce Saplings . . . . .
Study 2 - Spruce Progenies . . . . . . . .
Study 3 - Eastern White Pine Seedlings . .
Factor Analyses . . . . . . . . . . . .
Questions Raised from the Results . . . .
Practical Application of the Results . . .

iii

Page

vii

l4
l6

17

19
22

22

31

48

49
54

54
60
64
69
73
75

Chapter

III. BETWEEN-

AND WITHIN-TREE VARIATION

IN THE

CORTICAL RESIN ACIDS OF PICEA GLAUCA,

PUNGENS, AND PICEA GLAUCA X PUNGENS

Materials and Methods
Results and Discussion

PICEA

 

 

CONCLUSIONS AND RECOMMENDATIONS

LITERATURE CITED

iv

Page

76

76
89

95
98

LIST OF TABLES

Table Page

1. The number of trees and geographic origin of
each species sampled . . . . . . . . . . . . . . 20

2. Means (Y) and standard deviations (o) of the
monoterpenes commonly found in six species
sampled C C C C C C C C C C C C C C C C C C C C 23

3. Correlations derived from three types of
analyses showing only those consistent
relationships between certain monoterpenes . . . 24

4. Model for the analysis of variance of within-
tree variation in white spruce saplings . . . . 53

5. Model for the analysis of variance of within—
tree variation in spruce progenies . . . . . . . 53

6. Ranges in monoterpene content of the uppermost
samples from trees in the white spruce, spruce
progenies, and eastern white pine studies . . . 55

7. Summary of the analyses of variance and
additional statistics of the principal mono-
terpenes in white Spruce oleoresin . . . . . . . 56

8. Mean within-tree variation of the principal
monoterpenes in white spruce oleoresin . . . . . 57

9. Summary of the analyses of variance and addi—
tional statistics for the principal monoter-'
penes in seedlings of white spruce, blue
spruce, and white x blue spruce families . . . . 61

10. Mean within-tree variation of the principal
cortical monoterpenes in seedlings of white
spruce, blue spruce, and white x blue spruce
families . . . . . . . . . . . . . . . . . . . . 62

Table Page

11. Summary of the analyses of variance and addi-
tional statistics of the principal monoterpenes
of eastern white pine oleoresin . . . . . . . . 65

12. Mean within-tree variation of the principal
monoterpenes in eastern white pine oleoresin . . 66

13. Major factors in the factor analyses of
within-tree variation in the monoterpenes in
white spruce and eastern white pine oleoresin . 69

14. Major factors in the factor analyses of
within-tree variation in the monoterpenes in
seedlings of white spruce, blue spruce, and
white x blue spruce families . . . . .'. . . . . 72

15. Relative retention times of resin acid methyl
esters and unknown compounds found in white,
blue, and white x blue hybrid spruce cortical
oleoresins . . . . . . . . . . . . . . . . . . . 88

16. Major resin acids in the cortical oleoresin
of three white spruce seedlings . . . . . . . . 90

17. Major resin acids in the cortical oleoresin
of two blue spruce seedlings . . . . . . . . . . 91

18. Major resin acids in the cortical oleoresin
of five white x blue spruce seedlings . . . . . 92

vi

LIST OF FIGURES

Figures Page
1. Monoterpenes of spruce and pine oleoresins . . . 2
2. Resin acids of spruce and pine oleoresins . . . 4
3. Biogenetic pathways of plant terpenes . . . . . 9

4. Proposed biosynthetic pathyway for the
monoterpenes of pine (from Juvonen, 1966) . . . 12

5. PrOposed pathways of monoterpenoid
biosynthesis in Pinus and Picea . . . . . . . . 39

6. Mean within-tree variation of the principal
monoterpenes in white spruce oleoresin
variation in four trees with high concentra—
tions of 3-carene and four trees lacking
3-carene . . . . . . . . . . . . . . . . . . . . 58

7. Mean within-tree variation of the principal
monoterpenes in eastern white pine oleoresin:
variation in three trees with high concentra-
tions of 3-carene and three trees with low
3-carene . . . . . . . . . . . . . . . . . . . . 67

8. Linear calibration curve for levopimaric
acid (LPA) with methyl arachidate (MA) as
internal standard . . . . . . . . . . . . . . . 79

9. Linear calibration curve for dehydroabietic
acid (DHA) with methyl arachidate (MA) as
internal standard . . . . . . . . . . . . . . . 81

10. Chromatogram of the resin acid fraction of

white spruce, blue spruce, and white x blue
spruce cortical oleoresin . . . . . . . . . . . 86

vii

INTRODUCTION

Monoterpenes and diterpene resin acids are the
major components of the cortical oleoresin of coniferous
species. Oleoresin is a viscous liquid produced in espe-
cially large quantities in resin canals of the inner bark
of Pinus and Pigga_species. The predominant monoterpenes
found in pine and spruce oleoresins are illustrated in
Figure 1. Four types are common in the resin: the acyclic
and the mono-, bi-, and tricyclic. Only one monocyclic
monoterpene, myrcene, and one tricyclic monoterpene, cam-
phene, are present in oleoresin, while six monocyclic,
y-terpinene, a-terpinene, B-phellandrene, limonene, ter-
pinolene, and p-cymene, and three bicyclic, a- and B-pinene
and 3-carene are present. In Spruce and pine, the total
monoterpene fraction comprises between 20 and 50 percent
of the cortical oleoresin, depending on the species; resin
acids appear to take up most of the remainder. The resin
acids commonly found in spruce and pine oleoresins are
illustrated in Figure 2. Three major groups are common:
pimaric, abietic, and labdic. Other components of the

oleoresin may be fatty acids and sesqiterpenes.

Figure l. Monoterpenes of spruce and pine oleoresins.

on:— uooén. out: unan—
uzm¢<uun1 uzuzn£<u uzuzzufi uzquLo

6 We 4 .

condu— uooo. ooh:—
mzu£>uuQ «23022:: uzmzosz

a m

on»: coma. uninh— unso—
uzmunzfidzauu uzuzzaufih ”52.2.3-8 232.»:

a w w a

Figure 2. Resin acids of spruce and pine oleoresins.

5

Pimoric Typo Rosin Acids

         
    

  
   

“coon “coon
lcoplmcrlo cold I'm-lccplmcrlo cold
-\ ‘
“coon “coon
Sondcrcocplncrlo cold lecrlo cold

Ablclic Typo Ruin Acids

   

choplmcrlo cold Aclctlo cold Pcluctrlo cold

‘COOH
Ncccblcflo cold Dchydrccblctlo cold

 

Lobdio Typo Rosin Aoidc

” ” coon

Anlloopcllo cold Communlo cold

    

“coon
Strcblo cold

Anatomy

The cortical resin canals of spruce and pine are
long, narrow, circular or elliptical tubes, extending along
the stem axis. Two types of canals (or ducts) are in the
cortex: primary and secondary. The primary ducts develop
schizogenously early in the development of the seedling
(Werker and Fahn, 1969) and the bud (Carolin and Baxter,
1968). They are large and relatively constant in number.
The secondary ducts form schizogenously from divisions of
the primary ducts, and from redivision of themselves, and
usually lead into short shoots or scales on the stem (Werker
and Fahn, 1969). As the stem expands radially, both duct
types increase in diameter. In white spruce (Pigga glauca
(Moench) Voss), there is an exception to the system just
described; bulbous ducts, perhaps traumatically formed,
extend radially outward in the cortex (Thomson and Sifton,
1925).

A layer of thin-walled cells with dense protoplasm,
termed the epithelia, surround the cavity of each duct.
Histochemical (Werker and Fahn, 1968) and electron micro-
scopic studies (Wooding and Northcote, 1965) suggest that
these cells are the site of oleoresin synthesis. One or
more layers of sheath cells enclose the epithelial layer;
neither the function, nor the relation of these cells with
the epithelia, is known. In young stems of eastern white

pine (Pinus strobus L.), these cells appear to contain

higher concentrations of chloroplasts than the surrounding
parenchyma cells, excepting the cells adjacent to the

phelloderm (personal observation).

Function of oleoresin

 

Because monoterpene and resin acid quantities are
quite large in the cortex of spruce and pine, and this may
require a significant expenditure of energy, and because
these coniferous genera have developed specialized struc-
tures to contain these substances, one might assume that
they serve important functions in spruce and pine. But
until recently, few functions had been ascribed to them; now,
the monoterpenes, at least, have several tentative functions
including allelopathy (Asplund, 1968), insect attraction
(Renwick and Vite, 1970), and disease resistance (Shrimpton
and Whitney, 1968).

In addition, monoterpenes are important in human
activities. They are useful in taxonomic studies (Wilkinson,
et al., 1971); they form the basic raw material for turpen-
tine; they are allergenic agents in humans (Pirila,
et al., 1964); and their oxidized forms contribute to

natural air pollution (Rasmussen, 1964).

Biosynthesis of terpenes

The initial steps of biosynthesis appear to be

similar for all the terpenoids. Acetyl-CoA is converted

in a series of steps to mevalonic acid (MVA); this is then
phosphorylated and decarboxylated to form isopentenyl
perphosphate (IPP), the basic isoprene unit. IPP is the
branch point to the synthesis of higher and lower terpenoids.
A diagramatic representation of the biogenetic interrela-
tionships of plant terpenoids is presented in Figure 3.
Although all cell or tissue types are probably incapable

of synthesizing all the terpeniod classes, it is significant
in terms of the regulation of plant physiological processes,
that the terpenoids are related in a branched sequence and
that some of the plant terpenes have known biological func-
tions (eg., gibberellins and carotenoids).

Although a healthy volume of work on the biosynthesis
and metabolism of monoterpenes has been produced over the
past two decades, progress has been slow. Loomis (1967)
and Francis (1971) have reviewed much of this work and
have discussed the major problems associated with it. In
general, these problems center around difficulties both in
incorporating radiocarbon from labelled intermediates into
monoterpene products and in extracting functional cell-free
systems (Francis, 1971). The problems in 14C incorporation
have led some to suggest that the conversion of acetyl-CoA
to monoterpene products proceeds through enzyme complexes
(Banthorpe, et al., 1970; Cori, 1969).

Probably because monoterpene biochemistry has been

so difficult, it has lent itself to the development of

Figure 3. Biogenetic pathways of plant terpenes.

10

 

 

 

 

Photosynthesis Kaurenoic acid—- Gibberellinsl
Kaurenal
Kaurenol Larclenoidsj

 

 

 

Respiration
Kaurene Diterpenesl
\sin\ acids]

Acetyl - Co A Geranylgeranyl
l pyrophosphate
Mevalonic acid Farnesyl
pyrophosphate

 

   
 

lsopentyl pyrophosphate

\\

Dimefhylallyl
pyrophosphale

_ Sesquiterpenes,
Essential oils,
?

Abscisic acid

 

 
   

 

Squalene
Geranyl l

pyrophosphate

 

' Monoterpenes,

 

Essential oils

 

ll

hypothetical constructs. Most of the ones in use today are
variants of a model first developed by Ruzicka (1953).
Figure 4 illustrates one of these (from Juvonen, 1966),
with modifications incorporated from work reported by
Sandermann (1962) and Smith (1964). However, parts of

this scheme are open to question: Banthorpe, et al., (1970)
has reported experimental results which clearly contradict

those of Sandermann. Most of the models, as in Figure 4,

 

suggest that mono- and bicyclic monoterpene synthesis
proceeds through monocyclic carbonium ion intermediates.
In response to the problematic nature of monoterpene bio-
chemical experiments, some writers have resorted to using
correlation analyses of quantitative measurements of cortical
monoterpenes as a complementary means of critically examining
the biosynthetic models (Hanover, 1966a, 1971; Tobolski,
1968; Wilkinson, et al., 1971; Zavarin, 1970). However,
these studies have proved to be both difficult to interpret
and inconclusive.

Catabolism of oleoresin constituents has been
demonstrated in mint (Mentha spp.) (Burbott and Loomis,

1967, 1969), in Scotch pine (Pinus sylvestris L.) (Sukhov,

 

1957), and in maritime pine (P. pinaster Ait.) (Monteoliva,
1970). And metabolic turnover is implicit in the seasonal
fluctuation of the monoterpenes of Scotch pine (Juvonen,

1966; Tobolski, 1968).

12

Figure 4. Proposed biosynthetic pathway for the mono-
terpenes of pine (from Juvonen, 1966).

Goranylpyrophosphal

-2 PO;

ka—
—7

>=C>=

-H : Hm.»
—
/
.44-
Myrcon Ccrbonium-Ion l Ccrbonlum-lon l Corbonium-lon Vlll Sabinon
- H ‘

1H ;

d-Plnon
L ,1 \

it
Q» >=:>=
1

 

[.1
Q cu—

\
Oclmon / TH fi-Phcllcndron
_H -
__—) 7$ -——->
Ccrbonlum-lon V fi-Plnen
\ ._
vOl-l
leoncn Carbonlum-lon ll
\
\
\
\ -H
\ -——) ———>
\
Ccrbonium-lon VI Ccrbonlumolon VII Ccmphon
OH \
\
c-Tcrplnccl 1;“ ' OH”
HO
. -H' 9H
. . Tricyclen Borneo! Ccmphor
Carbonlum-loo Ill
a: Phollcndrcn \ -

 

/

4’ Corn y-Terpmgn Typing." ct-Torplncn

14

Very little is known about the biochemistry of
resin acids. Sandermann (1958) has shown by labelling,
that the abietic and pimaric types are synthesized from IPP
through an intermediate similar to pimarate and that the
abietic types arise from a l, 2 shift of the terminal methyl

group of the intermediate.

