SEED SOURCE X ENVWONMENT iNTERACTIONS
EN SCG‘FCH PINE

‘E‘hesis ifor iho Dogma of Ph. D.
MECHEGAN STATE UNIVERSI?Y'
fame-s 57'. King
2964

THESIS

This is to certify that the
thesis entitled
SEED SOURCE X ENVIRONMENT
INTERACTIONS IN SCOTCH PINE

presented by

JAMES PHILLIP KING

has been accepted towards fulfillment
of the requirements for

Ph. D. degree in Forestr!

 

- I
Date/[1; éclfl 2/; /5‘//f(

0-169

_.

L I B R A 1
Michigan 5
Universi~

  
    

 

twgvg‘"

We:

ABSTRACT
SEED SOURCE X ENVIRONMENT
INTERACTIONS IN SCOTCH PINE

by James P. King

This study was undertaken to determine the effects of
differing Michigan environments on the growth of Scotch pine

(Pinus sylvestris L.) seed sources.

 

Scotch pine seed, collected from 122 native stands
throughout the species range, was sown in the Michigan State
University forest tree nursery in the Spring of 1959. Each
seed lot consisted of seed from about ten trees per stand.

In 1961 two-year-old stock was used to establish perman-
ent test plantations throughout Michigan and the central
United States. The plantings follow a randomized block
design with seven to ten replications. The number of seed
sources per plantation varies from 50 to 100.

In 1962 following the second growing season after out-
planting, one plantation in central Illinois was measured
for height growth while in Michigan seven plantings were
measured for height growth, five for needle length, and seven
for color. In 1963 four Michigan plantations were measured
for height growth and needle length. Data from 6” seed
sources were used in this study.

Analyses of variance of each character was made for each
test plantation using plot totals as items. The individual

plantation analyses were grouped into various combinations

James P. King - 2

and the mean squares used to compute the resulting variance
components. These components were then expressed as a
percent of the total variance for comparative purposes.

Seed sources from Belgium, Germany, and Czechoslovakia
made the most height growth at all plantings in 1962 and
1963. Sources from Spain made the least. However in 1963
there were indications that the Spanish sources might out—
_grow the Scandinavian sources as the effects of tranSplanting
dissipated.

The planting sites also showed marked differences in
growth rate. They differed in the amount of growth per year
and in the change in growth rate with time.

Sources from central Europe had the longest needles and
sources from either the northern or southern extremity of
the Scotch pine range had the shortest needles. However,
this trend differed more by plantation than did height
‘growth. While the Scandinavian sources were relatively short-
needled in the northernmost Michigan plantation, they had
relatively long needles in the southernmost plantation. The
Spanish sources, on the other hand, all had relatively longer
needles at the northernmost plantation.

The mean plantation needle length differed sharply with
site and year. In 1963 the mean needle length decreased by
25 percent in the southernmost Michigan plantation while it
increased by 10 percent at the northernmost planting. It is
suggested that this year x plantation interaction was the

result of a late frost in May 1963.

James P. King - 3

The Spanish, Greek—Turkish, and south France sources
were the darkest green and sources from the Ural mountains
and Scandinavia were the most yellow. Sources from Scandi-
navia showed more yellowing in central lower Michigan than
at any other plantation but were still not as yellow as the
Ural mountain sources.

The seed source x plantation interaction of the indivi-
dual seed sources showed no relation to seed source location
or plantation location. Differences in performance of sources
between plantations seems more a result of temperature and
moisture variations than between-planting differences in
soil or photOperiod.

The component of variance resulting from seed source
x plantation interaction never accounted for more than six
percent of the total variation encountered in all plantings
in either 1962 or 1963. This interaction component was about
1/6 the seed source component for heightgrowth; 1/5 the
seed source component for needle length; and about 1/12 the
seed source component for color.

The effect of yearly fluctuations of climate on seed
source differences (year x seed source interaction) was also
very small in relation to the seed source differences.

Under the conditions of pest infestation encountered
in this test (very low), the researcher could gain little
more than a two or three percent increase in precision by

replicating measurements in time and Space.

James P. King — u

From a practical standpoint this data indicates that
the most_genetic gain per dollar spent would be obtained by
testing an increased number of seed sources in a single
planting and then replicating in time and Space only a few
of the best seed sources.

Data from other tests indicates that these results
Should also be applicable in mature trees throughout North—

central and Northeastern United States and central Europe.

SEED SOURCE X ENVIRONMENT

INTERACTIONS IN SCOTCH PINE

By

James P. King

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Forestry

196“

ACKNOWLEDGEMENTS

I am deeply grateful to Dr. Jonathan W. Wright, who
supervised this study, and to Drs. M. Wayne Adams, John E.
Cantlon, Fred C. Elliott, John E. Grafius, Terrill D.
Stevens, and Donald P. White, all of Michigan State Univer-
sity, for their many valuable comments andguidance.

Thanks are also due to John Bright and Walter Lemmien
of Michigan State University, Dr. Jalmer Jokela of the
University of Illinois, Dr. Donald T. Lester of the Univer-
sity of Wisconsin, and Eric Bourdo of the Ford Forestry
Center, L‘Anse, Michigan, for making available their test
plantations and aiding in the measurements.

The staff of the Computer Center of Michigan State
University helped with the data analyses.

The study was financed in part by funds from the
Cooperative State Research Service of the U. S. Department
of Agriculture as part of regional project NC-Sl entitled
"Tree Improvement through Selection and Breeding." This
project involves active cooperation of numerous federal, state,
and private agencies in the North Central United States.

A large number of c00perators in Europe and the United States

supplied seed.

ii

TABLE OF CONTENTS

INTRODUCTION ....................
REVIEW OF LITERATURE ................
Provenance Tests ...............
Agronomy Tests ................
Controlled Environment Tests .........

OBJECTIVES .....................

 

METHODS ......................
Material -------------------
Measurement ..................
Analysis ...................

RESULTS ......................
Height Growth .................

Needle Length -----------------

 

Color ---------------------
Field Measurements ............
Laboratory Measurements .........

APPLICABILITY OF RESULTS TO OTHER AREAS ------

 

DISCUSSION .....................

 

PRACTICAL APPLICATION OF RESULTS ----------

 

SUMMARY - — - - - - — - - - a ........ a--- -

LITERATURE CITED ..................

 

iii

13
1a
19
20
20
26
30
33
33
1+2
148
us
52
56
59
53
66

69

LIST OF TABLES
Table Page

1 Combined height growth analysis for thirty-one
Scotch pine seed sources grown in New York and
New Hampshire ----------------- ll

2 Seed source location data--grouped by region - 23

3 Summary of plantation location, size, and
method of establishment ------------ 25

4 Plantation climatic data ----------- 27

5 Analysis of variance form used for the combined
analysis of a single year's results ------ 32

6 Analysis of variance form used for the combined
analysis of yearly results ---------- 32

7 Height growth of Scotch pine seed sources - - - 3M

8 Combined analysis of variance results of 1962
and 1963 height growth for 55 Scotch pine seed
sources from four Michigan plantations - - - - 36

9 1962 and 1963 components of height growth vari-
ance as a percent of total variance--from 55
seed sources ----------------- 39

10 1962 components of height growth variance ex-
pressed as a percent of the total variance of
#2 seed sources ---------------- NO

11 Between-plantation correlation coefficients
using the mean 1962 seed source height growth - u3

l2 Needle length of Scotch pine seed sources - - - an
13 Combined analysis of variance for 1962 and 1963
needle length of 55 Scotch pine seed sources
grown in four Michigan plantations ------ #5

1n 1962 and 1963 components of needle length variance
as a percent of total variance--from 55 seed

sources -------------------- H7
15 Between plantation correlation coefficients
using the mean 1962 seed source needle length - H9

iv

Table

16

17

18

19

LIST OF TABLES (CONT.)

Color grades and date of scoring Scotch pine
seed sources ...................

Between-plantation correlation coefficients
using the mean seed source color grade for the
1961 and 1962 measurements ----------

1962 components of color variance as a percent
of total color variance-~from 57 seed sources

Results and analysis of laboratory color
measurement ..................

1962 combined analyses of variance of height
growth ....................

1963 combined analyses of variance of height
growth ....................

1962 combined analyses of variance of height

pgrowth ....................

1962 combined analyses of variance of needle
length --------------------

1963 combined analyses of variance of needle
length --------------------

Combined analyses of variance for color
scoring ...................

Analysis of variance of individual plantation
1962 height growth --------------

Analysis of variance of individual plantation
1963 height growth ..............

Analysis of variance of 1963 individual planta-

tion needle length --------------

1962 analysis of variance results of planta-
tion color scorings .............

Page

80

80

80

81

81

82

82

83

LIST OF FIGURES
Figure Page

1 Natural distribution of Scotch pine in EurOpe
(shaded) and provenances in this study (num-
bered dots) (Wright and Bull, 1963) ----- 21

2 Natural distribution of Scotch pine in Asia
(shaded) and provenances included in this
experiment (numbered dots) (Wright and Bull,

1963 -------------------- 22

3 Location of Scotch pine test plantations used
in the regional study ------------ 24

.vi

LIST OF APPENDICES

APPENDIX A —————————

METHODOLOGY ........

APPENDIX B .....................

ANALYSES OF VARIANCE CHARTS

vii

Page
75
76
79

8O

INTRODUCTION

 

Scotch pine (Pinus sylvestris L.) is the most import—

 

ant commercial tree species in Europe. Its natural range
extends from Spain and Turkey in the south to Scotland and
northern Finland in the north, and eastward across Russia
almost to the Pacific Ocean. In the southern part of the
range it is found in isolated stands on cool moist mountain
slopes. Toward. the north its occurrence is continuous
over large areas.

Although Scotch pine has been planted and grows well
over a wide area of the northeastern United States, American
foresters held it in low regard because of its poor form.
However, as evidence mounted indicating the poor form was
due to improper seed source, interest in the species was
renewed. The tree reproduces naturally in many parts of
the United States, is easily transplanted, and grows well
on a variety of sites. It is a favorite of Christmas tree
growers. In the past three years one-fifth to oneathird of
all stock growing in Michigan forest tree nurseries was
Scotch pine. It is logical then that this species should
become the subject of a comprehensive tree improvement
program.

Tree improvement research relies on the presence of
genetic variation. In the past, few if any estimates of

genetic variation of forest trees have been made in a

manner that excludes bias from seed source x environment
interactions. Yet, unless the magnitude of this bias is
known, a realistic estimate of the rate of improvement is
not possible. Furthermore, the magnitude of the seed source
x environment interaction determines the extent of the area
to which test results will apply and provides a measure of
the profit of breeding for specific locations.

Review 2: Literature

 

Provenance Tests

 

A single replicated seed source test_gives a valid
estimate of the error variance necessary for testing the
Significance of seed source differences, but confounds
variance due to seed source with variance due to seed source
x environment interaction. Tests of this nature have been
common with forest tree species since about 1930. Results
of such tests have been summarized by Wright (1962).

Unreplicated seed source tests repeated at several
locations confound seed source x plantation interaction with
within-seed-source variation. Thus the magnitude of the
interaction cannot be determined and conclusions from such
tests cannot be applied to Specific environments. Tests
in this category have been conducted with Scotch pine

(Wiedemann, 1930; Kalela, 1937), Douglas-fir (Pseudotsuga

 

menziesii (Mirb.) Franco) (Munger and Morris, 1936),

 

ponderosa pine (Pinus ponderosa Laws.) (Munger, 19u7;

 

Squillace and Silene, 1962), and Norway Spruce (Picea abies

 

(L.) Karst.) (Rubner, 1957; Scthbach, 1957).

Thus far only a few seed source tests

designed to

measure interaction have been reported. These include

maritime pine (Pinus pinaster Ait.), loblolly pine (P.

taeda L.), slash pine (P. elliottii Engelm. var. elliottii),

 

 

shortleaf pine (P, echinata Mill.), longleaf pine (P.

palustris Mill.), jack pine (P, banksiana Lamb.), eastern

 

 

white pine (P. strobus L.), ponderosa pine, Scotch pine,

Norway spruce, and Eur0pean larch (Larix decidua Mill.).

The majority of these tests have not been reported in

sufficient detail to allow a complete analysis of the inter-

action variance. But even a look at the general trends can

be revealing. For while a constant ranking of sources

between plantings does not preclude the existence of inter-

action, it does indicate that the seed source component of

variance is much greater than the interaction component.

Thus while the estimates of genetic variation from a

single test may not be free of interaction
of this bias may not be great. This seems
in most forest tree species.

Rycroft and Wicht (19H?) reported the
of a seven-origin teSt of maritime pine at
in western and southern South Africa. The

seed source x plantation interaction. The

bias, the effect

to be the case

ten-year results
eight test sites
data showed no

Portuguese source

was best at all Sites and the ranking of the remaining origins

remained constant from site to site.

Wakeley (1961) reported on the mean five-year height

of eight sources of loblolly pine grown in two plantings

in the southeastern United States. This test is part of the
Southwide Pine Seed Source Study. The southernmost plant-
ing was in Mississippi (latitude 30°HH'N.) and the
northernmost planting was in Maryland (latitude 36°35'N.).
The latitude of the seed sources ranged from 30°N. to 38°N.
The correlation of five-year height with seed source lati-
tude showed a significant positive relationship (r = .92
with 6 degrees of freedom) at the northern planting and a
negative, but non-significant, relationship (r = -.48 with
6 degrees of freedom) at the southern planting. These
results suggest an interaction due to the relatively better
height growth of the southern origins at the southern
planting.

Five-year height of shortleaf pine Shows the same
trend. Using seven seedlots whose source ranged from 31°N.
to u0°N. latitude and grown in plantings in Louisiana
(30°58'N.) and Tennessee (36°00'N.), there was a signifi-
cant negative correlation (r = -.97 with 6 degrees of free-
dom) between height and seed source latitude at the
Louisiana planting, but a non-significant correlation
(r = .3u with 6 degrees of freedom) in the northern planting.

Three-year height growth of shortleaf pine at plantings
in Louisiana (30°H2'N.), Tennessee (36°13'N.), and New
Jersey (39°36'N.) also bears out the trend toward increased

height growth of southern seed sources in southern

plantations. Using seven seed sources whose source latitude
ranged from 31° to ”0° N. the correlation between the three-
year height and seed source latitude were significant and
negative in the southern plantings (Louisiana r = -.88; and
Tennessee r = -.82, with 5 degrees of freedom), but signi-
ficant and positive in the New Jersey plantation (r = .9u
with 5 degrees of freedom).

Incidence of fusiform rust (Cronartium fusiform Hedge.

 

and Hunt) in the Southwide Pine Seed Source Study was reported
by Henry (1959). Comparing slash pine seed sources from
Florida, South Carolina, Alabama, Louisiana, and Mississippi
in fiveeyear old plantations in the same five states, signi-
ficant differences between sources appeared only in the
South Carolina planting where the Florida source was signi-
ficantly higher in infection than the other four sources.
Incidence of fusiform rust on loblolly pine in plantings
in Louisiana, Mississippi, Alabama, and North Carolina were
also discussed by Henry (1959). The Texas, Maryland,
Arkansas, and Louisiana sources fell into a relatively low
susceptibility_group as compared with the North Carolina,
South Carolina, Georgia, Alabama and Mississippi sources.
The Tennessee source was intermediate. One marked exception
to this relationship was the high infection of the Texas
source and the low infection of the Onslow County, North
Carolina source in the Talladega County, Alabama planting.
Snyder and Allen (1963) reported on ten-year height

growth of four sources (two from Alabama; two from

Mississippi) of longleaf pine. Seedscollected from these
sources in three successive years (19u7, 19u8, and 19u9)
were sown in two nurseries in Alabama and Mississippi and
then planted at two locations in Alabama and two in Missi-
ssippi. Their analysis of variance showed a highly
significant seed source x plantation interaction. The
nature of the interaction was not stated.

Five-year results of a 29-origin Lake States jack pine
test have been reported by Arend 33,31. (1961). Seed was
collected from 29 jack pine stands in Minnesota, Wisconsin
and Michigan and outplanted in a u-replicated randomized
block design at 17 locations in the 3 states. Arend's
report covered measurements of the three plantations in
lower Michigan. The data showed a non-significant seed
source x plantation interaction for height growth and inci-
dence of white pine weevil (Pissodes strobi (Peck)) in the
three plantations studied.

Seed sources from lower Michigan made the best height
growth as a group and the sources from Upper Michigan the
least. Seed source differences in weevil incidence were not
related to geography.

In the same Lake States jack pine test Rudolph (1962)
studied 1ammas growth, prolepsis and long bud formation in
10 selected origins at Six plantations for a two year period.
The origins were selected to cover the entire range of jack

pine in Minnesota, Wisconsin, and Michigan. Four of the

plantations were located in Minnesota and two in Wisconsin.
The data showed a Significant seed source x plantation
interaction for 1ammas growth and long buds, but none for
prolepsis. The data also indicated a significant seed
source x plantation x year interaction for 1ammas growth.
Several sources behaved differently between several plantings.
There appeared to be no relation between interaction, location
of source or location of planting.

In two Wisconsin plantations of this same test, the pre-
sent author found a significant seed source x plantation
interaction in susceptibility to jack pine needle cast

(Hypodermella ample Dearn.) One planting was located in

 

the western portion of Michigan's upper peninsula and the
other planting was located about 100 miles to the south in
central Wisconsin. Sources from northeast Minnesota showed
the highest susceptibility to this disease and sources from
lower Michigan the least.

Only one source showed any great difference between
plantings. This source, from lower Michigan, Showed a much
lower susceptibility in upper Michigan than in central Wisconsin.

Three-year height growth of Six seed origins of eastern
white pine common to replicated plantings in North Carolina,
Georgia, and Virginia were described by Sluder (1963). The
source from Georgia made the best growth at all plantings; the
Nova Scotia, Ontario, and Minnesota sources were the three
poorest at all plantings. The West Virginia and Pennsylvania

sources were not significantly different in the Virginia

planting; the West Virginia origin was significantly
better than the Pennsylvania origin in the Georgia plant-
ing; and the Pennsylvania origin was significantly better
than the West Virginia origin at the North Carolina plant-
ing. Thus, the performance of the West Virginia and
Pennsylvania provenances in the three plantings indicates
a significant seed source x plantation interaction.

Another range-wide provenance test of fifteen sources
of eastern white pine has been reported by Wright 3: 31.
(1963). At two plantings in southern Michigan there was no
significant seed source x plantation interaction for mortal-
ity, color and six-year height. The source from Tennessee
was the fastest growing at both plantings and was followed
by sources from Georgia and Pennsylvania. The two slowest
growing sources were from Nova Scotia and Minnesota. Wright
EEHEl' compared their results with ten sources in common
with a New Jersey test reported by Santamour (1960). Their
comparison shows a non-Significant interaction for height
.growth.

Height and diameter growth of one- and two—year old
ponderosa pine seedlings was recently described by Baron
and Schubert (1963). The seed in this test was collected
from five seed collection zones and grown in four California
nurseries. The collection zones were: (1) northern east
Side Sierra (high elevation, H000 to 6000 feet above sea
level); Q) northern west Side Sierra (low elevation, 2000

to uooo feet above sea level); (3) northern west side Sierra

(high elevation); (H) southern west side Sierra (low ele-
vation); and (5) southern west side Sierra (high elevation).
The four nurseries range in elevation from 226 to 3252 feet
above sea level and in latitude from 37°N. to H1°N.

At all four nurseries seed from zone four made the
best two-year height growth and was not Significantly ex-
ceeded in diameter growth by any other seed collection.

Zone one stock was poorest at all nurseries.

Evidence that a seed source x plantation interaction in
ponderosa pine may develop as the trees mature comes from
two reports by Mirov E: El. (1952) and Callaham and Liddi-
coet (1961). Both reports concern the same ponderosa pine
individual-tree progeny test. In this test seed was collected
from 89 trees along a narrow east-west transect on the west
slope of California's Sierra Nevada mountains. The elevation
of the 89 trees ranged from 125 to 6919 feet above sea level.
The progenies were grown at three planting sites--960, 2730,
and 5650 feet above sea level.

At the end of 12 years' growth, MirOV'EE_E£. reported
that sources from 1500 to 3500 feet in elevation made the
best growth at all planting sites, i.e. no major interaction.

At the end of 20 years, however, Callaham and Liddicoet
reported a distinct change from the 12—year trends. At
the two lower plantings the progenies from the lower eleva-
tions made the best height growth and progenies from high
elevations the least. At the high elevation planting site,

however, there were no significant differences between

lO

progenies.

Callaham and Liddicoet (1961) also reported on 20-
year height growth of 21 Jeffrey pine (Pi23§_jeffreyi,

Grev. and Balf.) progenies grown at the same planting sites
as the preceding ponderosa pine test. The jeffrey pine
show the same pattern as found in the pnnderosa pine, i.e.
high elevation progenies did relatively better at the high
elevation planting than at either lower elevation planting.

Two separate reports on the 1938 International Union
of Forest Research Organization Scotch pine test have been
issued for a New York and a New Hampshire plantation (Wright
and Baldwin, 1957; Schreiner, Littlefield and Eliason, 1962).
The authors published their data in sufficient detail to
permit a combined analysis for height of 31 origins common
to both plantings. The results of this analysis are shown
in Table 1.

The strong differences between plantations are partly
due to the fact that the New Hampshire data were from 17-year
measurements and the New York data were from 18-year measure-
ments. A careful examination of the data indicated that this
age difference did not contribute to the interaction.

The small but Significant interaction is probably due to
the fact that three of the five south Swedish origins (Wright
and Baldwin's "ecotype D") did better in relation to the
overall mean at New Hampshire than at New York, while eleven
of the eighteen German origins ("ecotype G") did better at

New York than at New Hampshire.

11

Table 1. Combined height growth analysis for thirty-one
Scotch pine seed sources grown in New York and
New Hampshire.1/

 

 

: Degrees :Component of Variance

Source of : of :as a Percent of Total
Variation : Freedom: Mean Square: Variance
Plantation 1 1601.Mu** 69.07
Rep. within

Planting u #53.25
Seed Source 30 38.03**_2_/ 22.88
Seed Source X **

Plantation 30 3.n2 2/ 3.76

Rep. within seed
Source 120 1.07 n.29

 

** Significant at the one percent level.

1/ New York data from Schreiner et a1. 1962.
New Hampshire data from Wrighf_afia Baldwin, 1957.

2/ Based upon seed source x plantation interaction as error
term.

3/ Based upon seed source x replicate-within-planting as
error term.

12

Holst (1963) recently summarized the results of the
International Union of Forest Research Organization Norway
spruce provenance tests begun in North America in 1938-39.
These tests are located in Michigan, Wisconsin, Massachu-
setts, New Hampshire, Ontario and New Brunswick. Because
the number and source of the seedlots varied between planta-
tions a direct measure of seed source x plantation inter-
action is not possible. However, a comparison of the
general trends at each planting suggests little interaction.
Sources from Poland, White Russia, Czechoslovakia, Yugoslavia
and Rumania made better height growth than other EurOpean
and Scandinavian lots in every planting.

The height growth superiority of east-central European
seed sources was also reported by Langlet (1963) for a 36-
origin Norway spruce provenance test in Danelt, Sweden.
This test was also a part of the 1938 International Union
of Forest Research Organization provenance tests.