Genetics of terpenes

 

As in the physiology of monoterpenes, much remains
to be learned about their genetics. The most definitive
work has been done with the oxygenated monoterpenes of
mint (Hefendehl and Murray, 1972; Murray and Lincoln, 1970).
Less is known about genetic relationships in coniferous
monoterpenes; all of the work done to date has been on the
cortical monoterpenes of spruce and pine. From this work,

a few general observations can be made:

1. Most of the monoterpenes are quantitatively
inherited and all of them are under strong genetic
control (Hanover, 1966; Squillace, 1971). This
fact strongly implies enzyme-mediated synthesis in
the monoterpenes.

2. Three-carene (Hanover, 1966c) and myrcene (Squillace,
1971) are controlled by relatively few genes.
However, in red spruce (Picea rubens Sarg.), varia-

tion in 3-carene appears quantitative because of a

 

15

normal distribution within a high concentration

range (Wilkinson, 1970).'

Nothing is known about the genetics of resin
acids. However, preliminary results, showing differences
between pine species, suggest that resin acids may also
be under genetic control (Joye and Lawrence, 1967).

The cortical oleoresin provides a unique system
for genetic and biochemical study. It is essentially a
closed system, limited in volume and is apparently
influenced internally only by the specialized cells sur-
rounding it. It can be extracted easily with little
contamination from other tissues. And it varies greatly
in internal and external properties: chemistry, vis-
cosity, crystalization, internal pressure, external flow
rates, color, etc.

The major components of oleoresin, the mono-
terpenes and resin acids, are elements of a microhomeo-
static system: because they possibly share a common
biogenetic pathway and share a limited space, they
necessarily limit each other in quantity. An unknown
biochemical-genetic control system effectively limits

the variation of individual monoterpenes and resin acids.

16

Objectives of the research

 

My general objective was to study oleoresin systems

to determine their constituents, how those constituents

vary, and what may cause that variation. My specific

objectives were:

1.

To compare and contrast simple correlation

analysis with partial correlation analysis

and factor analysis, using quantitative

measurements of the monoterpenes from selected

pine and spruce species.

To compare interrelationships among the monoterpenes
of selected pine and spruce species.

To determine the within-tree variation in mono-
terpenes of white spruce, blue spruce, white x blue
spruce, and eastern white pine.

To determine the qualitative and quantitative varia-
tion in resin acids of white spruce, blue spruce,

and white x blue spruce.

CHAPTER I

CORRELATION ANALYSIS OF CORTICAL MONOTERPENES

IN SPECIES OF PINUS AND PICEA

A number of writers have recently used simple
correlations or regression slope coefficients to infer
biosynthetic pathways for the monoterpenes in coniferous
oleoresins (Hanover, 1966a, 1971; Wilkinson, et al., 1971;
and Zavarin, 1970). Zavarin (1970) has reviewed much of
this work. But simple correlations can be helpful in bio-
logical interpretation only if the relationships involved
are clear and unambiguous. Thus, in multivariate situa-
tions, simple correlations may yield confusing and unclear
results; true relationships among a set of variables may
be confounded by spurious correlations between variables
with a common source of variation. This problem can be

illustrated by the following example:

/\

o--—--w

l7

 

18

In the figure shown here, the arrows show causal
relationships between variables A and B between A and C.
And the dotted line indicates the simple correlation
between B and C resulting from their common relationship
with A. Thus, in this situation all the variables shown
would be simply correlated.

The mathematical analysis of monoterpenes should,
at the outset, be classed as a multivariate problem. The
monoterpenes are thought to be synthesized from a common
precursor--either geranyl or neryl prophosphate.‘ In addi-
tion, both 32.21332 rearrangements of the monoterpenes and
structural similarities among them suggest that certain
linear and branched biochemical relationships may be pos-
sible. Further, the oleoresin containing monoterpenes is
essentially a closed system, and a certain amount of inter-
molecular substitution seems inevitable. From the above
considerations, it would seem that techniques more defini-
tive than simple correlation analysis are needed.

Here I present the results of comparisons between
simple correlation, partial correlation, and factor
analyses. All of these techniques were used in analysing
the monoterpenoid compositions of selected Pinus and Picea

species.

 

19

Materials and Methods

The cortical oleoresins of six species-~western
white pine (Pinus monticola Dougl.), eastern white pine

(P. strobus L.), Scotch pine (P. sylvestris L.), limber

 

pine (P. flexilis James), white spruce (Picea glauca

 

(Moench) Voss), and red spruce (P. rubens Sarg.)--were
sampled, and their monoterpene fractions were analysed

quantitatively by gas-liquid chromatography. The number

 

and geographical origin of the trees sampled for each
species are summarized in Table 1. Sampling and analytical
procedures used were previously reported by Tobolski (1968)
and Wilkinson, et al., (1971). Quantitative measurements
of the monoterpenes were expressed as a percent of oleo-
resin. The monoterpenes analysed in all species except
eastern white pine and white spruce included a-pinene,
camphene, B-pinene, myrcene, 3-carene, limonene, B-phel-
landrene, y-terpinene, and terpinolene. Because of their
low concentrations, camphene and y-terpinene were omitted
from the white spruce analyses and y-terpinene was omitted
from the eastern white pine analyses.

For each species, the following matrices were com-
, puted: simple correlation, partial correlation, and factor
vector. The matrices of partial correlations were computed
using the least squares method. To do this, each mono-

terpene was considered separately as the dependent variable

I

20

Table l. The number of trees and geographic origin of
each species sampled.

 

 

 

Species No. trees Geographic origin
Pinus strobus 284 From a New Hampshire

provenance test of
rangewide sources, 27
sources1

Pinus monticola 40 From a Michigan
provenance test of
Idaho sources,
MSFGP-8-70

 

Pinus flexilis 217 From a Michigan
provenance test of
rangewide sources,
MSFGP-11-64

Pinus sylvestris 54 From three diverse
sources in a Michigan
provenance test2
MSFGP-12-61

 

Picea rubens 136 From a provenance
test at Quebec,
Canada, 14 sources

Picea glauca 150 From a Michigan
provenance test of
rangewide sources,
MSFGP-6-63, 16
sources3

 

1From unpublished data supplied by Dr. R. Wilkinson
2See Tobolski (1968)

3See Wilkinson, et a1. (1971)

21

and regressed on the remaining monoterpenes. Partial cor-

relations were estimated using the following formula:

tb.
_ 1

r . . —
Y1°3k"° /Fb +n-K-1
i

 

 

where tbi is the t statistic for the regression
coefficient bi' Fb. is the F statistic for bi’ n is the
number of observations, and K is the number of independent
variables (Gustafson, 1961). This formula has been shown

to be equivalent to the one more commonly used, i.e.:

-a..

= 11
rij.k1... a a
ii jj

where aij’ aii' ajj are the off diagonal and the
two diagonal elements, respectively, of the inverse simple
correlation matrix (Draper and Smith, 1966; Steel and
Torrie, 1960).

A factor analysis was performed on each simple
correlation matrix using a computer routine and subroutines
adapted from Veldman (1967). An image covariance matrix
was first extracted from the simple correlation matrix,
leaving only a matrix of the common variation. For the
purposes of this study I am interested only in that
variation due to those variables included in the study,

i.e., the monoterpenes, and not in the "unique" variation

or that variation caused by unknown factors. Principal

22

axes extracted from the image covariance matrix were then
rotated to orthogonal simple structures by the Varimax

method (Cattell, 1965a; Veldman, 1967).

Results and Discussion

 

Comparison of the methods
of correlation analySis

Table 2 lists the mean and standard deviation for

 

each monoterpene in the species studied. A few, relatively
consistent correlations between certain monoterpenes were
found in all six species sampled. Because I found these

to be helpful in comparing the methods of correlation
analysis, I have listed them in Table 3. Both of these
tables will be used in later discussions on monoterpene
biochemistry.

Depending on the problem at hand, each of the
three techniques used in this study has its advantages and
disadvantages. In fact, the combined methods are helpful
in most situations. In the opposite relationship to that
given in the introduction, where A is instead a linear
combination of B and C, simple correlations are useful.
Here, A would be correlated with B and C, but B and C
would be uncorrelated.

Partial correlations show the relationship between
two variables after the linear effects of other variables

have been eliminated (Kempthorne, 1969). This property

23

 

 

 

 

 

 

 

 

 

 

 

 

 

mH.o mo.o mm.o mm.H mm.o mm.o ~m.o mH.o em.o mm.o mm.o mv.o mamaoawmnme
I I «H.o mm.o NH.o NH.o mo.o no.0 I I -.o ¢H.o cancmaumuu>
Hm.a No.H -.o mm.o mm.a vm.m mh.o hm.o ov.m mm.~ mm.H o>.H occupamaamsmIm
mo.m Hh.m om.o wm.o ah.v mm.m ~o.o no.H vm.m mm.H no.0 mo.H maoaoeflq
ov.H mm.H mm.m ov.mH vm.m mm.m mm.~ mm.H om.h mm.v mm.¢ o>.m mamHMOIm
ov.v Hm.m mm.o ab.o ha.v on.v mm.o vs.o ow.m mm.m vm.o on.H mason»:
hm.n mm.va MH.N ov.m vn.m ov.v Hm.m mm.mH o¢.m mm.mH mo.m o~.vm ocoaflmIm
I I NH.o mm.o mo.o ~m.o H~.o mm.o om.o mm.H na.o mm.o mconmsmu
mH.m om.m mm.H mm.m vo.H >m.~ mm.~ mm.v hm.m HN.NH mh.v hm.m mamaHmIa
I I I I I I I I I I I I I I camonomao mo usoonwm I I I I I I I
6 m b m o m o m 6 m o m occmuouoaoz
mmmmwm mmwmmm manpmo>flmm mwhmmwww manonum mamwwwmoe
modem moowm mpsflm mfinflm madam manna
.pmamfimm mmwommm xfim
may CH venom SHQOEEOU monomnmuoaoe any mo on mCOHDMH>mp pumpamum can Amy memo: .N manna

24

 

 

 

 

 

 

ma.OI nh.o Ah.o mH.o om.oI Nb.o
mm.OI mH.OI NN.OI mm.OI mw.o mm.OI uouomm
mo.o no.0I mH.0I «som.0I «omm.0I mo.o HMflunm
mamHMOIm
o«NN.o mH.0I «om.0I somv.OI osow.OI hN.0I OHQEHm Ocmcwmlm
mo.OI Hm.0I on.o mm.0I mm.OI va.oI
mv.o mm.o mm.OI mv.o ha.o mm.o Houomm
NH.o Ho.oI Ho.o. NH.o mo.o mN.OI Hmflnnmm
mamaoawmnmu
mo.o «omm.oI ««mm.0I «sHN.0I *«SN.0I HN.OI mHmEHm mamCHmIo
mo.o bb.o H>.o ww.o om.0I mm.0I
mv.o mm.OI mm.OI ma.oI 5H.OI ob.oI Houomm
mH.OI «nmm.0I mo.o «omm.oI «omm.OI mo.o Hoapumm
mamumoIm
mo.ou «Ism.on Ismm.ou Isom.on s44m.ou m~.ou mameHm mcoanaua
I mm.o mm.o mo.o m~.o ov.o
I ma.o mN.o mv.o mh.o mm.o “nouomm
I oomm.o «omv.o s«NN.o onam.o «omv.o HMHuHmm
madamEmo
I «omm.o «omm.o oovm.o ooom.o oom¢.o OHQEHm mcwcflmlo
lomaucv Asmaucc lemuav lsnmuav lammucc accuse mammamqm mama
IIlIll IIIIII mo
mosmam msmnfin mHHPmmPHmm mfi.axwdw mDQOHpm MHODHH:OE posuwz mammumuoaoz
mmowm mound madam madam macflm macaw
.mmammumuoaoa campnwo amm3uon madamCOAHMHmn pamumflmaoo
moon» haao maw3onm mommamcm no woman woman Scum pm>wuop mcofiumamnnou .m manna

25

 

 

 

 

 

 

 

 

 

 

I 00.0 00.0 00.0 I 00.0
I 00.0 00.0 00.0 I 00.0 “cause
I IImm.0I HH.0I 00.0I I 00.0 Hanuuma
mamaflmnmul>
I II0~.0 II00.0 IIHm.0 I 0~.0 maasnm mamumoIm
00.0I 00.0 00.0 00.0 00.0I mm.0
mm.0I 00.0I 00.0I 00.0I 00.0 0m.0I 000000
00.0 00.0I 00.0I «Imm.0 00.0I II00.0I Hmnuuma
oCmHoawmnmu
HH.0 «00.0I Imm.0I ImH.0I II~0.0I II00.0I mamsnm mcmcnaIm
mm.0I 00.0 H0.0 ~0.0I 00.0 00.0
mm.0I 00.0 00.0I ~0.0I 00.0 00.0I nouomm
00.0 II00.0 IIHm.0I HH.0 IImm.0I 0H.0 Hmnuumm
occupamaamnmIm
II00.0 II00.0 I00.0I 0H.0 00.0 00.0 mamsnm mamcnaIm
mm.0 00.0 00.0 00.0 00.0 00.0
~m.0I 00.0I 00.0 00.0 00.0 00.0 Hooomc
II00.0I 00.0 II~0.0I 00.0 IINm.0I 0H.0 Hanunmm
mcmcoEHH
IIsm.0I 00.0 00.0 I00.0 00.0I 00.0 massam mamanmIm
l0mauc0 l0maucv Avmncc Asamucv lemmuav onncc mammamq< Mums
MO
mosmam mcwnsu mwuumw>HMm memeHm msmonum MHOOHDGOE bonus: mammumuocoz
mwoflm mmowm macaw madam madam madam
.Aomscnuaooc .m magma