Although the same trends appeared at both the North
American and Swedish tests if seed sources are grouped by
general area, the data suggest that some individual seed
sources responded differently between countries. For example,
the source from Stolpce, White Russia S.S.R., exceeded the
plantation average by thirty percent at Harvard Forest in
New Hampshire but only equaled the plantation average at
Dbnjelt, Sweden. The source from Muntele, Rumania, exceeded

the plantation average at Danelt by twelve percent, but fell

l3

twelve percent below the plantation average at Manistee,
Michigan. It must be kept in mind, however, that these
comparisons are probably biased by the fact that thirty-six
origins were represented at Danelt while only 13 and 10
origins were in the New Hampshire and Michigan tests re-
spectively.

Vins (1963) reported on diameter growth of twenty-two
provenances of Norway spruce at two plantings in Czecho-
Slovakia. These plantings were located at 340 and 850
meters above sea level. The trend noted in height growth
was also noted in diameter growth, i.e. sources from east-
central Burcpe were best and sources from northern Europe
poorest. The low correlation between provenances at various
ages in two plantings (1989 r = .57; 1958 r = .71; 1962 r =
.61) suggests that there was a seed source x plantation
interaction between these two plantations.

Genys (1960) showed high correlations in heightgrowth
of EurOpean larch seed sources between United States and
EurOpean plantations. His correlations were based on lZ-year
height in New York and New Hampshire and: (l) 9-year height
in Scotland (r = .83 with 13 degrees of freedom); (2) 10-year
height in northeast Germany (r = .93 with 6 degrees of freedom);
and 11-year height in northern Italy (r = .87 with 12 degrees
of freedom).

Agronomy 22532
More detailed analysis of genotype x environment inter-

actions have been made in several farm crop plants. Gardner

1”

(1963) has recently reviewed interactions in cross-fertiliz-
ing crop plants and Matzinger (1963) has reviewed interactions
in self-fertilizing crOp plants. The theory and implications
of estimating_genotype x environment interactions has been
discussed by Comstock and Moll (1963).

Among the most important conclusions from these papers
are these:

1. As the genetic diversity of the material increases,

the relative magnitude of the interaction decreases.

2. The interaction components vary with the area in

the test.

3. The interaction components may be as large or

larger than the varietal component.

Item number one above is of special interest to tree
breeders. As foresters are presently provenance testing
diverse material from natural populations, the Size of the
interactions should be relatively small. However, as pro-
venance testing continues and narrows down the desirable
seed sources to the few best, genotype x environment inter-
action will assume an increasing importance.

Controlled Environment 23535

Numerous studies under controlled environment condi-
tions have shown interactions between seed source and
specific factors of the environment. Photoperiodic response
has been the object of the largest number of studies.

Vaartaja (195a) collected Scotch pine seed from several

trees in two Finnish stands located at 65° N. latitude and

15

at 60°30' N. latitude. These two seed sources were given
two photoperiodic treatments; one limited to two hours of
natural light, and the other treatment ten hours of natural
light plus continuous (2” hour) low intensity light. The
northern origin seedlings grew better than the southern
origin under continuous light, but under short days the
southern origin made the best growth. These growth differ-
ences were measured by needle number and dry weight.

Wassink and Wiersma (1955) collected Scotch pine seed
from Sweden (66° N.) and the "Massif Central" of France
(u5°3o'). They tested the seedling growth under two photo-
periods, one a 12-hour day and the other a 12-hour day plus
continuous weak light. The seed from northern Sweden made
only slightly less growth under long days, but much less
under short days.

Vaartaja (1959) tested Scotch pine seed from Finland
(60° N.) and Spain (Hl° N.). In this test Vaartaja simulated
four different daylengths by varying the length of the
period of continuous dark period. All plants received light
from 6:00 a.m. to 5:00 p.m. Then, for example, the very
long day treatment would receive one hour of light from
11:00 p.m. to 12:00 p.m. This limited the dark period to
six hours and was considered equivalent to an eighteen hour
day. With this method Vaartaja could hold the amount of
light constant and vary only the length of phot0period.
Daylengths of 12, 1M, 16, and 18 hours were simulated. As

in the previous tests northern origins grew for a longer

16

period and made better growth relative to the southern
origins under the long photoperiod.

Vaartaja (1959) has also reviewed the evidence for
"photoperiodic ecotypes" in other tree Species. In general,
the results agree with those of Scotch pine. That is, seed
origins from northern latitudesshow a greater response to

change in photoperiod than southern origins.

Karschon (19u9) studied the photoperiodic response of
Scotch pine seed from different elevations in Switzerland.
He used a short day treatment consisting of natural light
only and long day (15 hours) treatment with artificial light.
Using seed from five stands ranging in elevation from 385
meters to 1770 meters, he found a statistically significant
interaction between seed source and treatment in height
growth. The lower the elevation of the seed source, the

greater was the growth under short days.

Callaham (1962), using ponderosa pine progenies from
26 localities throughout the ponderosa pine range, measured
growth under nine combinations of three day temperatures
(30°, 23°, and 17° C.) and three night temperatures (22°,
1n°, and 7° C.) with a constant daylength of sixteen hours.
The results showed that seedlings from east of the Rocky
Mountains need high night temperatures for best growth;
seedlings from the southwestern United States did best with
cool days and hot nights; and Pacific coast seedlings grew
reasonably well at lower night temperatures.

Perry (1962) reported an interaction of day and night

l7

temperatures in red maple (A235 rubrum L.). Using seed
collected from three or four widely separated trees at six
locations from Florida to Ontario and Minnesota, the seed-
lings were grown for 3% days with a constant day temperature
of 23° C. and night temperatures ranging from 7° C. to 26° C.
Northern origins made their best growth with a night temp-
erature of about 17° while the southern origins did best
with a night temperature of about 20° C. Using only two
origins (Vermont and Florida), Perry grew the plants under
a constant night temperature of 17° C. at different day
temperatures of 17°, 20°, 23°, 26°, and 30°C. Again tree
growth of the two seed sources varied with temperature.
The Vermont seedlings made their best growth with a day
temperature of 26° while the Florida seedlings showed in-
creasing growth with increasing temperature. Perry's data
also indicated the presence of a seed source x temperature
x photoperiod interaction in red maple. This same type
second-order interaction was also suggested by Irgens-Moller
with Douglas-fir (1962).

Kriebel and Wang (1962) reported an interaction between
seed source and amount of pre-chilling. They subjected four

provenances of sugar maple (Acer saccharum Marsh.) to seven

 

durations of outdoor chilling in northern Ohio and observed
the frequency and timing of bud burst. The sugar maple of
southern origin broke dormancy with less outdoor chilling

than did the northern sources. The time between the end of

the chilling period and the time of bud break was reduced

18

more in the northern origins than in the southern origins.
The preceding tests should suffice to demonstrate why
it is so difficult to determine the cause of a particular
seed source interaction in a natural environment. For not
only do the seed sources interact with several factors of
the environment, but the environmental factors themselves

may interact with each other in their effect on the genotype.

OBJECTIVES

 

This study is part of a cooperative project entitled
"Tree Improvement through Selection and Breeding of Forest
Trees of Known Origin." This is regional project NC-51
of the United States Department of Agriculture and involves
active cooperation by the state experiment stations of ten
north central states. The regional project's objectives
are as follows: (1) Determine the range and pattern of
genetic diversity within selected forest tree species, (2)
utilize thegenetically most suitable material for breeding
purposes, and (3) provide genetically suitable material for
reforestation purposes.

The objectives of this study are as follows: (1)
determine the magnitude of seed source x environment inter-
actions over a variety of conditions in the North Central
United States, (2) determine the causes of these inter—
actions, and (3) determine the effect of these interactions

on a Scotch pine selection and breeding program.

19

METHODS
Material

Seed collected from ten average trees in each of 122
stands throughout the natural range of Scotch pine was sown
in the forest nursery at East Lansing, Michigan in the Spring
of 1959. Within each stand the parents were separated from
each other by 100 feet or more. The collection areas are
shown in Figures 1 and 2. Origin data are given in Table 2
for the 6% sources used in the present study.

Part of the seed was sown in a u-replicated randomized
block design to provide data oanenetic variation in seedling
performance. The remainder of the seed was broadcast sown
in large rectangular plots. Results of the seedling study
as well as a detailed description of nursery procedures have
been reported by Wright and Bull (1963).

In the spring of 1961, 2-0 seedlings from the broadcast-
sown seed were used to establish #1 permanent test plantations
throughout Michigan and the central states. The locations
of these plantings are shown on Figure 3. Most plantings
followed a randomized block design with from seven to ten
replications (Table 3).

Replacement stock was planted in the spring of 1962 to
fill gaps left by first-year mortality. These replacements
were easy to identify in the fall of 1962 by the stunted
.growth characteristic of Scotch pine in the first season

following transplanting. None of the replacements were

20

21

Figure 1. Natural distribution of Scotch pine in EurOpe
(shaded) and provenances in this study (num-

bered dots). (Wright and Bull, 1963)

 

0 has

Elva” ’
W“

 

   

 

 

 

5° :00 {5° 20°

22

Figure 2. Natural distribution of Scotch pine in Asia
(shaded) and provenances included in this

experiment (numbered dots). (Wright and

Bull, 1963)

 

I60 60
'-
f
, 1 I
/g’ ’ A].
. i, , «I : \
3 ,, \
I. _. ’1’\ I \ 50°
/ »’ ~
' fire/2 / i
‘(f ,4 ,1:
"-—‘ I, I.
\ 40°
4‘
IZO’ '30

x" 1
, / v“
"0 ;

 
     
  
     
    
  

     
   

    

.j____.x":\t\

\

\\

\‘i
‘w - \\\\\V
\\

I00.

100° "0' I20. I30‘

‘ 7t? 00’”

 

 

 

I‘LL-Burs 2"

23

10010 I. 8000 000110 I0000100 0000 - 3100000 by 003100.

 

 

110.100, I010! b“ Nov. legion. hr“ .00 .00.
0000017 lot. 100.. Country 1.00. In...
of 001010. 0! 011.10,
100100.. 000000.
u-dogrooo— 100'. m- 1".
0! 0!
{000 (000
1 ' I- 000 00.0 0.0 -
0010 000 00.0 101.0 00 000 001 «J ,4 .
01100 001 00.0 0.0 1 000 001 “J "' .
., 0- 000 00.0 1.0 10
000 -0: 00.0 10.0 0
0- 00.1 1.0 10
ma 200 00.0 20.0 1 p
001. 010 01.0 0.0 —
000 010 00.1 0.0 0
. 001. 00 00.- 0.- 10
000 010 00.0 0.0 0
I10 000 01.1 10.0 10
000 000 00.0 10.0 1 ‘ ’
000 0.0 01.0 10.0 1 0100 000 00.0 .J 00
000 000 01.0 11.0 1 100 000 000 100 «
000 000 00.0 10.0 1 ‘ '
000 000 000 100 0
' ' 01110 010 00.0 00.1 00
000 000 00.0 10.0 0 100 2" “J 01.0 .,
01.11 000 01.0 00.0 - 000 t" “J 00.1 00
m 000 01.0 00.0 00
101 000 01.1 00.0 -
000 000 00.0 00.1 00
000 001 01.0 10.0 0
us 011 01.0 00.0 00
"' 5" ”J 10.0 ‘ 0. 001 01.0 00.0 00
000 000 00.0 10.1 1
0010 001 00.0 00.0 0 I: g: 3'; :1 g:
010 000 00.0 111.1 00 ‘ ' '
010 000 00.1 00.0 10 i «3‘ 4,. 01 1
000 000 00.0 00.0 0 " g: :1: «,0 -0.0 00
000 000 00.1 4.8 1’
0001. 011 00.0 00.0 — 000 000- “J 4,0 00
0000 000 00.0 10.0 0 _‘
OH 803 “0. .0. ."
000 :01 00.1 11.0 —
000 210 00.0 10.0 -
cu 000 00.- 10.1 10
1:00 001 00.0 11.0 0
1:00 000 00.0 10.0 1
cu 000 00.1 10.0 00
1:00 010 00.1 10.0 10 '
000 011 00.0 10.1 10
01.0 012 . 00.0 10.1 00
0011 000 00.0 10.0 10
an 021 00.0 10.1 10

 

90:14:10.. 00000001000010. "Ila-d. M000,

0 “‘00 Ills-r1.