26

.>Hm>fluommmmn .mam>ma usmonmm H can m any um DGMOflmacmwm «I .o

.aowumamnnoo o>aummmc oumowpaw Guam mnemommo mo mmcflpmoa oHHn3 .poumHmnnoo ham>wuflmom

ma “and mammnmuocos map .amam dawn on» as: mmcflpmoH mo Mama 0 MH

.mflxm mean can nuflz

mammnmuoaoe Hmzoa 0:» mo mawpmoa msu ma Hones: HmBOH on» oHH£3 .mwxm as SDAB mammnwuoaoe

woman may no mafipmoH on» mucmmmnmmn Hogans momma one

.mflxm .HOHOMM M USN mQHQMHHM>

asapfl>wpaw amm3umn Ammaflpmoa HOV mcowumamnnoo mum who: pmuammmnm mmHDmHm one”

 

 

 

 

 

 

I 00.0 00.0 mh.0 I mm.0
I «0.0 0m.0 Nv.0 I No.0 Houomm
I soam.o «omv.o oomm.o I mm.o HMHHHMQ
mamaocflmumu
I II00.0 «I00.0 II00.0 I II00.0 mHmEAm mamansumuI>
0H.0 00.0 00.0 m0.0 00.0I mm.0
mm.0 00.0 H0.0 00.0 00.0I Nh.0 Houomm
osmm.o sohh.o «ovm.o ssmm.o «ovm.o somm.o Hmwphmm
mamaoaflmumu
sovm.o soom.o «owm.o «swm.o oovm.o osmh.o OHQEHm OGOHMOIM
l0mancc Ammancv lemuac Asamncc Asmmncv onucv mammamcs cams
IIIIIII. mo
flwwmwm memmm manumm>amm mwaamem msmonam mHOUHDGOE Genus: mammnmuocoz
mwofim mwoflm. macaw macaw mSCAm msawm

 

 

.Acmscnucooc .m magma

27

can lead to confusing results if partial correlations are
used alone. For example, when three variables, say, A,
B, and C, are all highly related to each other, the
partial correlations of each pair would be zero since
the variation in the pair had been accounted by the third
variable. The effect is illustrated in Table 3: in red
spruce, the simple correlations between a-pinene and
3-carene, a-pinene and terpinolene, and 3-carene and
terpinolene are all statistically significant; factor
analysis confirms these correlations. But the partial
correlation coefficient of a-pinene and terpinolene is
zero; the relationship between the two has been accounted
for by 3-carene. Therefore, given n variables, the rela-
tionships among them should be studied by comparing the
simple correlations with the partial correlation matrices
of size three through n.

The most powerful technique of the three used is
factor analysis. This technique converts a matrix of n
correlated variables to a matrix of k uncorrelated factors,
where k < n and the k factors account for most of the
variation in the image covariance. The factor matrix is a
set of k factor vectors of order n where the n variables
are variously "correlated" with the vectors. Those vari-
ables highly correlated (or "loaded") with the same factor
may be considered to be highly correlated with each other--

and vice versa. (For a rather comprehensive,

28

ordinary-language discussion of factor analysis, see
Cattell, 1965a, b.)

The factor analytic technique has the advantage
of enabling one to look at all the variation originating
from the data matrix: a single variable may correlate
with more than one factor and so may be associated with
more than one cluster of variables. Since each factor
accounts for a particular proportion of the total varia-
tion and each variable loads in varying degrees with the
factors, one can determine the relative importance of a
given variable's correlations with other variables.

But there are some problems and deficiencies
associated with factor analysis; the two most contro-
versial concern communalities and the rotation of axes.
Communalities are those values, inserted into the diagonal
elements of the correlation matrix, which reflect the true
amount of common variation in the matrix. So the problem
here is to arrive at a good estimate of the communalities
(Cattell, 1965a).

In this study, the image covariance method used
gives communalities which are recognized to be the minimum
estimate (Cattell, 1965a). And the method has the advan-
tage of mathematical elegance (Veldman, 1967). So because

of questionable additional advantages of alternative

29

methods of computing communalities and the given advan-
tages of the image covariance method, the latter method
was judged adequate for the purposes of this paper.

The rotation of factor axes is a more difficult
problem. If the factor axes are restricted to orthogo-
nality, simple structure often is not completely reached
(Cattell, 1965a). In a pure, orthogonal simple structure,
each variable would have only one non-zero loading in the
factor matrix. However, in biological work, this condi-
tion should not really be expected. In working with the
monoterpenes, for example, one would expect that, because
of their possible common biosynthetic points of origin,
those monoterpenes clustered on more than one factor would
be correlated with each other.

Evidence for nonorthogonality (or obliqueness) of
factor axes is seen when a variable loads on more than
one axis. In the three variable illustration given above,
variable A would be expected to load rather heavily on
both factor axes. And, in the data for western white pine
in Table 3, though terpinolene is significantly correlated
with both 3-carene (loaded 0.53 and 0.72, respectively)
and y-terpinene (loaded 0.39 and 0.62), the two pairs of
loadings occur on separate factors.

But complete oblique simple structure rotation,

though more desirable than rotation to orthogonal simple

30

structure, cannot be programmed on the computer at the
present time. And manual rotation of more than three
factors is quite tedious. Cattell and Foster (1963) have
proposed a program for visual rotation of axes on an
oscilloscope-computer combination. However, since that
program was not available to this laboratory, the Varimax
routine used in this study was then considered to suffi-
ciently approximate simple structure.

A further problem with factor analysis, as with
other multivariate techniques, is that it is quite
susceptible to violations of those basic conditions
necessary for statistical validity (Steel and Torrie,
1960). Univariate methods of analysis, such as simple
correlations are considered quite "robust," in that
violations of these conditions do not greatly alter the
essential validity of the relationships. However, the
interpretation of multivariate analyses may be seriously
questioned if these assumptions are in any way unfulfilled.

A final difficulty is that the factor loadings and
the variance component associated with each factor cannot
be tested for statistical significance. Thus, choosing a
minimum level for non-zero loadings is most difficult.
Intuitively, this minimum level would be expected to vary
according to the number of observations (n) in the

analysis: Cattell and Foster (1963) have suggested a

 

31

general value of I0.10 to be raised to 10.15 when n < 70

and lowered to 10.06 when n > 300.

Biochemical interpretation
cf the results

 

Of the correlations listed in Table 3, the white
Spruce data show the greatest divergence from patterns
established in other species. This probably reflects the
great amount of between-source variation in the population
sample. Common variation in the correlation matrix was
quite low; consequently, the factor loadings were also
low. If the between-source variation had been smaller,
common variation might have been greater and the correla-
tions stronger. In Chapter II, a factor analysis of the
monoterpene data of white spruce within-tree variation
supports this explanation: the analysis showed a nega-
tive correlation between a-pinene and 3-carene; a
positive one between a-pinene and camphene; and positive
correlations between 3-carene, y-terpinene, and terpinolene.
These results will be discussed in greater detail in
Chapter II.

In his review, Zavarin (1970) presented a bio-
synthetic model for monoterpenes, modified from one first
develOped by Ruzicka (1953). A model similar to Zavarin's,
proposed by Juvonen (1966), is presented in Figure 4.

Essentially, these models prOpose that the mono- and

 

32

dicyclic monoterpenes originate through hydride Shifts of
monocyclic carbonium ion intermediates, and that those
monoterpenes found in the "essential oil" or the oleoresin
are the end products of the biosynthetic system.

Using criteria established in his paper, Zavarin
biochemically interpreted certain correlations between
monoterpenes found in Pinaceae species. These criteria

and assumptions were: (1) Correlations between substances

 

are caused by the linkage of biosynthetic reaction sequences
and not by gene linkage of spurious correlations; (2)
Compounds correlated either positively or negatively are
closer biosynthetically than those which are uncorrelated;
(3) Positively correlated compounds are biosynthetically
closer than those negatively correlated. But criteria
such as these have no empirical base: there exist no
correspondence rules between statistical correlations,
however strong or consistent, and biochemical relations.
Thus biochemical interpretation of statistical correlations
would seem to be purely speculative. So in order to
develop some tentative correspondence rules, I will now
discuss the following factors which may influence both
individual monoterpene levels and correlations among them:

1. The closed system nature of the oleoresin.

2. Gene linkage.

3. Gene regulation.

33

4. Enzyme feedback mechanisms.

5. Biochemical reaction sequences.

6. Genetic segregation of Single and multiple enzyme
forms.

First, the oleoresin itself is the product of a
closed system: a tube of fixed diameter and length.
Whatever goes into the system must necessarily be limited
in quantity and prOportion. Therefore, it follows that
the monoterpenes would not only compete amongst each other
for the space available, but also with other components of
the oleoresin: sesqui- and diterpenes, fatty acids, and
other lipids. But there are two reasons why this does not
seem to be a regulating factor in this study: (1) Genetic
control over the total monoterpene fraction seems relatively
high. For example, Hanover (1971) found moderate heredi-
tability estimates for total monoterpenes in western white
pine. Also, the total monoterpene portion does not vary
much: in white and blue spruce, monoterpenes comprise
between 20 and 30 percent of the oleoresin, while in
eastern white pine, the percentage range is 25 and 40
percent. Most of this variation follows that of the major
monoterpene components, usually a- and 8-pinene and 3-carene.
(2) Only the correlations derived from common variation are
used in this study. Thus, effects due to variations in

non-monoterpenes are greatly reduced.

34

Second, gene linkage could be a factor in regulating
monoterpene levels, although this is improbable: Zavarin
(1970) came to the same conclusion in his review (cf.,
criterion (1) above). If gene linkage were important, one
would expect between-tree correlations to randomize within
the tree: the results presented in Chapter II show they
do not. One might also expect linkage to be detected in

parent-progeny analyses. Again, it has not at least in

 

western white pine (Hanover, 1971) and slash pine (Squillace,
1971). But because spruces and pines are closely related
cytogenetically, one would expect the overall biochemical-
genetic system to be similar in all of the species in the
two genera.

Third, gene regulation may affect monoterpene
levels. This mechanism is apparent in monoterpene
increases found in the developing seedlings of maritime
pine, (Pinus pinaster Ait.), (Monteoliva, 1970), although
enzyme activation could also cause such increases.
Similarly, this mechanism may be responsible for tissue
age differences found in eastern white pine and white
spruce stems (cf., Chapter II). And it may well be
responsible for variations (and covariations) in individual
monoterpene levels. But it is difficult to speculate
whether individual or group regulation is most active: if

a recent hypothesis on gene regulation (Britten and

35

Davidson, 1969) proves correct, then group regulation may
be the dominant mechanism.

Fourth, enzyme feedback mechanisms, possibly in
combination with complex enzymes and isozymic variation
(Datta, 1969), quite probably cause a portion of the
variation found in the monoterpenes. Most probably, these
contribute to correlations amongst the major constituents
of the monoterpene fraction and to correlations specific
to groups of taxa (generic, specific, and intraspecific).

Fifth, the biochemical reaction sequences them-
selves may influence correlations; though, as pointed out
above, their interpretation in statistical terms is very
difficult. A number of reaction types are apparent: irre-
versible synthesis from a common precursor, sequential
irreversible reaction sequences, and reversible reaction
sequences. If monoterpenes arose by irreversible reactions
from a common precursor (suggested in parts of Figure 4) but
no enzyme regulation was present, then since the precur-
sor(s) would be absent, covariation among the monoterpenes
would be either near zero or would vary negatively ac-
cording to the relative efficiency of the enzymes present.
And if synthesis arose through sequential, irreversible
reactions, again with no regulation, the results would
be the same as with the type just discussed. The oxy-

genated monoterpenes of mint, the carvone and the pulegone

series, apparently are related in a sequential system

 

36

(Loomis, 1967; Hefendehl and Murray, 1972), but with
enzyme regulation. Members in each series appear to be
positively correlated. But the series themselves may each
originate from a branch point: limonene, ring-substituted
at C3, may begin the carvone series, while terpinolene,
ring-substituted at C3, may begin the pulegone series. The
two series are negatively correlated: single gene pair
apparently controls the presence and absence of each of
the two series (Murray, 1960).

Reversible reactions, without regulation, would
probably behave like sequential irreversible ones with
regulation; given a Single enzyme acting in the reaction,
the precursor and product would be positively correlated.
But if multiple enzymes were present in the reaction, and
if these varied either between or within trees, the
correlations would probably be negative.

The final factor, genetic segregation of single
and multiple enzyme forms, interacts with others given
above: gene regulation, enzyme regulation, and reaction
sequences. Thus it would act as an overlay on the effects
of these other factors, either directly causing between-
tree variations or modifying within-tree variations.

Although gene control of coniferous monoterpenes
is not as Simple and clear cut as that in the mint oxy-

genated types, it may involve relatively few alleles in

3-carene (Hanover, 1966) and myrcene (Squillace, 1971).

37

But in the cortical monoterpenes as a whole, quantitative
and within-tree variation almost certainly indicates
multiallelic and isozymic variation.

The point of the above discussion is that for the
monoterpene systems, cases where there is a simple presence
or absence of a substance do not exist; variation even in
multimodal distributions is quantitative. And the factors
just discussed would all contribute to quantitative varia-
tion.

Now, while keeping these causal elements in mind,
how may statistical correlations be interpreted? First,
correlations consistent over many species Should be con-
sidered biochemically significant regardless of their
strength. Next, simply correlated pairs are meaningful,
while uncorrelated ones are not. However, simple cor-
relations which drop to zero in a factor analysis may mean
that a common element has been located; so this rule should
be applied with caution. Finally, negative correlations
are no less biosynthetically meaningful than are positive
ones. Obviously, this postulate conflicts with Zavarin's
third, given above. However, I feel the postulate is
justifiable for two reasons: (1) Most of the statistically
significant correlations found in this study are negative.
(2) And a negative correlation is consistent with three

biochemical-genetic events, isozymic variation in the

activity of two enzymes acting on a common precursor,

 

38

differences in the feedback inhibition of those two enzymes,
and isozymic variation in enzymes of reversible reactions.
On the other hand, positive relationships suggest either
one of two separate events. First, the two elements of

the correlation may be linearly related in a synthetic
sequence. Or they may be synthesized in initially uncor-
related parallel sequences, but have become correlated
through a feedback mechanism.

In the overall view of the species analysed here,
two relationships are apparent. One of these is the
consistency in the negative correlations between the
bicyclic monoterpenes, a-pinene and 3-carene, and B-pinene
and 3-carene, presented in Table 3. The other relationship
is the variability in strength and direction of the rela-
tionships between the mono- and the dicyclic and even among
the monocyclic monoterpenes, implying that the two groups
are not closely related biosynthetically: they may orig-
inate from parallel paths. Obviously, the strong positive
correlations amongst 3-carene, a-terpinene, and terpinolene
are exceptions to this observation; these will be dealt
with later. But my point is, the data imply that, from a
point following the condensation of IPP and DMAPP, the
individual monoterpenes arise through unlinked, concerted
paths.

My hypothesis for the biosynthetiprathways for

monoterpenes is illustrated in Figure 5. Although this

 

39

Figure 5. Proposed pathways of monoterpenoid biosynthesis
in Pinus and Picea.

PP PP
+ mu»
#1

 

 

41

hypothesis is quite different from the one usually pre-
sented, there is evidence supporting it. Gascoigne (1958),
for example, has argued that direct conversion of acyclic
to bicyclic is more probable on thermodynamic grounds:
the conversion of acylic to monocyclic and acyclic to
dicyclic monoterpenes involves a negative free energy
change, while the conversion of mono— to dicyclic is
positive. And a positive free energy change would require
a high energy compound to drive the reaction. In the
terpenoids, the only cyclization reaction described at
present supports this argument: the enzyme, squalene-
oxidocyclase, catalyses the concerted, nonstop cyclization
of 3-hydroxy squalene to lanosterol (Heftmann, 1965).

After a similar review of the literature, Monteoliva
(1970) developed a hypothesis much like the one presented
here. He also proposed that each monoterpene originates
from the pyrophOSphate of a distinct acyclic precursor.
In this scheme, the double bonds of the condensate from
IPP and DMAPP are first rearranged to a given precursor
and then the corresponding end product is formed, implying
that a series of isomerases and cyclases would be active
in the system. Portions of the scheme in Figure 5 have
been adopted from Monteoliva's hypothesis.

Further, recent evidence suggests that nerol or

neryl pyrophosphate may be the basic precursors of the

monoterpenes (Cori, 1969). Synthesis of neryl-PP may

 

42

occur through one or both of two routes: (1) IPP and DMAPP
condense to form geranyl-PP which then isomerizes to neryl-
PP or (2) IPP and DMAPP condense directly to neryl-PP.

Because total monoterpene levels are relatively
constant within a Species, implying genetic control,
regulation of these levels should take place prior to
monoterpenoid synthesis. Then control of monoterpenoid
levels might center on the direct condensation reaction
(Figure 5, Arrow A) or on the isomerization; and control
of overall sesqui- and diterpenoid levels may occur near
the condensation reactions of IPP and geranyl or farnesyl
pyroPhOSphate (Figure 5, Arrow B).

A feature of the scheme prOposed here is that it
offers greater possibility for regulation of individual
monoterpene levels, either at the point of isomerization
or at cyclization. Much of the covariation among groups
of monoterpenes is negative: a decrease in the levels of
one group is compensated for by an increase in the other
group. Thus, 3-carene may inhibit the synthesis of
a-pinene from neryl-PP and B-pinene from 7-methy1-3-
methylene-6-octen-1-y1-PP(I); or 3-carene synthesis
itself may, in turn, be inhibited by either a- or B-pinene.
Regulation of the monoterpenes in Pinus and Picea is
probably not at all simple: results from this study and
from Tobolski's (1968) show that correlations within a

species vary according to the sample used. In the

43

developmental study of eastern white pine in Chapter II,
the factor analyses of both the whole data set and the
high 3-carene subset Showed negative correlations between
3-carene and a-pinene, and 3-carene and B-pinene; but in
the subset with low or absent 3-carene, these negative
correlations are replaced by a negative one between a- and
B-pinene. The action of isozymes is suggested here.

Three omissions in the scheme in Figure 5 may seem
evident. First, the mono- and bicyclic monoterpenes may
each arise from a "core" precursor: the monocyclic ones
from a monocyclic intermediate and the bicyclic ones from
a bicyclic intermediate. But although the negative rela-
tionships among the bicyclic monoterpenes support this
alternative, both the general absence of a negative
relationship between the monocyclic monoterpenes and the
bicyclic monoterpenes and the inconsistent relationships
among the monocyclic monoterpenes negate it. Next,
reversible reactions have been omitted from consideration.
However, again because inconsistent correlations are present
in the data, this alternative seems unlikely. The third
omission concerns the synthetic mechanism for a-pinene
implied in Figure 5. Although this mechanism is con-
sistent with evidence presented by Sandermann (1962), it
contrasts with the recent work of Banthorpe, et al., (1970)
on thujane derivatives from Specifically labelled

mevalonate. If these results are extended to a-pinene,

 

44

they suggest that the secondary ring is formed at C-2 of
neryl-PP, with a shift of the double bond to C-3 on the
acyclic precursor. This mechanism could similarly be
applied to B-pinene (with neryl-PP as the precursor),
except that the double bond shift would come to rest at
the 3-methyl group. Therefore, both a- and B-pinene
would result from neryl-PP. Cori and his group have
found evidence in labelling experiments supporting this
postulate (Beytia, Valenzuela, and Cori, 1969; Cori,
1969). However, my correlation results neither support
nor conflict with either alternative.

Some of the consistent correlations found do not
fit into the system presented in Figure 5: it implies
negative relationships among the end products. The
a-pinene-camphene relationship, however, is positive in
the species studied here. But in the foliar oils and the
oleoresin of Scotch Pine (Juvonen, 1966; Tobolski, 1968)
and Pigga_spp. (von Schantz and Juvonen, 1966), the rela-
tionships are variable: in Scotch Pine, both a-pinene
and camphene levels are high in the foliage, whereas in
the cortex, a-pinene is high and camphene is low. In the
spruce, the relationship in the foliage varies according
to the species; while in the cortex, a-pinene is again
high and camphene low. Further, note in Table 1, that
in western and eastern white pine there is a dispr0por-

tionate increase in the level of camphene relative to

45

that of a-pinene. Relatively high levels of camphene are
characteristic of eastern white pine cortical oleoresin
(Hanover, in preparation). These results suggest that the
relationship is more than just a simple one; a regulated,
branched relationship is implied. One reasonable explana-
tion for the positive correlation could be in_yitrg
rearrangements: under prOper conditions, a-pinene will
convert to camphene (Finley, 1969). Then it is possible
that a-pinene and camphene are both initially synthesized
through the normal biosynthetic pathways, but then a-pinene
is partially converted i2_yiE£g to camphene. The dashed
arrow from a-pinene to camphene in Figure 5 indicates this
possible synthetic route.

Other anomalies in the consistent correlations are
the positive ones among 3-carene, y-terpinene, and terpin—
olene. But the explanation for these correlations may be
the same as that for a-pinene and camphene: their cause
may originate from a combination of 12.2122 and i2_yiE£g
events. In the xylem resin of BishOp pine (Pinus muricata
D. Don), for example, 3-carene and terpinolene are nega-
tively correlated: when 3-carene is high, terpinolene is
low and vice-versa (Forde and Blight, 1964). And in the
cortical monoterpenes of the species in this study, the
ratio of the quantity of 3-carene to that of terpinolene
is usually at least ten to one (cf., Table 1). The only

exception to this rule is in trees lacking 3-carene:

46

terpinolene is still present, though in low amounts. Thus
levels of terpinolene are generally low in pines and
spruces, as are levels of y-terpinene; individuals with
high amounts of either in cortical oleoresins are rare.
Further, because the cyclopropane secondary ring of
3-carene is relatively less stable, it is more prone to
cleavage. Accordingly, Simonsen (1920) reported that,
in hydrochloric acid solution, 3-carene degraded to
limonene and sylvestrene (l,8(9)-mfmenthadiene). But
in nature, this degradation is unlikely since sylvestrene
apparently does not naturally occur. Thus, what could
occur is either the enzymatic degradation of 3-carene to
y-terpinene and terpinolene or the incomplete synthesis
of 3-carene, leaving y-terpinene and terpinolene as by-
products. To illustrate the importance of 3-carene in
the variation of terpinolene, a factor analysis of white
spruce lacking 3-carene, shows the percent common varia-
tion in terpinolene is 50, while an analysis of Spruce
high in 3-carene shows the percent common variation for
terpinolene increases to 98 percent (cf., Chapter II).
The dashed arrow in Figure 5, from 3-carene to y-terpinene
and terpinolene suggest the possible secondary synthetic-
catabolic paths.

Although correlations between B-pinene and
limonene and B-pinene and B-phellandrene are not consist-

ent across all the species listed in Table 3, those

47

correlations are consistent in the spruces: the correla-
tion between B-pinene and limonene is negative, while that
between B-pinene and B-phellandrene is positive. But in
the white spruce within-tree sample in Chapter II, a factor
analysis of only those trees low in limonene and high
B-pinene showed a positive relationship between B-pinene
and limonene; the negative relationship was confirmed in
the trees with low B-pinene and high limonene. The
B-pinene-B-phellandrene relationship was confirmed in both
analyses. In addition, B-pinene and B-phellandrene are
positively correlated in blue spruce (Pigga’pungens Engelm.)
and in the hybrid between white and blue spruce (cf.,
Chapter II). Thus variation between B-pinene and limonene
seems to occur in white Spruce partially through gene
segregation and enzyme regulation. But in the B-pinene-
8-phellandrene relationship, enzyme regulation Specific

to the spruces may cause the correlation.

CHAPTER II

DEVELOPMENTAL VARIATION IN THE CORTICAL

MONOTERPENES OF PICEA GLAUCA, E.

 

PUNGENS, B. GLAUCA X PUNGENS,

AND PINUS STROBUS

 

Within-tree variation of cortical monoterpenes
has been described in pine (Hanover, 1966a; Tobolski,
1968) and fir (Zavarin, 1968). Since the overall objec-
tives of these studies were chemotaxonomic or genetic
oriented, these authors concluded that for their purposes,
the variation was small, and could be made less signifi-
cant by sampling equal-aged tissue. In addition, Hanover
(1971) has described developmental variation in B-pinene

levels between western white pine (Pinus monticola Dougl.)

 

parents and their seedling progeny: B-pinene levels were
higher in the progeny.

During the course of spruce monoterpene studies in
this laboratory, we noted higher 3-carene levels in the
cortex of the upper portion of Seedlings than in the lower.
Thus a series of studies were undertaken to examine this
variation in greater detail: one of selected white spruce

saplings, another of white, blue, and white X blue hybrid

48

49

Spruce seedling progenies, and a third of selected eastern
white pine seedlings. Here I present the results of these

studies.

Materials and Methods

 

In the study of selected white spruce (Picea glauca

 

(Moench) Voss) saplings, eight trees of known cortical
monoterpene content were selected from a Michigan prove-
nance test (MSFGP-6-63) (see Wilkinson, et_al., 1971).
Four trees were selected for high levels of 3-carene; the
remainder lacked 3-carene. The group as a whole was
selected to obtain as much genetic variation in a-pinene,
B-pinene, and myrcene as possible.

In each tree, oleoresin samples were taken from
the upper and lower portions in each of three consecutive
internodes, beginning with the most recently formed inter-
node. Sampling was done in July and August, 1971.

In the study of spruce full-sib progeny, controlled

pollinations of white spruce and blue spruce (Picea pungens

 

Engelm.) parents were made in May, 1969; in the hybrid
cross only blue spruce pollen parents were used. Pollina-
tion and seed collection, and preparation procedures have
been previously reported (Hanover and Wilkinson, 1969).

The stratified seeds were sown in pots, at the rate of ten,
evenly-spaced seeds per pot, in the greenhouse. New

germinants were grown for 18 months under continuous light

50

at 25-27°C. Then the seedlings were transferred, in the
pots, out to nursery beds.

Four white spruce, three blue spruce, and five
hybrid full-sib families were randomly selected for sampl-
ing. In the white and blue spruce families, three seed-
lings per family were sampled; but in two hybrid families,
two seedlings were sampled in one family, while only one
seedling was sampled in the other; three trees were sampled
in each of the remaining three hybrid families. Oleoresin
samples were taken in June, August, and September, 1971,
from the upper and lower middle portions of each seedling.
Most of the seedlings grew continuously for the entire 18
month growing period, and thus could be considered having
only one internode. Those that did not grow continuously,
ceased growth shortly, and then continued growing for the
remainder of the period. Seedlings in some of the hybrid
families were resampled either in November, 1971, or in
February, 1972, to check monoterpene content for the resin
acid study in Chapter III.

For the study of eastern white pine seedlings,
six Six-year-old potted seedlings of known monoterpene
content were selected from a group of southern Appalachian
sources: three with high and three with low 3-carene
levels. These seedlings had been first grown from seed
sown in nursery beds and then transplanted into pots after

two years. Both selection procedures for the remaining

51

monoterpenes and procedures for oleoresin sampling were
equivalent to those in the white spruce sapling study.
Sampling was done in December, 1971.