M
“0110, I! ~00 York, 00000,. 001000. ".0000, ”All. III“, 1'00.-
0l0vi0. UM! Ion-001.0.

P 30040 01100100! In. 111000001 0000‘“

211

Figure 3. Location of Scotch pine test plantations used

in the regional study.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

can 000 can 000 000
~
an . _ n...
. s u n .
.0 1
..-...+ ...... 5-..... u ........ _
J
71“) The 0 .1 1.
xx c. To» _ null!!!
\ z .
t. _ M.» a. _.
a 0 a. 0
. 0
. _ ‘ xx x 73
_ 0 __ _ w 0
_ 0‘ ’lIlIII
“ .109 . 70” \\/l Ill-LN]! I, 1.1! 0*
00¢ _ a J. 0
_ _ _ . ~0N.
\ I 5.17"”. ~11 _
— O .0.” I .
z . 11. 0n. o I III 1V /
’l—s.‘ 000 n .O-- O“. . fl —
_ o -3 .. 7,
“ 0.0 0.1”
4 \\. ,1- ._ .0. ~llllllll l _ Ill_
2 x . . 2 f I a
, a: .. 2... .0 . , 4
3.0 700 o a 7.» z I
I \ . TO. i Nd .\ II _
_ a __ To .0 i .0 _
. .1111r...‘ Pr N “NON ,
V1. \)ll/ II N 0 C J
a. x 0 I , . km» .. 73 n .3
0 \. \ L /lll/ Tn. .0-N I .D “ IIIIII I
M / u .N I
I, I; “-0.“ _
\ .1“ _ .OO“ .—
\ _ r 0’ 0I m ’
\1I(\ \ / .— l\ IMIH~ —.
.‘u. L)! s.
_ . I
.2. .2. .o. .8 .3 .8 .oo.

25

 

 

 

 

.Hopucoo owes HMOHEmaw .omaammm .Umucmaa unwgonm .omusmaa scum Soosopnnm \fll
mm S SS SSH S.SS S.SS SnocSHHH
.mepmzem Swansea HS1SS
mm S S. SS S.SS S.SS cmmwaoaz
.copcwzom copnmsom HSuSN
2 S S SS S.SS S.SS cmmwnowz
.mmmmmzmwnm mxmq mmom lema
2 S SS NS S.SS S.SS cmmanoaz_
.cmmmaa< cmmmHH< SSuaa
2 S S SS S.SS S.SS cmmasoaz
.Spomzmeo mama Scammam HSuSH
2 S SS SSS S.SS S.SS cmmwnoaz
SowSMsz owmmzmz Hmam
mo S SH SSH S.SS S.SS cmmanoaz .mmmo “mason mmsm Hmua
\Mmm S SH SSH S.SS S.SS cmmagoaz
l .oonmemamx pmmnom wwoaaox Hmum
amnesz mmmpmma
convex "poam\u mCOSu «mmonsom urmcoqu.pma.. mpmum .zwcsoo " mEmz upwnssz
wcfipcmam uwmmneu1mowaammu ommm upmoz "nufiozu u“ an:
.ucmEnmenmumm mo conuwe can .mnwm Scowumooa cowvmucmac mo mamEESm .m magma

26

were measured for this study. The trees planted in the
Spring of 1961 all made apparently normal growth in 1962
and 1963.

Climatic data for the eight plantings used in this
study are shown in Table H.

In general, the soils are coarse-textured ranging from
sands to loamy sands. The Allegan planting has by far the
coarsest—textured soil and is the most moisture deficient
site. The Edwards County, Illinois planting, on the other
hand, has the finest-textured soil.

Measurement

 

In the fall of 1962 height growth was measured on
eight plantations and needle length on five. In the fall
of 1963 height growth and needle length were measured on
four plantings. Color was scored in six plantings in 1962
and none in 1963.

Height growth was measured to the nearest centimeter.
Total height growth for all four trees on the plot was
recorded. If less than four trees were present, the height
‘growth was converted to a four-tree basis before recording.
In 1962 the height of the fastest growing tree on each plot
was also recorded.

One needle fascicle from the middle of the current
terminal growth was removed from each tree and measured to
the nearest millimeter. The total length of the four
fascicles was recorded. If less than four trees were present,

the total length was converted to a four-tree basis.

27

 

 

 

 

 

 

ll
T0010 0. Plantation cit-11c 1101.0. -
Plant-11011
kilo“ lint-0 Allegan mmoa lu-yn I000 lul 00000-00
M01. in. MI. Pore-t
\
Mel-00m: Ly--- 7.
April. 1003 00.0 01.2 01.0 30.1 42.0 00.0 00.4 9.8
.y. I.“ 0.. S... “.1 52.7 60.2 “.1 “.0 12.5
m. 1.0 “.0 0.2 “.0 50.0 60.0 01.0 ”.0 73.0
0.211. 1000 00.0 00.0 00.0 00.1 — -- - -
.y. 109 00.0 01.0 01.0 00.1 - - - -
Juan. "0 C... “.3 “.3 00.2 "' 1" '- —
Am] .000. s
In 40.1 43.0 08.4 30.0 «.1 41.8 40.3 '-
hrtod 501.0. fin
.10. 0101- d: _,._
33:'-0’°’ 100 110 100. 100 100 100 100 -
00.0.0000) 000 100 000 - 100 101 100 100 -
'3 '41.”) 101031 110 1014/ ‘ 110 - - - -
Precipitation; '-‘L‘-
1001 Mun-l . It." 40.10 34.” II.“ 34.2! 36.11 8...- H.“
I!!! Annual “.00 88.40 20.“ 80.05 23.15 20.“ 83.“ -
urn-M- 1"! 0.10 0.00 0.00 1.00 10.00 0.10 — 00.0:
”f“-m In .015 .0. O.“ '0' 5.“ ’0.l .0” .02:
Avril-JI— 1'0 0.01 10.10 0.00 0.10 -— -- —

I]

- “11001 VI hut-01' 000000 0000100 100001: lotto“ bun-cull lab I... 000-; lint-0 lab-Ital” uh;

Allen-411ml: 0000.0 "alumnus-muo- PM Akron; “no-h1g0 Duly Ibo; loco lab ‘00
III..." .010. 101-I; lu- brat-00.31001; 000-000-010". Ill.
- .00 00 011-00 "CI [011 "1100 not 011011.010.

28

Large samples of ten to fifteen fascicles per tree
were collected at several plantings in 1962 for comparison
with results of the single fascicle measurement. This com-
parison showed that a single fascicle per tree gave almost
identical results as obtained with a larger sample of fas-
cicles (See Appendix A). This was due to the small within-
tree variation in needle length.

Color was scored on the basis of ten color grades--
grade 1 being the yellowest in the planting and grade 10
the bluest. This scoring system permits within-planting
analysis but may bias between-planting comparisons. To
determine the constancy of color grades between plantings,
several samples were scored in the field and then brought
immediately into the office for comparison with Munsell
color charts. The following tabulation shows the entire
range of Munsell color grades and the field score of the

selected samples.

 

Plantation and date sample taken

 

 

 

Munsell : Higgins Lake Newaygo Russ Forest
Color : October 20, October 22, November 13,
: 1962 1962 ' 1962
Field Score l] '
2.5 Y 1 l
5.0 Y 1
7.5 Y
10.0 Y 2 2
2.5 GY
5.0 GY 6 6
7.5 GY 5
10.0 GY 9
2.5 G 9 9
5.0 G
ll 1 = yéllowest; 10 = darkest_green.

29

There isgood general agreement between plantings among
the greener shades. However, there are indications that
the observer recognized fewer shades of yellow than of green.

The color scorings were made over a fairly wide range
of dates. However, subsequent analyses showed that the date
of scoring was relatively unimportant. Because of a heavy
snowstorm in early December it was not possible to score
the plantings when color differences would be at their
peak-~about mid-December.

One special color scoring was made the first week of
December. This scoring was designed to provide precise
between-planting comparisons with the effect of date of
scoring held constant. Samples of nineteen sources common
to three plantings were collected. The needles were removed
from each tree, stapled onto a wide card with plot identi-
fication on the back, and then brought into the laboratory
for direct comparison under uniform light. Collection and
scoring were completed within two days. Using this system

it was possible to recognize nineteen color grades.

30

Analysis

The number of seed sources used in the analyses was
determined by the number of seed sources the measured
plantations had in common. There were #2 seed sources common
to all 7 plantings measured in 1962; 55 seed sources common
to the u Michigan plantings measured in 1962 and 1963; and
57 sources common to the six plantations scored for color.

Data for each characteristic and plantation were indi-
vidually subjected to analysis of variance using plot totals.
Plantation analyses were then grouped in various combinations
and tested for treatment differences and seed source x
plantation interaction.

Such combined analyses may be Open to question because
of possible heterogenity of error and interaction variances.
Thus the Edwards County, Illinois, plantation was not in-
cluded in the combined analyses. This plantation had two-
tree plots whereas all the other plantings had four-tree
plots.

Excluding the Edwards County planting, treatment differ-
ences in the combined analyses were so much greater than the
interactions that there is no doubt as to their significance.

The "conservative" approach of Cochran and Cox (1957)
was used to test the significance of interactions. That is,
if one of the plantings had a much higher error variance than
any of the others, the tabular F-value used was that with
(s-l) and n' degrees of freedom--where s is the number of

seed sources and n' the number of error degrees of freedom

31

in the plantation with the highest error variance. There
were no instances in the entire study, however, where this
conservative approach changed any conclusions as to signi-
ficance of interactions.

Significance of differences between plantations was
determined by the "t-test" since the combined analyses do
not provide a valid error term for comparison of plantation
means.

Components of variance were determined by setting the
computed mean squares equal to their expected mean squares
(Tables 5 and 6) and solving for the components.

Correlations between plantations, using seed source
totals as items, were made to determine the degree of

similarity between plantings.

32

101110 5. mly010 ot 001-10000 ton 13011 101- :00 0001111100 01:11010 of 0
0111310 y00r'0 r00u110.

 

Soure0 o! nr10tton 9031-000 0! P0r000:01-0 0001-:00 by .00 0am"

 

1mm:-
. . 3 3 3
P1011:0:1011 (9-1) 6". 0:5. 506',
2 '3 3
80011 0ourc00 (0-1) 6'. 0:0}; $50",
Pl0n:0:1on0 X a 2
800d 0ourc00 (p-l)(l-U O". +- p.
3
Irror [:(rv-l )] (0.!) d".

 

p - 111.110: of ”manner: 0 a 111-b0r o! 0004 0ourc00; :- - 01-110: of ml1c0:1on0
’01- pl011:0:1011| 1' - tr, fir, ’

6.: 10 :110 eta-10011011: o! v01-100c0 11110 to nude. v0r10:1o00 01:11“ '10“ 01:1110
phantlonl.

d": 10 :110 couponont 01 000101100 11100 :0 1111101111000 Ian-0011 11100t0t1000.

6": 10 :210 130.0000: 01' 001101100 (1110 :11 1111101011900 00:0000 0000 001111000.

6;: 10 :110 eta-110000: of 001-10000 11110 :0 :110 111:0r0ct1on of 01000011000 004 00011

IOWCOU 0

701110 8. Amly010 or v0r100c0 torn 11006 for 1100 0001111104 0-1y010 o!

y00rly "011110.

 

Sour-1:0 0! 00r10t1oa D0¢m0 o! P0r0-0:0r0 00:1-:0d by .00 unfit-00

 

{mac-
2 3 3 T 7
You-0 (y-l) 6', 070;” airpd'; 4706;, +£5.00",
- 3 3 3 3 3
Pl0nt0t1on0 (pl) 6". +?O;’. +Fy6;. 0'06;’ 5’06;
Y00r0 X 2 2 2
Pl0u:0:10n0 (y-l)(p-1) 6'. 0.6,}. 5:06.;P
3 ‘ 3 3 3 3
80011 0011rc00 (0-1) 6". 01%;” +Fy0;. 1136;. +3135»:
8004 00111-c00 x 2 _ 2 2 I
Yuu (0-1)(y-l) o", ”6;” +£rpq'
800d 00111-1100 x 2 _ a _ a
Pl0n:0:1on0 (0-l)(p-l) r. ”6;” 0nd}:
8001! 0011:1300 X
P100:0:1on0 x 2 _ 2
1001-0 (I-IHD-IHV-l) o". 014;"
3
bro: [fin-’48 (04} r.

 

y - 01-b0r o! ’00"; p - 111.1101- 101 pl011:0:1o=0; 0 - 01-h0r of 00011 0011:1100;
rp . manor of npl1c0:1on0 p0: 0100:0t1001 r - trp / Ir,

10 :h0 man: of 1101101100 6100 :0 20000. v0r10:10110 01:1110 010:0 01:11111
pl0n:0:1on0.

10 :110 «.0000: 01 001-10000 :0 «11110000000 50:0000 y00r0.
10 :b c.0000: of v01-10m0 :0 011100004000 00011. 0100:0“000.
10 "10 «1.011011: 01 v0r1011c0 :o d1110r00c00 0000. 00011 MI.
10 :110 own: of 00r100°0 :o 10:01-0c:1011 of mu 00¢ p100:0:1on0.

10 :00 c.0000: 01 0011100110 :0 10:0r0cuol1 of 1001-0 0011 0000 001111100.

33????

10 :00 c.0000: 01 001-10000
0000 00111-1100.

:0 111:0r0c:1ou of 9100:0“000 00¢

\

i

10 :b man: of v0r10000
01111 0000 low-col.

o 10:0r0c:1ou of y00r0, phnuum,

iqu fin§fiu3x .m‘qwfi 00 I3 00

' RESULTS

In the presentation of results the seed sources are
grouped according to the geographic ecotypes delineated by
the three-year nursery measurements of Wright and Bull (1963).
It should be pointed out that these ecotypes were determined
on the basis of a multi-character analysis. Thus these
,groups may not conform exactly to a grouping based on height
growth, needle length, or color alone.

Height Growth

Sources from Western West Germany and Belgium (Group H)
made the best height growth throughout all the plantations.
This group made nearly twenty percent more height growth
than the average tree in the planting (Table 7).

The Southern East Germany and Czechoslovakian sources
(Group G) made the next best height growth over all planta-
tions. This group grew from ten to fifteen percent faster
than the plantation average.

Group F was slightly behind Group G in overall growth.
This group, probably because it was represented by only two
Polish sources, showed much variation between plantings. The
differences in height growth between groups P and G are
probably not real.

The percentage by which group H and G exceeded the
plantation average remained nearly constant for all planta—
tions and years of measurement.

The Spanish sources (Group N) made the least growth

33

311

101110 7. --Ib1uht growth 01 Scotch 111110 00001 00111-000.

 

 

 

FFETSFi

Country [0110113 01103011 ”[3100 lbuflhton Ill. h,” m u. 00001-0.
01 orla1n. Faro“ 1.0110 Poul:

BK; 310. 1932 1963 1203 .12“ 1093 1203 1203 1003 [”3 1.“ a "2

3...": 01 9100101100 2::

 

 

 

c W120, 70,3 «.1 10.0 00.1 00.0 10.1 10.0 10.0 00.0 01.0 02.1 00.0
s... 222 95.3 00.2 00.0 90.0 03.0 01.0 100.1 92.0 11.0 00.1 00.1 00.0
FIN 230 09.9 51,2 “.3 01.3 07.0 33.. 33.3 “.0 - - — —
Rm 273 13.3 “0. ”01 .200 ”0. 1.0g ”0‘ .0. “.0 1W.° 1.0, ”.7
10011210 93.1 11.1 11.0 11.2 00.0 71.0 02.0 07.3 00.3 00.0 10.3 101.1
s“: 521 92.1 00.1 10.0 00.0 120.2103.0 101.3 30.0 ”.0 103.0 00.0 00.3
511525 000 00.5 01.1 00.0 113.0 01.0 00.0 00.11 00.0 01.0 100.0 00.1
3111:5210 96.2 00.0 00.1 ”.1 02.3 00.3 00.0 00.0 11.2 00.0 01.4 00.1
501: 503 10.1 19.1 70.3 73.2 100.1 00.0 00.3 01.0 07.3 ”.0 101.0 103.3
5"; 54‘ .301 ”0‘ “0‘ .110 ”0' .‘0. ”I, ”01 ”0. lu0, .10: .10.
501.505 02.0 01.0 10.0 02.0 100.0 31.0 00.0 02.0 .. .. ._ _
AvorIl'o 35.0 "0. “a: ”0. ”0' ”0. “0, “0° ”0. ”0’ '0. .g.

0 M1” 223 01.2 00.0 00.1 01.1 107.2 103.0 110.3 103.0 112.1 00.0 ”,1 121,3
[AT 320 ”.0 10.0 00.1 01.0 00.1 01.0 101.1 01.0 ”.2 00.0 01.0 111.3
50: 501 110.1 100.0 99.0 100.0 130.3 110.1 110.0 05.0 100.0 00.3 101.3 115.0
S012 302 110.1 00.1 00.0 00.3 110.0 103.0 ”.1 101.3 03.3 00.3 100.1 ”.0
$01: 330 00.1111.0 00.0 00.0 00.1 00.1 01.3 00.0 — .. .. _
Avorngo 102.1 00.0 03.1 01.1 100.1 102.0 103.0 07.1 00.3 00.0 100.0 110.3

x 519 221 00.0 03.0 09.0 03.0 110.1 01.0 105.1 03.5 .. .. .. ..
1311250 101.3 03.3 111.3 103.3 110.0 01.1 110.0 00.3 100.1 101.0 01.0 100.0
Avenue 00.0 00.3 100.3 03.1 110.0 00.0 110.2 00.0 100.1 101.0 01.0 100.0

1 001 211 110.3 110.0 110.3 110.0 00.3 110.1 113.0 111.0 112.0 111.1 130.0 110.0
001 311 112.1 105.0 130.0 119.3 113.9 113.1 103.1 110.5 130.0 131.0 111.3 131.5
’Avernxe 113.2 101.9 123.0 111.0 105.1 111.1 101.0 111.1 110.5 110.1 130.0 131.5

a GER 202 111.5 119.0 90.3 100.3 100.1 100.0 109.0 113.1 100.0 110.3 103.1 101.0
can 200 95.5 100.5 93.0 05.0 09.0 103.0 00.0 00.1 100.3 00.0 01.0 00.0
fill! 200 122.0 116.0 111.0 122.1 100.0 120.7 100.3 101.3 '- - - -
can 210 120.1 119.0 115.2 120.0 100.0 101.9 110.1 109.0 110.0 100.1 100.2 .0
ex: 305 110.0 123.1 120.2 125.0 103.1 103.3 109.3 105.0 - -— -. ..
czs 300 95.1 100.0 130.0 100.0 91.0 100.1 111.0 120.0 105.1 110.1 113.0 03.1
C2: 301 111.3 113.0 100.0 135.0 119.0 111.3 101.0 101.0 -- - —- ..
czs 300 103.3 111.0 135.0 109.1 129.5 111.0 00.3 110.1 - -— - .-
cza 309 113.0 113.0 90.1 91.0 95.9 105.3 113.5 100.3 110.0 105.0 101.0 00.3
can 310 130.3 133.0 110.1 113.1 103.0 130.0 131.9 113.0 135.3 115.0 105.1 .0
r2: 311 121.0 113.3 111.0 119.0 90.5 101.0 130.1 115.0 - - - -
oz: 312 131.9 111.0 120.0 121.0 110.5 130.0 110.0 100.0 110.0 128.1 111.3 119.0
can 525 110.0 115.0 122.1 119.0 130.1 126.0 120.1 112.0 104.9 113.1 100.0 100.0
can 521 109.1 100.0 100.1 131.1 120.0 130.5 110.1 125.1 131.0 109.3 100.0 111.0
Avor0go 115.1 115.5 119.1 110.9 111.1 113.3 111.0 111.1 110.1 100.0 111.0 103.0

M can 200 100.3 109.5 93.9 102.5 11.1 111.1 103.0 110.1 119.0 00.1 103.0 100.0
rug 2.. 119.2 131.9 119.1 119.0 131.9 129.0 119.0 110.0 - .. .. ..
can 251 129.3 131.1 125.0 119.0 100.0 130.0 150.1 135.1 100.3 105.2 105.1 110.5
53“ 253 110.2 130.2 133.1 129.0 133.0 120.0 109.3 115.5 122.1 139.1 133.0 100.1
BEL 313 125.5 135.5 132.2 130.1 131.0 129.2 133.1 125.0 132.1 100.0 123.2 139.1
33L 530 132.0 125.9 119.1 121.0 09.3 100.9 111.1 130.0 136.1 111.5 151.0 100.0
nun 553 110.5 90.1 119.5 110.1 130.3 119.3 120.2 133.5 -— -. .. ..
xv 225 100.0 90.1 100.0 101.0 131.1 100.3 113.3 100.5 131.0 90.2 01.1 100.2
Average 111.5 120.0 119.0 110.0 110.0 111.9 121.0 130.0 130.5 131.0 135.0 110.0

J rnA 235 90.1 121.0 110.1 113.3 90.0 90.1 91.1 111.1 93.3 110.0 110.9 131.0
100 202 90.5 100.0 100.1 03.0 01.0 90.1 93.0 90.0 110.9 120.0 115.2 00.0
Avcn‘o 00.0 110.0 103.1 00.0 02.3 00.1 03.3 103.0 102.1 121.0 113.1 110.1

K TUR 210 90.9 00.0 09.0 00.0 03.3 99.3 13.0 03.0 15.1 00.0 00.3 00.1
101 220 00.1 05.0 05.0 00.3 11.0 00.0 00.0 100.0 101.0 00.1 00.0 01.0
TUR 221 10.0 93.0 01.1 00.2 53.1 10.0 03.5 10.3 01.0 11.3 103.0 101.0
can 203 03.0 09.3 11.1 01.3 19.1 05.1 03.0 10.0 93.0 100.9 10.1 100.0
can 211 91.0 92.0 103.1 100.0 100.0 109.0 111.0 113.1 .- .. .. ..
on: 551 99.9 103.0 19.0 01.0 00.0 05.0 01.0 90.0 -- - - -
Aver-go 05.1 91.0 .05.0 00.5 10.1 00.5 05.0 01.3 00.0 09.1 00.0 90.2

I SPA 210 73.1 33.3 72.3 32.0 22.3 33.7 70.3 25.3 “.3 In“ 13., .1"
SM 210 78.0 ”.0 05.5 ”.0 02.0 77.7 ”.3 00.1 ”.0 00.0 00.0 ”.3
500 205 10.5 19.9 13.5 10.3 55.1 12.0 10.0 91.0 01 3 00.3 01.1 00.0
SPA 206 01.1 91.3 90.0 92.0 55.3 77.3 11.1 03.0 50.0 . 01.0 51.0 01.0
590 301 13.3 19.1 05.3 11.3 10.9 19.3 00.1 01.0 11.0 03.0 53.0 00.3
Avorlgo 11.0 10.0 01.0 10.0 00.3 01.0 10.1 05.1 15., .._° ‘3', ,3_.