Monoterpene sampling and analytical procedures
were similar to those outlined by Wilkinson, gt_gl. (1971),
but with the following exceptions: (l) Oleoresin samples
of 2—5 ul were drawn into 20 pl calibrated micropipettes.
(2) These samples were diluted with spectra-grade acetons
to a ratio of 15:1, acetone to resin, for the 5 01 samples,
and to 16:1 for smaller samples. These particular dilu-
tions were made to simplify subsequent calculation proce-
dures. (3) A 4 ul aliquot of the resin-acetone solution
was injected into the gas chromatograph. (4) Peak areas
were measured electronically on a Hewlett-Packard 3370A
Digital Integrator.

Monoterpene quantities were expressed as a percent
of the total volume of oleoresin sampled. Nine monoter-
penes were consistently observed: a-pinene, camphene,
B-pinene, myrcene, 3-carene, limonene, B-phellandrene,
yeterpinene, and terpinolene.

An analysis of variance was performed, in each of
the three studies for a-pinene, B-pinene, and 3-carene,
using an alternative least squares solution described by
Draper and Smith (1966). This solution is especially
suited for unbalanced designs as in the study of spruce

progenies. The three monoterpenes analysed were selected

52

for two reasons: (1) because they accounted for a majority
of the variations in all the studies and (2) because all
the remaining monoterpenes were found to be strongly cor-
related with at least one of these three.

The experimental design for the white spruce sap-
ling study is given in Table 4. In generating the expected
mean squares, all the effects in the design were considered
to be fixed. The design for the white pine study is the
same as that given in Table 4, except that the degrees of
freedom are 5, 2, 10, l, 5, 2, and 10 for the tree, inter-
node, tree X internode, position, tree X position, inter—
node X position, and error effects, respectively. The
design for the study of Spruce progenies is presented in
Table 5; the family and trees-within-families effects were
considered to be random, whereas the position effect was
fixed.

Intraclass correlation coefficients (Kemp-
thorne, 1969) and coefficents of multiple determination
(R2) (Draper and Smith, 1966) were calculated for each of
the tree selected monoterpenes in each study.

Factor analyses were made on the monoterpene cor-
relation matrices in each study. In the white spruce
sapling study, the following data sets were analysed: the
complete set, the subset with high 3-carene, the subset
lacking 3-carene, and each of the high and low limonene

subsets. And in the spruce progeny, these subsets were

Table 4. Model for
variation

53

the analysis of variance of within-tree
in white spruce saplings.

 

 

 

Source df Expected mean squares
2 . 2
Tree 7 O TIP + 190T
Internode 2 02 + tpo2
TIP I
Tree X Internode 14 02 + p02
TIP TI
Position 1 02 + tic2
TIP P
Tree X Position 7 02 + i02
TIP TP
Internode X Position 2 02 + to2
TIP IP
Error 14 02
TIP

 

Table 5. Model for
variation

the analysis of variance of within—tree
in spruce progenies.

 

 

 

Source df Expected mean squares

. 2 2 2
Family 11 OF(TP) + pOF(T) + kpoF

. . . 2 2
Trees Within family 21 OF(TP) + p°F(T)

. . 2 2 2
POSition l GF(TP) + kOFP + fkoP
Famil X Position 11 02 + kc2 1

Y F(TP) FF
2
Error 21 °F(TP)

 

1k is an average of the "trees" coefficent, ad-
justed for the unequal sample size in this experiment (Sokal

and Rohlf, 1969).

54

analysed: the white spruce families, the blue spruce

families, and the hybrid families. In the eastern white
pine, the complete set and the high and low 3-carene sub-
sets were analysed. Analytical procedures were the same

as those in Chapter I.

Results and Discussion

 

Ranges in concentration for the nine monoterpenes
in uppermost resin samples in each of the three studies
are listed in Table 6. The major monoterpenes in the
cortical oleoresins of all the species studied were a- and
B-pinene, 3-carene, and limonene. However, in terms of
within-tree variation, those of a-pinene, B-pinene, and

3-carene were most significant.

Study 1 - White Spruce Saplings

 

A summary of the analyses of variance for the
three major monoterpenes in the white spruce saplings is
presented in Table 7. In each analySis, the model ex-
plained a high prOportion of the total variation. The
results of the analyses in Table 7 can be best discussed
in combination with the data of Table 8 and Figure 6 below.
Table 8 shows mean variation within and between internodes,
whereas Figure 6 graphically illustrates mean variation
between internodes. When 3-carene is high there is a

negative interaction between 3-carene and a- and B-pinene,

 

55

00I0~ qusa mmIsa 0~IN~ mMImm mocmaumuocos annoy

m.HIH.0 m.0Im.0 N.HIm.0 m.0I0 HI0 mawaoaflmnme
H.0I0 H.0I0 m.0I0 H.0I0 m.0I0 mamawmnmuI>
0.HIm.0 H.HI0.0 0.0IH.0 NIH 0.HIN.0 madnpcmHHmnmIm
0Im.0 mIH 0I~ mIH ~0Im.0 camcoqu
0HI0 0IH 0I~ 0.0IH.0 0I0 mamuMOIm
H.HI0.0 0.0I0.0 0.0I0.0 NIH 0IH memos»:
HmIm 0I~ AIS mHIm HmIH mamcflaIm
m.HIN.0 0.0I0 0.0I~.0 N.0IH.0 0.0IH.0 mcoaasmo
sflIN SIN HHIs 0Im NHIN mcmcHaIa

Aucooummv mmaam> mmnu CH mmasm

 

 

mmawapmmm UHHQMSHmsHQ wwwumm mwsmmm mbaHHQMm
mafia mufln3 x 0.33 am p.53 mosnmm monomnmuocoz
ammummm mwflcmmoum mcflapmmm moanmm mean:

I I

.mmwpdpm mafia wufi£3 ammummw can .mmflammonm madnmm .oosnmm muwnz
one SH mmmup Ecum.monEMm umOEmems as» no uaouaoo mammduoaoe SH momamm .0 dance

56

Table 7. Summary of the analyses of variance and additional
statistics of the principal monoterpenes in white
spruce oleoresin.

 

 

Level of significance

 

 

Source df . . 3
a-pinene B-pinene A -carene
Tree 7 <0.0005 <0.0005 <0.0005
Internode 2 NS NS <0.0005
Tree X Internode l4 <0.0005 <0.0005 0.0030
Position 1 0.0010 NS NS
Tree X Position 7 0.0230 NS NS
Internode X Position 2 NS NS NS

Additional Statistics

I 0.70 0.90 0.60

R 0.99 0.99 0.97

 

the intraclass correlation coefficient.

'1
II

N
W
II

the coefficient of multiple determination.

with an increase in tissue age 3-carene decreases and both
a- and B-pinene increase. But in trees lacking 3-carene,
a-pinene levels remain essentially constant with increas-
ing tissue age and B-pinene decreases. These differential
responses in a- and B-pinene with 3-carene levels probably
caused the high significance levels in the tree-internode

interaction in the analyses of variance for both

57

Table 8. Mean within-tree variation of the principal
monoterpenes in white spruce oleoresin.

 

 

 

 

Tissue
3-carene age (yrs.) Monoterpene
concen- and
tration Internodal _ inene 8- in 3_
positionl a p p ene carene
percent of oleoresin
High 1 U 3.65 9.53 6.24
l L 4.48 10.22 4.47
2 U 6.16 12.18 2.43
L 6.35 12.43 2.89
3 U 6.66 11.92 1.67
3 L 6.88 12.95 1.94
Lacking l U 6.89 7.41 --
1 L 8.72 7.54 --
2 U 7.37 5.91 --
L 8.48 6.00 --
U 7.52 5.55 --
L 8.42 5.50 --

 

1U = upper internodal position; L = lower inter-

nodal position.

a- and B-pinene.

for most of the variation in d- and B-pinene between inter-

nodes.

And the interaction seems to have accounted

In addition, a-pinene is consistently higher in

the lower portion of each internode, accounting for the

high significance level of the position effect (Table 7).

58

Figure 6. Mean within-tree variation of the principal
monoterpenes in white spruce oleoresin;
variation in four trees with high and four
trees lacking 3-carene.

Percent of oleoresin

Nahum-400:0

 

59

d-Pinene

 

 

\mé-Corene

O-— -Trees with high 3-Coren?‘~ .0
D—Trees lacking 3-Corene

l 1 I

 

O

l 2 3
Internode (age of tissue in years)

60

Greater differences in a-pinene between the upper and lower
positions in those trees lacking 3-carene probably caused
significance in the tree-position interaction.

In the experimental design of this study, the
intraclass correlation coefficients (Table 7) may be con-
sidered more a measure of repeatability, since the within-
tree samples are in a sense, repeated measurements on a
tree. However, the variation expressed here is best con-
sidered a function of developmental effects; thus these
estimates of intraclass correlation or repeatabilities are
a measure of the ratio of genetic variance to phenotypic
variance (phenotypic variance here equals the sum of gene-
tic, developmental, and unexplained environmental variance
plus error) (Falconer, 1960). In this study, though the
intraclass correlations are quite high in all three
monoterpenes, develOpmental variation is nevertheless

significant.
Study 2 - Spruce Progenies

In the second study of spruce families, variation
in a-pinene, B-pinene, and 3-carene paralleled that in the
previous study. Table 9 gives a summary of the analyses
of variance and Table 10 lists the mean within-tree‘varia-
tion in each family of the three species. Positional
differences were quite strong, especially in a-pinene and

3-carene; the B—pinene variation pattern between positions

61

was not as great and even reversed in one family (cf.,

Table 10, white spruce family number 23 X 52).

The family-

position interaction was high for a-pinene; this interac-

tion is analogous to both the tree-internode and

tree-position interactions above, except that this varia-

tion is according to family rather than tree-to-tree

differences.

Table 9. Summary of the analyses of variance and additional
statistics of the principal monoterpenes in seed-
lings of white spruce, blue Spruce, and white X
blue spruce families.

w

Level of Significance

 

 

 

 

Source df 3
a-pinene B-pinene -carene
Family 11 <0.0005 <0.0005 <0.0005
Family (tree) 21 0.0300 <0.0005 NS
Position 1 <0.0005 <0.0005 <0.0005
Family X position 11 0.0400 NS NS
Additional Statistics
rIl 0.35 0.74 0.35
22
R 0.92 0.98 0.93
23
R1 0.17 0.13 0.15
1rI = the intraclass correlation coefficient.
2R2 = the coefficient of multiple determination.
3Ri = the proportion of the total sum of squares

due to within-family effects.

62

 

 

 

 

 

Table 10. Mean within-tree variation of the principal
monoterpenes in seedlings of white spruce, blue
spruce, and white spruce X blue spruce families.

Number Monoterpene
Species Family of Positions a- 8-
trees pinene pinene 3-carene
percent of oleoresin
White 23x52 3 01 5.99 13.12 0.20
Spruce L 6.37 12.85 0.06
25x15 3 U 4.04 12.54 2.79
L 5.15 14.33 1.89
15x16 3 U 3.72 10.37 2.23
L 6.86 11.25 1.31
64x54 3 U 5.25 11.55 2.04
L 6.58 12.14 0.89
White 15x12 3 U 5.89 5.37 5.02
Spruce L 10.72 6.50 1.69
x 23x19 3 0 2.59 6.47 4.80
Blue
L 2.97 9.65 2.49
Spruce
25x21 1 U 3.98 6.57 4.73
L 6.18 10.15 1.21
64x21 2 U 5.74 9.03 1.70
L 8.90 10.66 1.31
30x12 3 U 6.14 4.41 4.74
L 9.74 6.10 2.80
Blue 1x19 3 U 4.79 5.43 6.27
Spruce L 4.85 5.88 3.16
12x4 3 U 8.30 1.54 4.19
L 13.79 1.58 1.97
2x1 3 U 7.62 5.14 2.90
L 10.36 5.87 1.09
1U = upper stem position; L = lower middle stem

position.

63

In this study, a variant (Ri) of the coefficient
of multiple determination (R2) was calculated: it is the
portion of the total variation due to unexplained within-
family effects (within-family sum of squares divided by
the total sum of squares). I found it interesting that
this portion was nearly the same in the three major mono-
terpenes (about 15 percent). Again, the proportion of the
total variation explained by the model (R2) was high.

The intraclass correlation coefficient here must
be interpreted differently from the one in the previous
study. Here the notion of genetic variance is vague;
although these families in the study are from full-sib
crosses, they had not been arranged in an orderly design.
So estimation of broad-sense genetic variance from the
between- and within-family effects would be difficult.
Thus the intraclass correlations in this study must be
strictly defined as that prOportion of the total variance
of a randomly selected seedling due to family. The pro-
portion is quite high for B-pinene but low for a-pinene
mainly because the estimated phenotypic variance is
equally divided among all the effects. In the case of
3-carene, though the correlation is low, the between-
family and between position effects together take up 70
percent of the phenotypic variance. The low amount of
within-family variation in 3-carene suggests that 3-carene

levels are quite uniform in each family and that

64

heritability in 3-carene may be quite high. This corrobo-
rates the hypothesis that the inheritance of 3—carene

involves few genes (Hanover, 1966c).

Study 3 - Eastern White
Pine Seedlings

 

 

For the eastern white pine seedlings, most of the
results are the same as those for the white spruce saplings.
The model explains most variation, tree-internode inter-
actions are high for a- and B-pinene, and positional
effects are high and in the same direction for a-pinene.
These results are given in Tables 11 and 12, and Figure
7. A few major differences in these results compared with
those for white spruce are evident. First, the between
internode effects are high in all three monoterpenes.
Next, B-pinene, on the average, first decreases in the
second internode, then rises slightly in the third of
trees high in 3-carene and a- and B-pinene covary nega-
tively in trees low in 3-carene. Finally, the intraclass
correlations, equivalent to those in the white spruce
study, are here very high.

The variations and covariations found in the three
studies presented here are similar to those reported in
other studies. Hanover (1966a) found a general increase
in a- and B-pinene and a decrease in 3-carene, with in-
creasing tissue age in the cortical oleoresins of Pinus

monticola (western white pine). Eastern and western pines

 

65

Table 11. Summary of the analyses of variance and addi-
tional statistics of the principal monoterpenes
of eastern white pine oleoresins.

 

 

 

 

Significance

Source df . . 3

a-pinene B-pinene A -carene
Tree 5 <0.0005 <0.0005 <0.0005
Internode 2 0.009 0.001 0.024
Tree X internode 10 0.009 0.021 NS
Position 1 0.037 NS NS
Tree X position 5 NS NS NS
Internode X position 2 NS NS NS

Additional statistics

 

 

r11 _0.92 0.88 0.97
22
R 0.99 0.99 0.99
1r = the intraclass correlation coefficient.
2R2 = the coefficient of multiple determination.

are thought to be closely related species. In the branch

resin of scotch pine (Pinus sylvestris), Juvonen (1966)

 

reported an increase in a-pinene, a decrease in 3-carene,
and an increase, then a decrease in B-phellandrene with
increasing tissue age. Tobolski (1968) reported similar
results in scotch pine. And in slash pine, Roberts re-

ported an increase in a-pinene with age in the branch

66

Table 12. Mean within-tree variation of the principal
monoterpenes in eastern white pine oleoresin.

 

 

3-carene Tissue age Monoterpene
(yrs.) and
concen-
tration Internodal a- inene 8- inene 3—caren
position1 p p e
percent of oleoresin
High U 3.28 9.51 13.56
1 L 3.66 9.79 13.34
U 3.90 7.45 12.24
L 4.51 7.56 11.21
U 5.25 7.05 10.87
3 L 6.60 8.57 11.47
Low 1 U 14.38 15.73 0.05
1 L 14.92 16.13 0.09
2 U 15.83 13.23 0.03
L 16.59 13.03 0.04
U 14.64 13.93 0.05
L 15.53 14.06 0.04

 

1U = upper internodal position; L = lower inter-
nodal position.

cortex of trees high in B-phellandrene. However, from a
survey of within-tree variation in the cortical monoter-
penes of Abi§§_spp., Zavarin (1969) found that, though
changes were significant in the youngest tissue, they

varied according to the species.

67

Figure 7. Mean within-tree variation of the principal
monoterpenes in eastern white pine oleo-

resin: three trees with high and three
trees with low 3-carene.

Percent of oleoresin

 

68

    

 

 

I7-
l6- d-Plnene
l5-
I4-
.3. \ fi-Pinene
\\
I2- \O~-“
"' 3-Carene
IO'
o\\
9" \
\\ fi—Pinene
8" \ __,——-G
‘o—-""'
7..
6' ’0
/
5 . / /P
c- inene
4)- ”’O’
or’
3..
2 O---Trees with high 3-Carene
- EI— Trees with low 3-Carene
" 3-Corene
I 2 3

Internode (age of tissue in years)

69

Factor Analyses

 

To determine how correlations within trees of the
three studies, compared with those between trees, factor
analyses were performed on each species. The important
factor vectors in the analysis of the monoterpenes in
white spruce saplings are presented in Table 13.

Table 13. Major factors in the factor analyses of within-

tree variation in the monoterpenes in white
spruce and eastern white pine oleoresin.

 

 

Eastern

White spruce white pine

 

 

 

Monoterpene
Factor Factor Factor Factor Factor
1 2 3 l 2

a-pinene ++2 0 — __ +
Camphene ++ o _
B-pinene 0 ++ 0 0 ++
Myrcene - -— _ ++ 0
3-carene 0 + ++ ++ _-
Limonene 0 -- - ++ -
B-phellandrene 0 ++ 0 — o
y-terpinene 0 ++ ++ _
Terpinolene 0 ++ ++ -

 

1Each factor is a vector taken from the factor
analysis of the monoterpene simple correlation matrix in
each species (cf. Chapter I).

2++ or -- = a positive or negative loading, re-
spectively, which has an absolute value exceeding 0.50;
+ or -,= a positive or negative loading which has an

absolute value between 0.15 and 0.49; 0 = a zero loading--
has an absolute value less than 0.15.

70

This analysis shows correlations consistent with those
given in Table 3. with the exception of the weak positive
correlation, in Factor 2, between B-pinene and 3—carene.
In addition, the correlations among B-pinene, limonene,
and B-phellandrene, unique to the red and white spruces

in Table 3, were also consistant in Sign in this analysis.
However, because of built-in correlations in the data, the
subsets of data with high 3-carene, no 3-carene, high
limonene, and low limonene were analyzed.

These analyses showed the following: (1) when
3-carene was high, B-pinene and 3-carene were negatively
correlated, which is also indicated in Figure 6. Thus,
the positive relationship between B-pinene and 3-carene
in Table 13 could be seriously questioned. (2) Terpino-
lene and Veterpinene maintained a strong, positive rela—
tionship, even when 3-carene was absent. (3) Beta-pinene
and limonene were negatively correlated only when limonene
was high, otherwise they were positively correlated. This
suggests that partial replacement of B-pinene with limonene,
and vice-versa, occurs only when limonene is a major com-
ponent of the oleoresin (see the discussion in Chapter I).
(4) The correlation of a-pinene with myrcene varied ac-
cording to the subset analysed. (5) In the high limonene
subset, the B-pinene-limonene and the B-pinene-B-phellan-

drene relationships each loaded on separate factors,

71

suggesting that limonene and B-phellandrene levels are
controlled separately by levels of B-pinene.r

The factor analyses of the white spruce, the blue
Spruce and the hybrid families, illustrated in Table 14,
contain the same main relationships as in the analysis of
the white spruce in Table 13. But the analyses of blue
spruce and the hybrid indicate that those relationships
associated with limonene and with myrcene, though differ-
ent in the two species, are intermediate in the hybrid.
This suggests, then, a hybrid effect on the correlations.

Results of the factor analysis of eastern white
pine are summarized in Table 13. The same correlations
shown for eastern white pine in Table 3, were also found
here. Again, because built-in correlations are present
in the data, the complete data set was split into high and
low 3-carene subsets. Two of the correlations found in
the subsets are illustrated in Figure 7, that is, a
negative relationship between a- and B-pinene when 3-carene
is low, and negative relationships between 3-carene and
a-pinene and B-pinene when 3-carene is high. In addition,
when 3-carene is high, a- and B-pinene are uncorrelated,-
implying that 3-carene is the common, regulating factor
in the relationships between 3-carene and a-pinene and
3-carene and B—pinene.

In the low 3-carene subset, there was a strong

positive correlation between B-pinene and 3-carene. This

72

.mH.0 asap mmma oaHm> muaa

Iowan am mmaIImaHpmoH anon a u 0 “00.0 paw mH.0 ammzumn msaa> ouaHOQO as as:

30033 maflpmoH o>0uamoa no m>flufimom a u I no + “0m.0 maapmmoxm maam> muaaomnm
as man £0033 .wam>fluommmmn .maflpmoH o>wummma no m>HuHmom a H II no ++N

.AH Hmummao .mov mmwommm comm a0 xflnumfi acauaamnnou mHmEHm mammnmu
IoaOE on» no mammamam nouoam on» Baum amxmu nouom> a ma Manama commH

 

 

+ I I 0 ++ 0 I ++ mamaoaflmnoa
++ 0 0 I + 0 0 ++ mamaflmnmuIr
0 ++ 0 ++ 0 0 ++ 0 mcmuscmaamraIm
0 0 I 0 + 0 II 0 mamaoaflq
o I I 0 ++ o I ++ OCOHMOIM
0 + 0 ++ 0 0 0 ++ 0:000»:
0 ++ 0 ++ I I ++ 0 macaHmIm
0 .0 ++ 0 I ++ 0 0 mamamfinu
0 0 ++ I I ++ 0 «I oaoaHaIa
m m H m a m m ”a

nap Hop non uou nou HOD Hop no»

Ioam Iomm Iomm Iomm Iomm Iomm Iomm loam oammumuoaoz

 

 

 

 

woaumm maam x

monumm muޣ3 madman maam modumm 09033

 

 

I
ll

.mmaaflfiam moanmm oSHQ.x muanz
can .moanmm mafia .woanmm muwn3 mo mmaflapmmm SH mmaomumuoaoe may
aw aowumflnm> monuIawnuH3 mo mommamam Houomw on» as mnouomu sons: .00 manna

73

reversal in correlations is analogous to the situation in
white spruce saplings above, where the correlation between
B-pinene and limonene reversed when limonene ceased to be
a major component in the oleoresin. Thus when 3-carene
becomes a trace component, its effect on B-pinene is
altered.

In summary, major within-tree correlations, at
least in eastern white pine and white Spruce, do corrobo-
rate those found in between-tree samples. However, there
are indications that these change with gross variations
in the levels of individual components. The former obser-
vation suggests that.these correlations are biogenetically
significant; the latter suggests that gene and enzyme

regulatory mechanisms may be major causes of them.

Questions Raised
from the Results

As many questions have been generated from this
study as have been answered: Is within-tree variation
physiologically continuous or fixed? That is, do the
monoterpenes vary as the tissue ages or do they remain
unchanged relative to each other following the initial
formation of the oleoresin? Further, what are the sources
of genotype-environment interaction in monoterpenes?;
what combined biochemical and genetic events affect these

differential changes?

 

74

Studies of seasonal variation in the monoterpenes
of Scotch pine spanning one year (Juvonen, 1966; Tobolski,
1968), showed that variations were more or less random,
implying that monoterpene composition remains fixed
within a tissue. However, some of the trees in the hybrid
families in this study, resampled in.November, showed
significant increases in a- and 8-pinene levels and a
decrease in 3-carene levels. But the monoterpenes in
those trees further resampled in February, changed insig-
nificantly. Then, because the stems of those trees which
changed composition, has been sampled soon after they had
ceased growth, there may be a "maturation period" for the
oleoresin. During this period relative and absolute com-
positional changes occur after which composition varies
randomly. Evidence for maturation changes has been shown
in developing seedlings of maritime pine (Monteoliva,
1970): although the per gram tissue quantities of a- and
B-pinene increased rapidly, the two monoterpenes varied
negatively relative to each other.

In the final question above, answers are purely
Speculative at present. However, a preliminary resolution
can be had, initially through breeding for monoterpene
covariation. Between-progeny analyses should indicate the
extent of gross genetic effects; within-tree analyses of

these progeny should show biochemical-genetic interactions.

75

Practical Application
of the Results

 

 

Chemotaxonomic investigations up to the present
have centered on the comparative variation of natural sub-
stances between taxa. But tentative differences in
monoterpene correlations between white spruce and blue
spruce Show that between- and within-individual covaria-
tions may be useful in taxonomic and evolutionary work.

In addition, within-tree covariation in a particu-
lar pine or spruce species should be considered in
breeding for a desired monoterpene content at a given stem
position. For example, Renwick and Vité (1970) have de-
scribed systems of chemical communication in bark beetles
specific to pine species in which monoterpenes interact
with pheromones produced by the insect, accelerating its
attack on the tree. Moreover, the stem position attacked
is species-specific for a given insect (Dr. L. F. Wilson,
personal communication). Therefore, if one were to breed
for low amounts of the interacting monoterpene at a given
stem position, within—tree covariation of that monoterpene
with the others must first be accounted for. In breeding
for low a-pinene in the upper stem of eastern white pine,
for example, one might select those trees low in a-pinene

and high in 3-carene.

 

CHAPTER III

BETWEEN- AND WITHIN-TREE VARIATION IN THE

CORTICAL RESIN ACIDS OF PICEA GLAUCA,

 

PICEA PUNGENS, AND PICEA GLAUCA

 

 

X PUNGENS

Very little is known about the resin acids in the
cortical oleoresin of spruces and pines. All the work
done has been mainly survey-type, involving pine resin
acids. In one such survey, Joye and Lawrence (1967)
found differences over a range of hard and soft pine
species, suggesting the possible taxonomic utility of
the resin acids.

This study was initiated to gain an understanding
of how resin acids vary qualitatively and quantitatively
in the cortical oleoresin of three spruce species: white

Spruce (Picea glauca (Moench) Voss), blue spruce (Picea

 

pungens Engelm.), and white X blue spruce (Picea glauca

 

X pungens).

Materials anddMethods

 

Cortical oleoresin was sampled in February, 1972,

from the upper and lower stem positions of three white

76

77

spruce, two blue spruce, and five hybrid spruce seedlings.
These seedlings had been selected from the spruce families
described in Chapter II. The cortical monoterpene content
of each seedling had been previously determined.

The sampling method was as follows: a razor inci-
sion was made in the stem cortex. Between 1 and 5 ul of
the exuded oleoresin was drawn into calibrated 20 ul
capillary pipettes, placed in a sealed 1 m1 vial, and re-
frigerated at 5°C for no more than two days before analy-
sis. Duplicate samples were also obtained for monoterpene
analysis so that these could be checked for possible
changes in composition from previous analyses.