Ivan p10n10110n

M13111 arc-1h
(cenu-ueru)

13.01 22.10 0.00 12.00 11.91 10.50 12.39 30.22 13.05 0.00 10.05 7~"

Standard error

of srod source °

.00n0 30 a per-

cent 01 p10n100

llon noun 6.00 3.03 0.20 0.31 3.17 0.33 3.30 3.01 3.18 7.33 3.11 .'32

 

1/
- 0151411110. czzchoslonhu. 11111-110. 11111100. Git-00y. 001000. 110130". MTV/10, my, P0101101, 81hr“. 0P0“,
501201011. Ylno010v10. UMI bunutm.

35

over all the plantations measured in 1962, the second year
following planting. These sources were the slowest growing
group in every planting except Russ Forest. At Russ Forest
the Spanish sources made twenty-five percent less growth
than the plantation average but were equaled in slow growth
by sources from southern Scandinavia (Group C).

In 1963 the picture began to change. Both the Spanish
sources and the Greek-Turkish sources (Group K) did better
relative to the plantation mean at every planting in 1963.

At the same time the southern Scandinavian group and the

two sources from western Siberia (Group E) all did relatively
poorer in 1963. At Kellogg Forest the growth of the southern
Scandinavian and Spanish groups were equal, while at Houghton
(the northernmost planting) the Spanish sources outgrew the
southern Scandinavian sources. That this change is real is
indicated by the significant seed source x year interaction
term in Table 8.

Kellogg Forest, Russ Forest, and the Houghton County
plantations made the best overall growth in 1962. And while
all plantings made better growth in 1963 than in 1962, this
yearly acceleration in height growth was not the same for
all plantations. Kellogg Forest made 1.8 times as much
growth in 1963 as in 1962, Higgins Lake and Houghton both
_grew about 1.6 times the 1962 growth, and Allegan grew about
1.3 times the 1962growth. That these differences in growth
acceleration are real is shown by the highly significant

year x plantation interaction in Table 8.

36

Table 8. Combined analysis of variance results of 1962
and 1963 height growth for 55 Scotch pine seed
sources from four Michigan plantations.

 

 

:Degrees: :Component of Variance
Source of : of : Mean : as a percent of
Variation :Freedom: Sguare : total variance
Years 1 726951.83** 8.17
Plantations 3 171288.05** 28.27
Seed Sources 58 7853.02** 31.33
Years x Seed **
Sources 58 500.27 l/ 1.61
Years x Planta- **
tions 3 u920.85 l/ 2.88
Seed Sources x **
Plantations 162 #99.56 l/ 0.7%
Seed Sources x
Years x *2/
Plantations 162 191.u7 _ 0.58
Error 329k 160.32 26.u2

 

Significant at the five percent level.
Significant at the one percent level.

1/ Based upon seed sources x years x plantations interaction
as error term.

2/ Based upon seed source x replication—within-plantation as

error term.

37

The Edwards County, Illinois plantation was not in-
cluded in the combined analyses. An examination of the
data, however, suggests that ecotypes did not behave markedly
different at that plantation. The means of groups C (a
northern group), H (a central European group), and N (a
southern group) are identical at Edwards and Houghton, two
plantings which are about 575 miles apart in a north-south
direction.

Group G, from southern East Germany and Czechoslovakia,
seems to differ markedly at Edwards. This group grew about
ten percent less at Edwards than at any other planting.

There seems to be no explanation of why several of the
sources in this_group did so poorly while groups H and F,
also from central EurOpe, did as well at Edwards as at any
other planting.

The individual seed source x plantation interactions
were random. That is, there was no relation between location
and elevation of the seed source and its performance in a
given plantation. For example, sources MSFG 2H3 and MSFG
271 are both from Greece and from the same latitude, longi-
tude, and elevation.- They made almost identical growth at
Kellogg Forest in both 1962 and 1963. But at Allegan,
Higgins Lake and Houghton they seldom exceeded the plantation
average by more than fifteen percent.

Source MSFG SHS, a high elevation source from central
Sweden, made its poorest showing in both years at the

northernmost plantation--Houghton County.

38

In 1962, between-seed-source differences accounted for
28 percent of the total variation encountered in the Michi-
gan plantings (Table 9). This drOpped to 19 percent in
1963 as differences between plantings became more pronounced.

If the variance component due to plantation differences
is not considered, the seed source component consistently
accounted for between 32 and an percent of the remaining
variance in both years.

The differences in growth acceleration greatly increased
the percent of variance due to plantation differences in
1963. For the four plantings measured in both years the
percentage of total variance due to site differences increased
from 20 percent in 1962 to 52 percent in 1963.

The plantations were grouped into a number of combined
analyses. This was done to determine whether any particular
plantation was responsible for most of the interaction. Such
was not the case. The seed source x plantation interaction
for the #2 sources measured at the seven Michigan plantings
in 1962 was significant for any combination of three or more
plantings (Table 10). For example, neither the combined
analysis for the Kellogg Forest-Allegan plantings nor for
the Kellogg Forest-Russ Forest plantings gave a significant
interaction. But when these three plantations were combined,
a significant interaction was found. Furthermore, the percent
of total variation due to interaction was the same for these
three closely situated plantings as for all seven Michigan

plantings.

39

.NIm
pom Hum moanme .m xfiocoma< ca czonw mommpcoopoa oceanopoo 09 cows moundom coo: \H
.Ho>mH Humanom mco may um ucmoflmwcmwm I «a
.Ho>oH ucoonoa o>wm any um pancamwcmwm I «
pQMOfimwcmeIcoz I mz
.mwcwpcmam
cwnpfia mvoam coozpon wcowPMflsw> Eoocmh ow mop moonwpm> mo pcocomeoo on“ ma m o

.mwcwycmHm cam
moomsom comm mo COflpompovcw may Eopm mcflvaswoa ooCMw9m> mo “cocoaeoo 0:9 we on o

mu.
.mmonsom comm cmospon wmocmhommwo ow map oUCMflhm> mo unocomaoo on# ma filo

N

o

 

.mopflm mafipcmam coozuon moocmpommwo op map wocmwpm> mo vcmcomeoo may ow .19
IF

cougwsom
IflQIIMMWIQGIIfiflfiwfimflmflMPuhMWméMIIIfiflfimflfl&
, counmsom

Imwwm I .. WWW mesa. I IewmueI mem.mmI eemmmI were. I Mem.m I I I Immem meMmmem

mxmq mcwmmwm
Hmonoh MmeHoM

.HNI m

i
5
k9
I
I

Imbmh I I .m...Mml Rab...» l lukzhbl bfbfiml fir».me h.
, I . . I covnwsom

mxma mcwwmfim
IfiEIIhsfistsIIfififiebfifuaafufisfiméfiIIIfimfimflflé

 

covawsom
ammoaa<
mxmq mawwwfim
~.o~ :.oe ee:.m ««H.m «em.ma ee:.m~ «so.Hm «em.o~ umoeom mmoflaox
mocmwpm> Hm90# mo unmouom .
meme . wood meme Nome . mama . mama . mama Nome” mewsHmsm ooeeoeoo
mro mole mIo oIo « on mcoeumusmao
N. N N u

 

“accomEop mocmmnm>

_ _f.mooh30m comm mm EonMIImoamwam>
amuOP mo “cochmm m mm mocmwpm> nvsopw unmwmn mo mucocoasoo mmma can «mmH .m magma

#0

Table 10. 1962 components of height growth variance
expressed as a percent of the total variance of

#2 seed sources.i

 

' Variance Component

Plantations in 2 _2
Combined Analysis: 0‘? ' 0‘s "‘ O’ps ‘ ‘ 0 E

Percent of Total Variance

Kellogg Forest 16.67** 28.17** M.6l** 50.55
Russ Forest

Allegan

Newaygo

Higgins Lake

Rose Lake

Houghton

Kellogg Forest 22.90** 27.7H** H.77** nu.59
Russ Forest

Newaygo

Allegan

Houghton

Kellogg Forest 21.83** 28.73** n.1u** #5.30
Russ Forest

Allegan

Houghton

Kellogg Forest 28.16** 26.u2** u.99** #0.H3
Allegan
Russ Forest

Eeil'é'g'g 50:35; " " 3.30%" 3573's}? ' 1:39;." " 3571+? " " '
Higgins Lake
Houghton

Kellogg Forest u3.10** 25.70** 0.00NS 31.20
Allegan , I I - - ' 3

Kellogg Forest l8.09** 25.97** 8.u8** u7.M6
Rose Lake . . . . . . . . I _

Kellogg Forest 3.02NS 38.58** 2.03NS 55.31
Russ Forest - . -

Kellogg Forest 2.HONS 35.90** 1.72NS 59.98
Houghton 8 ~ - ~ -

Kellogg Forest 7.09** 33.7u** 6.76** 52.ul
Higgins Lake - - - - . -

o. no
N

Continued

#1

Table 10. (Continued)

 

Variance Component

Plantations in : 2 ’ 2 ' 2

Combined Analysis-

2
OjP 0's ' fps O-E

 

Kellogg Forest 19.69** 30.13** 3.20** u6.98
Newaygo

 

is the component of variance due to differences between
planting sites.

is the component of variance due to differences between
seed sources.

is the component of variance resulting from the inter-
action of seed sources and plantations.

is the component of variance due to random variation
within plots within plantations.

Non-significant.
Significant at the five percent level.
Significant at the one percent level.

Mean squares used to determine percentages shown in
Appendix B, Table B-3.

02

As may be seen from Table 11, which shows correlations
between 1962 seed source totals for all combinations of
plantings, there is no relation between size of correlation
coefficient and size of interaction. The highest correlation
coefficient shown (.82) is between the two plantings with
the highest interaction component--Kellogg Forest and Rose
Lake. The coefficient of determination between Kellogg
Forest and Rose Lake is (.82)2, or .67. This leaves 33
percent of the variance to be divided amongst interaction
and experimental error. Whether a significant amount of
interaction is present depends upon the size of the experi-
mental error.

Needle Length

Sources from central Europe (Groups F, G, and H) had
the longest needles and sources from the southern extremes
of the Scotch pine range had the shortest needles (Table 12).

Group C, from southern Norway and central Sweden had
relatively longer needles in 1963 than in 1962 (note signi-
ficant seed source x year interaction in Table 13). However,
these sources still averaged intermediate in needle length
for both years.

At the Houghton plantation the Spanish sources consist-
ently had longer needles than at any other planting. The
Scandinavian sources, on the other hand, tended to have
relatively shorter needles at Houghton. As a result, in
1962 the Spanish sources were longer needled than the Scan-

dinavian sources at Houghton while in 1963 the Spanish,

H3

Table 11. Between-planatation correlation coefficients
using the mean 1962 seed source height growth.

 

4.:
(.0
o 0
LI 44 .54
o m m
m o A m
a c .x
m o c o m o m m
m In m m; c u .4 o
O U.) >3 ~r~l ,c: LI
H U) (I) ‘6 b8 b0 (1) TU
H m .a 3 m) s m 3
o 5 ca m -H o c U
x m < z m m m m
Kellogg Forest 1.00
Russ Forest .78 1.00
Allegan .80 .77 1.00
Newaygo .81 .66 .80 1.00
Higgins Lake .72 .67 .68 .65 1.00
Houghton .81 .80 .80 .72 .77 1.00
Rose Lake .82 .78 .85 .70 .59 .75 1.00
Edwards .70 .63 .67 .72 .57 .68 .73 1.00

 

All values are significant at the 0.1 percent level where
r = .36 with M0 degrees of freedom.