Analysis of the samples was done by gas-liquid
chromatography using an F & M model 700 instrument equipped
with dual columns, a hydrogen flame ionization detector,
and a Hewlett-Packard 3370A digital integrator. The 15 ft
by 3/16 inch O.D. c0pper column was packed with 60 to 80
mesh Chromosorb W-NAW coated with 8 percent Versamid 900.
Column temperature was 240°C, detector and injection ports
300°C; and helium flow rate was 80 ml/min. Resin acids
were converted to their methyl esters by pyrolysis of
their tetramethylammonium salts in the injection port of
the chromatograph (Hetman, et_al., 1965).

Quantitative measurements of the resin acid methyl
esters were made by the internal standard method (Barbato

and Umbreit, 1966) using methyl arachidate as the internal

78

standard. Linear calibration curves for 1ev0pimaric acid
and dehydroabietic acid, each in relative response com-
pared with methyl arachidate, were prepared; these are
presented in Figures 8 and 9 respectively. Methyl levopi-
marate has been previously reported as having a relative
response different from the other resin acid methyl esters
(Joye and Lawrence, 1967).

The method of calibration was as follows: a solu-
tion of 6.25 mg/ml methyl arachidate (MA) in spectro-grade
anhydrous methanol and 0.5 mg/ml each of levopimarate (LPA)
in methanol and dehydroabietate (DHA), methanol were pre-
pared. Next, 20 ul of the levopimarate solution was added
to each of three 1 m1 vials and 10, 20, and 40 ul of the
methyl arachidate solution was added to vials one through
three, respectively. Weight ratios of LPA to MA for vials
one through three were 0.64, 0.32, and 0.16, respectively.
The same procedure was repeated for DHA except 30 01 MA
solution was added to vial three; thus respective weight
ratios DHA to MA in vials one to three were 0.64, 0.32,
and 0.213. Each of the six standard solutions were
titrated to a phenolphthalein end point with a 0.10 N
tetramethylammonium hydroxide solution. Duplicate 4 ul
aliquots of each standard were injected into the gas
chromatograph. Peak areas computed by the digital inte-

grator were in uV sec.

79

Figure 8. Linear calibration curve for levopimaric acid
(LPA) with methyl arachidate (MA) as internal
standard.

80

 

LOO

0.50
Wclaht rotlo (LPA/MAI

 

 

LOOF

w
J
O
Evita: 2:: :3 zoom

81

Figure 9. Linear calibration curve for dehydroabietic
acid (DHA) with methyl arachidate (MA) as
internal standard.

82

LOO

0.50

rotio(DHA/MA)

Weight

 

 

 

LOOP

a
5.
O
2.2}19 3:: 020 0.0:.

83

The respective mean peak area ratios (LPA/MA) and
weight ratios (LPA/MA) for each LPA-MA standard are given

below (see also Figure 8):

 

 

 

Solution Area ratio (LPA/MA) Weight ratio (LPA/MA)
1 0.311 0.640
2 0.148 0.320
3 0.075 0.160

Mean 0.178 0.373

 

Thus, the relative response of LPA to MA is:

Mean area ratio
Mean weight ratio

 

Relative response =

0.178
0.37

= 0.48

Joye and Lawrence (1967) reported a relative response of
0.56 for LPA. However, they used a graphic method in
computing peak areas. In checking for the source of this
difference in relative response, the latter method was
used in calculating peak areas from the chromatograms in
my study: the resulting ratio was equivalent to that

reported by Joye and Lawrence.

84

The respective mean peak area ratios (DHA/MA) and
weight ratios (DHA/MA) for each DHA-MA standard are given

below (see also Figure 9):

 

Solution Area ratio (LPA/MA) Weight ratio (DHA/MA)
1 0.429 0.640
2 0.231 0.320
3 0.142 0.213

Mean 0.267 0.391

 

Here the relative response (DHA to MA) is 0.68, calculated
in the same manner as done above. When the graphic method
was used in computing peak areas in the DHA-MA samples,
the resulting relative response was again equivalent to
the corresponding ratio reported by Joye and Lawrence
(1967). Samples with LPA/MA and DHA/MA weight ratios
greater than those shown in the respective calibration
curves of Figures 8 and 9, gave relative responses equal
to those from the curves.

Oleoresin samples were prepared by first dissolv-
ing the samples in 15 to 20 pl of spectra-grade anhydrous
methanol. Then each sample was titrated to a phenolphtha—
lein end point with the 0.10 N tetramethylammonium hydroxide
solution and evaporated to approximately 10 01 on a steam

bath under a nitrogen stream. A measured volume of the

85

methyl arachidate solution was next added to the sample;
initially, 4 ul of methyl arachidate were added to each
of the first five samples; later 5 01 rather than 4 01
were added to the remaining five. A 4 pl aliquot of the
mixture was then injected into the chromatograph. Figure
10 illustrates a typical composite chromatogram of the
resin acid fraction in spruce oleoresin sampled in this
study.

The resin acid methyl esters were tentatively
identified by comparing relative retention times with

those of the following known compounds: A8(9)

-is0pimarate,
communate, anticopalate, primarate, sandarac0pimarate,
levopimarate, isopimarate, dehydroabietate, strobate,
abietate, and neoabietate. Neither communate nor antico-
palate were detected in the spruce samples. Strobate and
dehydroabietate were found to have equal relative retention
times. The retention times for the resin acids and un-
known compounds found in the spruce oleoresin samples are
listed in Table 15.

Quantitative measurements of both the resin acid
methyl esters and the unknown compounds were made using
the following procedure: first, peak areas on each
chromatogram were corrected for baseline drift. Next,
the area for methyl arachidate was divided into the peak

area for each resin acid methyl ester and unknown. The

resulting area ratios were divided by the corresponding

86

Figure 10. Chromatogram of the resin acid fraction of
white spruce, blue spruce, and white X blue
spruce cortical oleoresin.

87

 

OIIcIccccu a:

sunny 2

p
cmcus [cucIcchpluca 2

a
cIIchIdch c -

 
    

 

OIJOOI|Od/9|JIUIOOAO1 a
cIuIcIdcchcpIIcs 2
cum-Id

”ovum-av mum

IOUDQ‘O'

      
    

 

 

 

 

III ulcutcc

TIcIc

Table 15.

88

Relative retention times of resin acid methyl

esters and unknown compounds found in white,

blue, and white x blue hybrid Spruce

cortical oleoresins.

Resin acid methyl esters

A

:—
1

Relative retention time1

 

\oaoqamee-wNI-I
C

N N N |—‘ l—' H H H H H I'" l" I"
c c c c c c c c c c c c c U" m c
c c

23.

Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
08(9)-isopimarate

Unknown

Pimarate

Sandaracopimarate
Levopimarate + Palustrate
Isopimarate

Unknown

Dehydroabietate + Strobate
Unknown

Abietate

Unknown

Neoabietate

Unknown

Unknown

Unknown

0.47
0.52
0.55
0.63
0.70
0.75
0.81
0.94
1.11
1.18
1.23
1.28
1.38
1.48
1.67
1.80
1.89
1.99
2.08
2.27
2.33
2.56
2.76
2.89
3.40

 

1Retention times are calculated relative to
methyl arachidate, the internal standard (1.00).

89

relative response, 0.48 for the levopimarate area ratio
and 0.68 for the other area ratios; this gives the weight
ratio for each compound relative to methyl arachidate.

The absolute weight (in mg) of each compound in the sam-
ple was determined by multiplying its weight ratio by the
weight (in mg) of the methyl arachidate added to the
sample. Because the densities of the resin acids are not
known, the absolute weight calculated was considered to
roughly approximate volume in 01. The volume measured
was then expressed as a percent of the volume of oleoresin
sampled. Total resin acid composition is the sum of the

concentrations of all the individual compounds.

Results and Discussion

 

Thirteen major components were present in the
resin acid fraction of the oleoresin of white spruce,
blue spruce, and white X blue spruce seedlings. These
are presented in Table 16 for the white Spruce seedlings,
Table 17 for the blue spruce seedlings, and Table 18 for
the white X blue spruce seedlings. Because unknowns 2, 3,
and 4 each eluted close together on the chromatograph and
varied in the same direction, their quantities were
summed.

No clear differences were found in the resin acids
between species. The only distinct difference between

species was in the sum of unknowns 2, 3, and 4. It was

90

.0090» 0:0 :0 p0©a0oa0 00a mvaaomfioo Hoa0E no 850 mpa0oa0 0000090

.0 cam .m .N mGSOaMaD mo EdmN

.maouma0mfi 0:0 00 a00000m 00300 pas 00mm: on» on 00m00 00300 cam 00mmo0

 

 

 

 

 

00.00 00.00 00.00 00.00 00.00 00.00 00.00 m0000.0.
00.0 00.0 00.0 00.0 00.0 00.0 00.0 00000000002 .00
00.0 00.0 00.0 00.0 00.0 00.0 00.0 00000000 .00
00.0 00.0 00.0 00.0 00.0 00.0 00.0 0300000 .00
. . . . . . . 00000000
00 0 00 0 00 0 00 0 00 0 00 0 00 0 + 00000000000000 .00
00.0 00.0 00.0 00.0 00.0 00.0 00.0 0300000 .00
00.00 00.00 00.00 00.00 00.00 00.00 00.00 00000000000 .00
O O O O O O O mpmuumslfimm
00 00 00 00 00 00 00 00 00 00 00 00 00 00 + 000000000>00 .00
00.0 00.0 00.0 00.0 00.0 00.0 00.0 00000000000000000 .00
00.0 00.0 00.0 00.0 00.0 00.0 00.0 0300000 .0
00.0 00.0 00.0 00.0 00.0 00.0 00,0 0300000 .0
azoaxaa .0
00.0 00.0 00.0 00.0 00.0 00.0 N00.0 0300000 .0
I CHMOHOOHO m0 #COOHOA $300,300.03 I N
com: 003mg momma. 00300 00mm: 000300 00mm: 000000 0%:009
m OOHB N OOHB H OOHB UHUM fiflmmm

l‘
i I
. II

.mma00cmmm 0050mm 00003 00030 00 a00000000 00000000 000 :0 mp0om aflmmu 00002 .00 00308

91

.0030“ may :0 cmcsaocfl 00c mccsomEoo HOQHE mo Sam mwsaocw mamuoam

.0 can .m .m mc3ocxc: mo Edm0

.msmumGHME may 00 soapuom Hm300 can 00mm: 000 00 Hmmmu 00300 can ummmo0

 

 

 

 

00.00 00.00 00.00 00.00 00.00 m00000.
00.0 00.0 00.0 00.0 00.0 00000000002 .00
00.0 00.0 00.0 00.0 00.0 00000000 .00
00.0 00.0 00.0 00.0 00.0 0300000 .00
00.0 00.0 00.0 00.0 00.0 00000000
+ 000000000000000 .00
00.0 00.0 00.0 00.0 00.0 0300000 .00
00.00 00.00 00.00 00.00 00.00 00000200000 .00
mumuumsamm
00.00 00.00 00.00 00.00 00.00 + 000000000300 .00
00.0 00.0 00.0 00.0 00.0 00000000000000000 .00
00.0 00.0 00.0 00.0 00.0 0300000 .0
00.0 00.0 00.0 00.0 00.0 0300000 .0
:3ochD .0
00.0 00.0 00.0 00.0 N00.0 0300000 .0
afimwnomao mo unmoumm G3ocxco .m
:00: 00300 00mm: 000300 00mm: mumumm Hmnumfi
N 0009 0 0009 @000 GHmmm

'F

i
I

.mmcwacmmm

0050mm @503 03» mo cflmmnomao 00000000 may c0 mvflom awmmu Honmz .50 00309

92

.mEmumchE may

.mHQmu map 2H UmvsHocH uoc mccsomeoo HOGHE mo Edm mUSHoaH mHmuOBm

.w cam .m .N mc3ocxcs mo EsmN

mo GOHMHOQ Hmon cam Hmmmn 0:“ cu Hmmmu Hm3oH cam “mums”

 

 

ov.mm mm.mm mm.~n H¢.Hm «v.ow mm.vm m>.¢m mm.mm ~N.hm om.>w MH.mm mHmuoe
m>.m mm.w hm.m mm.v mm.m ¢H.m mm.m mo.¢ m¢.m Hv.v hn.w mumumHnmomz .om
mm.v mm.m mv.m mm.~ Hm.¢ mm.m mm.v mm.m OH.m mo.m hm.m mumumHn¢ .mH
mv.o mm.o mv.o Hm.o nm.o mm.o mm.o mm.o om.o mm.o hm.o Q3OGMGD .nH
O O O O I I O O O O O mquO-Hpm +
co H HH H on H mm H mH H mv o mm o mm H Ho H up o mu 0 mumumHnmouUmsma .mH
NH.o oo.o oo.o oo.o oo.o H~.o mw.o oo.o oo.o mm.o mm.o c3ocxco .mH
mm.HH Hm.MH hm.vH MH.m MH.oH mm.~H mh.~H ~o.~H mo.HH Hm.m Hm.OH mumnmEHQOMH .vH
O O O O O I O O O O I mpmuumglfimm
OH mm mm hN mo om mm mm om mm mo vm mm Hm om vm av mm mm vm mH mm + mumumEHmo>mH .mH
mm.o 05.0 Hm.o mmao mm.o mm.o o>.o Hm.o vm.o mv.o mm.o mumumEHmoomgmvcmm .NH
mm.v Hv.m mv.m mm.m nH.m mo.h mm.v o~.m mw.~ mm.m mm.~ czocxcb .m
no.o 0H.o mo.o 0H.o MH.o no.0 vo.o vo.o vo.o mo.o no.o c3ocxco .m
csocxca .v
mm.H mH.m vm.H mm.H mm.o No.m vH.m no.H Hv.o mo.H ~mm.o c3ocxcb .m
:Hmmuomflo mo unmoumm czoqxcb .m
cams Hmsoq ummmo uw3oq Hmmma Hm30H ummms uw3oq memp HHm3oq human

 

mmHB

v mmna

m mmne

N mmna

H mmua

mnmumm Hanumfi
UHom chmm

 

 

.mmcHHUmmm Gwyn»:
mosumm msHQ x muHss 0>Hm mo chmHomHo HMUHuHoo map SH mUHom :Hmmu Mona:

.mH GHQMB

93

low in white spruce, high in blue spruce, and intermediate
in the hybrid. Within-tree variation was found in abietate
and neoabietate; these decreased in the lower stem, except
for abietate in blue spruce.

The resin acid fraction quite clearly is the major
component in the oleoresin; total monoterpenes in the
seedlings ranged between 18 and 29 percent. Total resin
acids generally followed variations in total monoterpenes
in the seedlings; the latter component was lowest in the
blue spruce.

I have one comment to make with regard to the
technique for quantifying the resin acid measurements.
Caution must be taken to be sure that the quantity of
methyl arachidate added is not too large relative to the
resin sample volume or to the final sample volume. Other-
wise, the arachidate tends to crystalize out of the
solution.

In summary, the results of this study show that
white spruce and blue spruce are quite similar in resin
acid content in the cortical oleoresin. This contrasts
with the differences in monoterpenes between the two
species (cf., Table 10). However, white and blue spruce
do interbreed freely; thus their similarity in resin acid
content might be expected. It is possible that further

improvements in technique and a broader sample of white

94

and blue spruce genotypes will lead to a more complete
understanding of the relationships between the resin

acids of these species.

CONCLU S IONS AND RECOMMENDAT IONS

Between- and within-tree relationships among mono-
terpenes appeared to be very similar in this study.
However, the reasons for this covariation are not known;
this information must come from more direct approaches
such as radiotracer studies and work on cell-free systems.
Radiotracer studies on intact pine or spruce plants may be
difficult to perform, especially if the study is of time—
course incorporation. Few samples can be taken from a
single internode on the stem before oleoresin pressure
drOps, eliminating the internode as a source of labelled
resin. The problem might be overcome through two experi-
mental designs, but both are less desirable than the method
above. In one, the internodes from lateral branches of
the same age could be sampled as each becomes useless; in
the other, cuttings or grafts from equal-aged sources of
tissue might be used.

Time-course work on the incorporation of carbon
label would give preliminary indications of the physiology
of the monoterpenes, whether the label goes into each
compound at different and varying rates. Carbon—l4 dioxide

seems to be the best source of label for this purpose.

95

96

For more detailed work, cell-free systems would be
more suitable. The best source of material would probably
be dormant buds, basically because the prOportion of resin
canal volume to the total is quite large in these organs.

The significant developmental differences found in
the cortical monoterpenes point to the need for more de-
tailed work in this area. Repeated samples of cortical
monoterpenes over a long period of time--two years or more--
will show whether the differences are continuous or fixed
or both. Of particular interest would be work on the com-
position of buds and flowers. Attention should be given
not only to differences between these organs and adjacent
tissue but also to changes during the development of each.

Because gibberellins and abscisic acid are both
terpenoids, develOpmental changes in the monoterpenes of
any of these organs--stems, buds, and flowers--may have
significance in the regulation of plant physiological
processes.

Since very little is known about the resin acids in
cortical oleoresin, much remains to be done. Natural vari-
ation, modes of inheritance, and age and tissue effects:
these are all areas in which considerable information has
been compiled in the monoterpenes, but where information
is lacking in the resin acids.

However, work is also needed to determine the inter-

relationships between cortical monoterpenes and resin acids.

97

Knowledge of how much variation in the two components is
independent versus how much is correlated should give in—
sights into how they are related physiologically.

Finally, the oleoresin itself is growing in impor-
tance in studies on pest resistance. Thus it will be
necessary to relate changes in monoterpene and resin acid
levels to the physical prOperties of oleoresin as mentioned

earlier.

L ITERATURE C ITED

LITERATURE CITED

Asplund, R. O. 1968. Monoterpenes: relationship between
structure and inhibition of germination. Phytochem.
7:1995-1997.

Banthorpe, D. V., J. Mann, and K. W. Turnbull. 1970.
Terpene biosynthesis of thujane derivatives in

Thuja, Tanacetum, and Juniperus species. J. Chem.
Soc. (C7:2689-2693.

 

 

Barbato, P. C., and G. R. Umbreit. 1966. Internal standard
technique in quantitative gas chromatographic analy-
sis. Facts and Methods (Hewlett-Packard Corp.),
7(1):lZ-14.

Beytia, E., P. Valenzuela, and O. Cori. 1969. Terpene
biosynthesis: formation of nerol, geraniol, and
other prenols by an enzyme system from Pinus radiata
seedlings. Arch. Biochem. Biophys. 129:346-356.

 

Britten, R. J., and E. H. Davidson. 1969. Gene regulation
for higher cells: A theory. Science 165:349-357.

Burbott, A. J., and W. D. Loomis. 1967. Effects of light
and temperature on the monoterpenes of peppermint.
Plant Physiol. 42(1):20-28.

, and W. D. Loomis. 1969. Evidence for metabolic
turnover of monoterpenes in peppermint. Plant
Physiol. 44:173-179.

Carolin, V. M., and J. W. Baxter. 1968. Bud characteris-
tics of ponderosa pine related to potential damage
by the European pine shoot moth. USDA, Forest Serv.
Res. Note PNW-90. l7pp.

Cattell, R. B. 1965a. Factor analysis: An introduction
to essentials. I. The purpose and underlying
models. Biometrics 21:190-210.

. 1965b. Factor analysis: An introduction to

essentials. II. The role of factor analysis in
research. Biometrics 21:405-435.

98

99

, and M. J. Foster. 1963. The rotOplot program
for multiple, single-plane, visually-guided rota-
tion. Behav. Sci. 8:156-165.

Cori, O. 1969. Terpene biosynthesis: Utilization of neryl
perphosphate by an enzyme system from Pinus radiata
seedlings. Arch. Biochem. BiOphys. 135:416-418.

 

Datta, P. 1969. Regulation of branched biosynthetic path-
ways in bacteria. Science 165:556-562.

Draper, R. G., and H. Smith. 1966. Applied Regression.
Analysis. J. Wiley and Sons, Inc., New York.

Falconer, D. S. 1960. Introduction to Quantitative
Genetics. Ronald Press, New York.

Finley, K. T. 1969. The development of the carbonium ion
hypothesis. Eastmas Organic Chem. Bull. 41(2).

Forde, M. B., and M. M. Blight. 1964. Geographical varia-
tion in the turpentine of Bishop pine. N. Zealand
J. Bot. 2(1):44-52.

Francis, M. J. O. 1971. Monoterpene biosynthesis. In:
Goodwin, T. W. (Ed.), Aspects of Terpenoid Chemistry
and Biochemistry, Academic Press, New York.

Gasciogne, R. M. 1958. Entropy changes in isomerisations.

Gustafson, R. A. 1961. Partial correlations in regression
computation. J. Amer. Stat. Assn.:363-367.

Hanover, J. W. 1966a. Environmental variation in the
monoterpenes of Pinus monticola Dougl. Phytochem.
5:713-717.

 

. 1966b. Genetics of terpenes. I. Gene control
of monoterpene levels in Pinus monticola Dougl.
Heredity 21: 73-84.

 

. 1966c. Inheritance of 3-carene concentration in
Pinus monticola. Forest Sci. 12:447-450.

 

. 1971. Genetics of terpenes. II. Genetic vari—
ances and interrelationships of monoterpene concen-
trations in Pinus monticola. Heredity 27(2):237-
245.

 

100

, and R. c. Wilkinson. 1969. A new hybrid between
blue spruce and white spruce. Canad. J. Bot.
47(11):1693-1700.

Hefendehl, F. W., and M. J. Murray. 1972. Changes in
monoterpene composition in Mentha aquatica produced
by gene substitution. Phytochem. 11:189-195.

Heftmann, E. 1965. Steroids. In: Bonner, J., and J. E.
Varner (Eds.), Plant Biochemistry, Academic Press,
New York.

Hetman, N. E.; H. G. Arlt, Jr.; R. Paylor, and R. Feinland.
1965. Gas chromatographic analysis of tall oil
rosin acids. J. Am. Chem. Soc. 42:255-256.

Joye, N. M., Jr., and R. V. Lawrence. 1967. Resin acid
composition of pine oleoresins. J. Chem. and
Engineering Data 12:279-282.

Juvonen, S. 1966. Uber die die Terpenbiosynthese
beinflussenden Factoren in Pinus sylvestris L.
Acta Botanica Fennica 71:1-92.

 

Kempthorne, O. 1969. An Introduction to Genetic Statis—
tics. Iowa State University Press, Ames, Iowa.

Loomis, W. D. 1967. Biosynthesis and metabolism of mono-
terpenes, p. 59-82. In; Pridham, J. B. (Ed.),
Terpenoids in Plants, Academic Press, New York.

Monteoliva, M. 1970. Biosynthesis of terpenes in pine.
Mimeo. Tech. Report No. FG-Sp-145, Univ. Granada,
Granada, Spain.

Murray, M. J. 1960. The genetic basis for a third ketone
group in Mentha spicata L. Genetics 45(7):931-937.

 

, and D. E. Lincoln. 1970. The genetic basis of
acyclis oil constituents in Mentha citrata Ehrh.
Genetics 65:457-471.

 

Pirila, V., E. Siltanen, and L. Pirila. 1964. On the
chemical nature of the eczematogenic agent in oil
of turpentine. IV. The primary irritant effect
of terpenes. Dermatologica 128:16-21.

Rasmussen, R. A. 1964. Terpenes: Their analysis and
fate in the atmosphere. Ph.D. Thesis, Wash. Univ.

lOl

Renwick, J. A. A., and J. P. Vité. 1970. Systems of
chemical communication in Dendroctonus. Contrib.
Boyce Thompson Inst. 24(13):283-292.

 

Roberts, D. R. 1970. Within-tree variation of monoterpene
hydrocarbon composition of slash pine oleoresin.
Phytochem. 9:809-815.

Ruzicka, L. 1953. The isoprene rule and the biogenesis
of terpenic compounds. Experientia 9:357-396.

Sandermann, W. 1958. fiber de Biogenese der Terpene und
Terpenoid. Proc. Fourth International Congr.
Biochem., Vienna, Vol. II - Biochemistry of Wood:
29-47.

Sandermann, W. 1962. Biosynthetische Untersuchungen on
verschiedenen Kieferarten. Holzforschung 16(3):
65-740

Shrimpton, D. M. and H. S. Whitney. 1968. Inhibition of
growth of blue stain fungi by wood extractives.
Canad. J. Bot. 46:757-761.

Simonsen, J. L. 1920. The constituents of Indian turpen-
tine from Pinus longifolia Roxb. Part I. J. Chem.
Soc. 117:570-578.

 

Smith, R. H. 1964. Variation in the monoterpenes of Pinus
ponderosa Laws. Science 143:1337-1338.

 

Squillace, A. E. 1971. Inheritance of monoterpene compo-
sition in cortical oleoresin of slash pine.
Forest Sci. 17:381-387.

Steele, R. G., and J. H. Torrie. 1960. Principles and
Procedures of Statistics. McGraw-Hill, Inc., New
York.

Sukhov, O. V. 1957. The use of radiocarbon in the study
of biosynthesis of terpenes. Internat. Conf.
Radioisotopes in Sci. Res. UNESCO, Paris, p. 1-11.

Thomson, R. B., and H. B. Sifton. 1925. Resin canals in
the Canadian spruce (Picea adensis [Mill.] B.S.P.);
an anatomical study, especially in relation to
traumatic effects and their bearing on phylogeny.
Phil. Trans. Royal. Soc. Land. (Ser. B) 214:65-111.

 

102

Tobolski, J. J. 1968. Variations in monoterpenes in
Scotch pine (Pinus sylvestris L.). Ph.D. Thesis.
Michigan State Univ.

 

Veldman, D. J. 1967. Fortran Programming for the Behavioral
Sciences. Holt, Rinehart and Winston, New York.

von Schantz, M., and S. Juvonen. 1966. Chemotaxonomische
untersuchungen in der gattung Picea. Acta Bot.
Fenn. 73:51pp.

Werker, E., and A. Fahn. 1968. Site of resin synthesis
in cells of Pinus halepensis Mill. Nature, Land.’
218(5139):388-389.

 

, and A. Fahn. 1969. Resin ducts of Pinus halepen-
sis Mi11.--Their structure, development and pattern
of arrangement. Bot. J. Lin. Soc. 62(4):379-411.

 

Wilkinson, R. C. 1970. Chemical analysis of species re-
lationships in northeast american spruces. Ph.D.
Thesis. Michigan State Univ.

; J. W. Hanover; J. W. Wright; and R. W. Flake.
1971. Genetic variation in the monoterpene composi-
tion of white spruce. Forest Sci. 17:83-90.

Wooding, F. B. P., and P. H. Northcote. 1965. The fine
structure of the mature resin canal cells of Pinus
pinea. J. Ultrastruct. Res. 13:233-244.

Zavarin, E. 1968. Chemotaxonomy of the genus abies--II.
Within tree variation of the terpenes in cortical
oleoresin. Phytochem. 7:92-107.

. 1970. Qualitative and quantitative co-occurance
of terpenoids as a tool for elucidation of their
biosynthesis. Phytochem. 9:1049-1063.

N

M'TITI'I‘I’IflfiwiflfljfljfirfltfllfllijffltlflfglfljflrfililTl'Es