1+0

 

 

 

 

$01. 10. M. m o! loot“ '1. .000 m.
l
m.
Gantry 0.110“ 011.- m m 0.0
o! "1.1-, m1 lab I.”
”10 lo. 10” 10” 10” 10! 10” 10” 10” 10” 10”
:_____: .‘ ’um‘m W
1
c In :01 00.9 01.0 “J 01.0 5.9 .J 00.1 ”.0 ”.0
8" m ”a. "a. ”I, l“.. ”a. .0, .0. .0. ”a.
'1. m "0' .o. "a. .0. ”.0 ”.1 .0. ‘0. -
II 010 00.0 1”.0 01.0 1”.1 00.0 ”.0 ”.0 ”.0 ”.1
II 010 00.0 ”.1 ”.1 01.0 ”.0 00.0 10.0 00.0 ”.0
m ”I .‘o. 10": ”0' 1“.. “0! '0‘ .0. .0, "0.
m m .l-s l“.. “.1 l“.. .0, .00 .0. ”0. "a.
0'0 000 ”.0 1”.0 100.1 00.0 01.1 ”.0 01.0 ”.0 ”.0
III 000 ”.0 110.1 ”.0 ”.0 00.1 01.0 ”.1 101.1 ”.0
"I 000 00.1 100.0 01.0 01.0 00.0 ”.0 ”.1 ”.1 ”.0
000 000 00.0 ”.0 00.0 1”.0 00.0 ”.1 01.0 ”.0 —
0100‘” no. two, ”.0 .0. ”o. .10, “a. “o. “0’
0 MT 000 ”.1 00.0 00.0 00.0 100.0 ”.0 100.0 ”.1 ”.0
MT 000 ”.0 ”.1 00.5 ”.0 00.0 ”.0 00.1 101.1 ”.0
m 001 100.1 100.0 100.0 100.0 100.0 100.0 101.0 1”.0 101.0
m 000 00.1 110.0 100.0 1M.0 00.1 110.0 00.1 100.1 ”.0
III 000 ”.0 100.0 100.1 100.0 00.0 00.0 1”.0 101.0 -
0101‘“ ”cl lab. ”0. l”.0 1‘0. l”.. I... :as. “o.
m 000 ”.0 100.0 100.1 100.0 00.0 1”.1 01.1 ”.1 00.0
Avon" 100.0 110.0 1”.0 111.0 1”.0 101.0 100.1 100.0 ”.0
l' ’01. 011 110.0 100.0 100.0 110.1 100.0 110.0 1”.0 100.0 101.0
RI. 011- 110.1 101.1 110.0 100.0 100.0 110.0 1”.0 110.0 100.0
Avon" 110.0 1”.0 111.0 101.0 1”.0 110.0 100.1 1”.0 100.0
a an no: 111.. 100.: ' 100.: 100.0 I 101.1 112.0 100.0 100.0 7 110.0 7
”I 000 101.0 00.0 100.0 100.0‘ 1”.0 ”.1 100.0 ”.1 00.0
on 000 110.0 101.1 100.0 100.0 100.0 100.0 110.0 100.0 -
an 010 100.0 100.0 110.0 110.0 ‘ 100.0 100.0 101.1 110.0 100.1
C10 000 111.0 110.0 ‘ 110.0 100.0 1”.0 111.0 ”.0 101.1 -
CZ! 000 100.0 101.1 100.1 100.0 . ”.0 100.0 110.0 1”.0 1”.0
ca 301 114.0 100.1 111.0 110.0 1 100.1 110.0 110.0 1”.0 —
ca 000 100.0 100.0 111.0 100.0 110.1 110.0 110.0 100.0 —
ca 000 100.0 100.0 101.0 00.0 1”.0 100.0 100.0 100.0 1”.0
ca 010 113.1 100.1 . 100.0 1”.1 . 110.0 110.0 00.0 1”.0 111.0
C" all ‘18.! lu.’ ‘ ‘uo. ”I. { ‘1... lac! m.’ l“.. -
010 010 110.0 110.1 . 1”.0 100.0 ‘ 110.0 1”.1 1”.0 1”.1 110.1
out 000 00.0 01.0 100.1 01.0 . 100.0 1”.1 110.0 1”.0 100.0
an 001 00.0 01.0 ’ 1”.0 1”.0 . 110.0 1”.0 00.0 101.1 ‘ ”.0
I
Am no.a 1a.. 100.. 101.0 ‘ 101.. 100.1 101.. 1a.: 1.0..
I m 000 111.0 100.0 110.0 110.0 1 101.0 100.1 101.0 101.1 111.0
I'M 001 100.0 100.0 100.0 110.0 100.0 110.0 110.1 101.1 —
an 001 111.0 100.0 110.0 111.0 111.0 100.0 , 100.0 110.0 1”.0
(II 000 110.0 110.1 100.0 111.0 110.0 100.0 111.0 110.0 101.0
”I. 010 101.0 100.0 100.0 110.1 I 110.0 110.0 110.0 111.0 101.0
”I. 000 100.0 110.0 110.0 110.1 110.0 111.0 111.0 110.0 100.0
In so: 101.0 ".1 101.: 100.: * 114.1 111.: 110.0 10.0 -
I! 000 100.0 00.0 1 100.0 ”.1 111.0 1”.0 100.1 1”.1 —
b
Avon” 110.0 101.1 ’ 111.0 110.0 I 110.0 111.0 111.0 110.0 1”.0
J m 000 100.0 100.1 ’ 100.1 ”.0 ”.0 ”.0 110.0 1”.1 ”.0
m 000 100.0 ”.0 1”.1 01.0 ‘ 100.0 1”.0 1”.0 1”.0 110.0
Avon" 100.0 00.0 101.1 01.0 { 100.0 ”.1 1”.1 1”.1 1”.0
I m 01: 00.0 11.0 01.0 00.0 . ”.0 ”.0 00.0 00.0 ”.0
m a” ..‘ ”.' ”a. ”a, ‘ no. no. lo... .0! “0'
m 001 ”.0 00.0 01.1 ”.0 ”.0 10.0 ”.0 00.0 ”.0
m 000 00.0 01.1 ”.1 10.0 ”.0 01.0 ”.0 10.0 10.0
m an "o. “o. ”to "0. "u. “a, '0' .‘ol -
”I 001 ”.0 10.0 ”.1 ”.1 01.0 10.0 11.0 ”.0 -
Avon.- 00.0 00.0 01.0 00.1 ”.0 ”.0 00.0 ”.1 01.0
0 000 010 11.1 11.0 " 00.0 ”.1 ”.0 00.1 ”.0 ”.0 11.0
SPA 010 10.0 ”.1 j ”.1 00.0 01.0 11.0 101.0 00.0 10.1
an 000 ”.0 10.0 ”.1 ”.0 ”.0 ”.1 ”.0 ”.0 ”.0
000 000 ”.0 10.0 77-. ”.0 10.0 11.0 01.1 ”.0 10.1
an an an 10.0 00.0 00.5 01.0 10.1 0.1 01.0 01.!
O
‘w ”o. "a. . “0° “0° “0. ”I. u.‘ ”a. ”o.
l ‘ I
r 0
ha- nut-un- ‘
noodle 10ml ' '
(31111-1020) .
01.01 00.00 00.” 01.00 01.00 00.00 00.00 01.” 00.01
Ito-fire onu-
01 so“ we.
”II as o prmt
of phluuu ”II

0.” 0.17 0.00 0.10 0.00 0.” 0.00 0.00 0.00

 

1/ _
- ”utu, enemlonua, "Ila-d, was, my, ”loco, “I", firm,
my, man, 0100:“, IPA“, 81000., mung, m1 “101..

,_7W.__...__.__!__ .... .__. -__- _ _

M5

Table 13. Combined analysis of variance for 1962 and 1963
needle length of 55 Scotch pine seed sources
grown in four Michigan plantations.

 

 

Source of : Degrees : : Component of Variance
Variation : of : Mean : as a percent
: Freedom : Square' : of total Variance

Years 1 5518.83** 7.90
Plantations 3 3185.03** 7.27
Seed Sources 50 273.93** 12.78
Years x Seed

Sources 58 15.19**1/ .73
Years x Planta-

tions 3 2170.1u**l/ 31.05
Seed Sources x 1/

Plantations 162 12.HO*_ 1.36
Seed Sources x

Years x

Plantations 162 8.88**3/ 2.78
Error 3186 5.36 36.08

 

** Significant at the one percent level.
* Significant at the five percent level.

1/ Based upon seed sources x years x plantations inter-

action as error term.

2/ Based upon seed sources x replication-within-plantation
as error term.

H6

Greek-Turkish, and Scandinavian groups did not differ in
needle length.

At the Kellogg plantation, however, the opposite trend
was observed. While in 1962 the Scandinavian sources were
intermediate in needle length, in 1963 these sources were
nearly equal to the long-needled German-Belgium-Czechoslo-
vakian sources.

Thus there was a tendency in both years for sources
from the southern part of the range to have relatively
longer needles in the northernmost plantation, while sources
from the northern part of the range had longer needles at
the southernmost plantation.

There were large year to year changes in the mean
plantation needle length. Houghton, which had the lowest
mean needle length in 1962, had, with Higgins Lake, the
highest mean needle length in 1962. While the mean needle
length increased at Houghton in 1963, it decreased at all
the other plantations. The ratio of mean 1963 needle length
to mean 1962 needle length was .70 for Kellogg Forest; .82
for Allegan; .94 for Higgins Lake; and 1.11 for Houghton.
That the differences in these ratios are real is shown by the
significant plantation x year interaction (Table 13).

The amount of variation due to seed source x plantation
interaction (Table 18) in 1962 was small compared to the
variation due to seed source differences. However, this

interaction component showed a definite increase between

07

moanmh .m xHocomo< cw czogm ohm mommpcmohnm ocflshopoo 09 com: monmsom coo: \H

.mlm cam :Im

.Ho>oH ucoopom 0:0 0:9 um HQMOflmwcwwm «a

 

 

.Hm>oa “cuckoo o>wm may yo pcmowmwcmwm «
.pcmowmwcwwmIcoz mz
.mmcwpcmam
cwngz mpoaa canvas mcofiumwhm> Eoocmp 0» man oocmwnm> mo “cocomsoo on“ ma filo
.mmcwpcmam can a
moonsom comm mo cowuomnopcw may 80pm mcwpasmon oocmwpm> mo “cocoaeoo mnu ma mAIo
.moopnom ooom coozyon mmocopommwo ow moo mocmwnm> mo ucmcooeoo may mw wlo
.movwm wcwpcmam comzpon wooconmmmwo 0» map mocmwpm> mo pcocooeoo onu ow mlo
nil
0x04 unflmmflm
.l_lphom _MJWm «aoJm «H.~ ««~.m: «am.:~ m2m.o «au.mm counmsom
.I.I_I I.I.I.I.I.I.I.I.1.I:I.I_I.I_I_I.I_I_I_I_I_I_I.I_I.IsMummmom
.l_lfihpm _H.mm «0H.» «so.m ««~.ow «aa.oa «am.oa «ao.mm Hmohom wwoaaox
IIIIIII+IIIIIIIIIIIIIIIIIIIIIoMommfiwam
m.o: h.H: ««N.o «o.m «as.wm «ao.om «0o.oa c«m.oa vmmmom wmoaawx
.I_I_I_I I.I.I.I.I.III_I.I.I.....I...I.I.1.I...I.I.I.I.I..........I.I.I_I.mowumd«
a.mm e.n~ .. .m .a.a a.m.om .«o.o~ .aa.o~ .«m... unonou mmoaaox
_I I I.I I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I.I:IsMaMwMom
ammoaa<
oxmg wcwmwwm
m.mm a.mm eao.m «.m.m «aa.m~ .«e.m~ a.m.am aaa.am pnonom mmoaaox
mommHmm> Hopov mo vcoonom
moma Noma mood mooa . mooH Nooa mood woman wfiwmwwcm confinsoo
o mAIo filo Alo « cw wGOfiumvcmHm
w m m

 

HMPOH mo pcoopom o no mocmwhm> nvmcoa oaoooc mo upcocomeoo mood cam wood

pcmcomsoo oocmwhm>

..\m.m00950m comm mm EOQMIroUGMMQM>
.za canoe

#8

1962 and 1963 (Table 13--note significant seed source x
plantation x years interaction). Although never accounting
for more than eight percent of the total variance, in some
instances the interaction component was more than one-quarter
the seed source component in 1963.

As with height growth, the interactions of individual
seed sources within groups showed no relation between loca-
tion of seed source and location of planting.

The lower Michigan plantations were all very similar
in 1962 seed source needle length (Table 15). The Houghton
plantation, however, showed the lowest degree of similarity
with the other Michigan plantations.

Color

 

Field Measurements.--The color differences between seed

 

sources closely followed the results reported by Wright and
Bull (1963) as well as the 1961 plantation scoringsl/ (Tables
16 and 17).

Northern seed sources were the yellowest and southern
seed sources the most blueggreen. This trend was not com-
pletely related to latitude however. Group E, from the Ural
Mountains of Russia--an area of intermediate latitude--was
the yellowest in all plantations. Sources from Spain (Group
N) and southern France (Group M) were the darkest green in all
plantations.

The plantings at Higgins Lake and Newaygo--the two

 

ll The 1961 color scorings were made by J. W. Wright.

#9

Table 15. Between plantation correlation coefficients using
the mean 1962 seed source needle length.

 

+1

U)

m o

p x

o m

m A o

x c

m. m c m 0

b0 I: It! .4 4.1

o -a m: .c

H b0 m o m

H bi) H (D :3

m oa r4 0 o

x :2 < a: m
Kellogg Forest 1.00
Higgins Lake .88 1.00
Allegan .92 .83 1.00
Rose Lake .92 .83 .88 1.00
Houghton .70 .70 .75 .76 1.00

 

All values are significant at the 0.1 percent level where
r = .36 with no degrees of freedom.

50

1:01. 10. .Color groloo nod Into o1 ooorlo‘ loocol ploo ooo0 ooorooo.

 

Kogloo. loco Loto 01331-0 loooygo Iollog; Allogoo I...
Country , loko Ibroot lbroot

of 011310, 00 1 0: r1
‘3’“ ”' on. 10 on. 1% 0".1'. 2": “on. u Iov. o In. 10

 

-——-----Porooo0 of plantation oooo--—-——--—-—-

1 3133/2114 11 as u « u a
c won 201 10 00 12 00 01 11
01! :22 14 00 01 00 00 01
'00 210 80 00 10 10 00 00
000 210 00 00 00 00 00 00
0'0 021 01 00 10 00 00 00
III 028 03 02 02 01 01 ' 01
0'0 000 02 00 00 09 11 00
8'! 020 00 00 00 00 11 II
010 503 80 00 00 00 01 01
8': 000 10 00 00 00 01 01
80: 000 10 00 00 00 00 01
".r.‘. '09. ‘.05 6'9. .2.‘ 02.3 “a.
1) LA! 220 10 03 10 01 00 01
LA? 220 80 01 00 10 01 03
5!: 541 at 1: 03 is a? 11
0'0 000 01 00 01 00 10 00
Aver-go 13.0 00.0 08.0 01.3 01.8 01.3
l 013 000 31 30 21 30 00 31
UIA :08 00 30 33 31 60 30
“mg. 41.5 30.0 27.0 35.5 50.0 30.0
P 101 011 loo 2 99 91 09 01
POL 311 102 81 101 90 00 00
Averago 101.0 88.0 103.0 03.0 01.0 03.5
6 sun 20? 91 90 100 100 110 100
all L03 111 103 100 110 110 110
an: $07 115 00 130 113 100 100
658 208 111 111 113 110 100 111
6:0 210 109 00 100 103 101 00
C1: 105 115 no 113 '0 00 110
CI! 300 I20 100 100 101 106 110
C1! 309 100 III 110 106 109 111
C2! 009 100 103 120 110 99 00
Clfi 3|0 [06 103 111 100 192 108
C28 all 111 90 120 09 101 111
C1! 312 100 81 110 00 100 110
“BR 505 100 100 100 113 100 115
all 521 93 93 101 00 100 00
Avorogo l01.0 99.0 139.8 106.1 100.0 109.1
I max 200 109 120 12? 113 110 1’0
K Y 335 113 110 I00 135 115 131
FRA “11 111 111 116 113 113 120
GER 231 111 [30 141 12F 101 121
G!“ 052 l13 ll? I10 120 101 125
NE“ 353 Ill I25 13? 123 109 113
"EL 319 111 125 131 113 109 110
REL 330 190 111 138 ll? 107 122
"L‘ ”53 100 109 I”? 113 l00 103
‘V'TOJQ 110.1 121.3 130.0 120.0 109.0 100.9
J Pu; 215 12? 131 100 12? 113 139
YUG :02 120 1?? I20 110 105 122
Av-rono 131.0 100.0 130.0 110.5 109.0 130.0
I TEN SIT I?" 101 131 130 111 130
r!“ :“0 II? ln1 130 123 II? 121
TUB 321 I?! 151 130 133 113 101
Gut 203 115 130 I40 lfil 112 120
ORE :00 110 Ill 13! 120 109 123
Av-rogo 119.0 101.0 130.0 120.0 112.0 129.0
K FHA "00 lflb 100 l03 106 l33 101
FHA 239 122 159 175 109 126 151
Av-rogo l”0.0 101.0 109.0 107.0 109.0 100.0
I SPA 21! 131 110 I10 109 131 101
311 210 13: 110 151 141 123 143
81A 045 132 110 110 109 I30 100
SPA :40 179 160 101 120 1!" 131
SPA 20? 130 113 101 100 100 130
Avorfigo 131.0 113.0 105.0 100.0 100.0 100.0

 

Ioon 71.010110.
co1or grodoi/ 0.10 0.10 0.00 0.00 0.11 0.10

aloud-rd orror
of oood ooureo
noon to o porcooc-
0! .100101100 loo:
' 0.00 0.10 1.00 0.01 0.01 0.10

 

_l/ 1111.1“, Cinclvonlouun, hung-o, Grit-any, Giza-bonuwgmry, 1.11-v“,
NOquy, 101-ad, Slflvzln, SlAin, Sabina, YUGooIQVIo, LEA] mouutulno.

o
-l
-l Color :rodoll I o Jolie's-1| IO o darkoot rroon.

51

Thhle 17“ Between-plantation correlation coefficiente ueing the seen
eeed eource color grade for the 1961 end 1962 leeenre-ente.

I
6.9
:I
‘1. ..
s. .
Ln. 0
in ‘6
can 1!-
e c
z- e
._1 '
c a g
2‘.
1901 Neeeurenente
Kellogg Fire-t 1.00
Rune Poreet .91 1.00
Newaygo .92 .91
Btggine Lekc .90 .80
Michigen State .91 .88
nureery
I962 Icnenrelente
Kellogg Poreet .90 .92
lane Foreet .96 .90
Neveygo .96 .90
fliggtnl Lnke .91 .89
Allegen .96 .88
lane Lake .96 .91

Newaygo

1.00
.88
.93

.90
.96
.96
.90
.93
.93

Biggine Lake

1.00
.82

.81
.81
.89
.82
.89
.89

Michigan State
nnrecry

1.00

.96
.96
.93
.90
.93
.92

1962 Meneuronente
hellogg Fore-t

1.00
.91
.90
.96
.90
.90

0
o.‘ .I
. .5 .
1- .8
:3 8. 2 5 S
h - U
I 3 EL 0 0
I h u no I
5 0 '- III 0
a: 2 B 1 2
1.00
.91 1.00
.96 .93 1.00
.90 .90 .91 1.00
.90 .96 .91

.96 1.00

 

A11 velaee ere eignifieent et the 9.1

.66 de‘rcee of freedcl.

percent level ehere r e .31 with

52

northernmost plantations scored for color--showed the
greatest color differentiation. These plantation means
were lower and they show the greatest range in plot means.
Allegan-~the next northernmost plantation--showed the least
differentiation. Since Allegan was one of the last planta—
tions measured this suggests that the plantation itself is
a more important source of color variation than the date

of scoring.

There was a small but significant seed source x plant-
ation interaction among the plantings. But this interaction
never accounted for more than six percent of the total
variation and was very small in relation to the variation
due to seed source differences.

Sources from southern Scandinavia (Group C) were much
more yellow at Higgins Lake than at any other planting.
Source MSFG 521 from southern Sweden was about 1 1/2 grades
yellower at Higgins Lake than at any other planting. The
Spanish sources appeared bluer at Higgins Lake than anywhere
else, but this was probably a result of the scoring method.
Laboratory Measurements.--As mentioned previously (See
Methods-~measurement) a special color scoring was made in
the laboratory under uniform lighting conditions. This was
-done to eliminate observer bias when evaluating color at
different times and places.

The results of this laboratory test (Table 19) strongly
confirm the conclusions from the plantation measurements.

There are real differences between plantations. Southern

53

Table 18. 1962 components of color variance as a percent

of total color variance--from 57 seed sources.1/

 

Variance Component

 

 

Plantations in _2 g_} _2 _2
Combined Analysis . A.. 0 p --. 0‘s' - 0 pg 0 E
Russ Forest, Allegan, 3.18** 69.15** 5.73** 21.9%

Kellogg Forest, Newaygo,

Russ Forest, Allegan, 3.3u** 69.“3** 5.8u** 21.39
Kellogg Forest, Newaygo,
Higgins Lake 0

Russ Forest, Allegan, 2.28** 69.H2** 5.68** 22.67
Kellogg Forest, Newaygo,
Rose Lake

Newaygo, Rose Lake, 2.22** 68.u8** 5.59** 23.71
Higgins ‘ . ' ' . .

Kellogg Forest, Newaygo, 0.25NS 77.05** 0.80* 21.90
Russ Forest a . .

Eel-13g; FOFeEtT " " ' " — " 3.31%" "afib’u'if 3.32%" 3975?

Russ Forest

 

2
0_b is the component of variance due to differences between
planting sites.
2
‘3 is the component of variance due to differences between
seed sources.

0 p5 is the component of variance resulting from the inter-
action of seed sources and planting sites.

0 B is the component of variance due to random variations
within plots within plantings.

NS Non-significant.
* Significant at the five percent level.
** Significant at the one percent level.

1/ Mean squares used to determine percentages shown in
~Appendix B, Table B—6.

51+

fehle u. Ieeelte end mlyeie e1 1eheretery celer lees-recent.

 

 

 

 

 

 

 

1
:3; 2m
61 eflgin, Neveyge Higgile Like 11099 1‘39
In In.
lee- ceIer gredey
c null, 201 6.6 0.2 1.0
”I 212 1.0 6.2 11.6
In 621 6.8 6.8 6.6
0 1.69 228 9.6 3.0 8.0
I 919 266 6.0 6.2 0.6
096 266 6.6 6.6 8.8
9 an 202 10.9 10.9 12.0
on 206 19.6 18.2 a, 10.8
ca 619 11.0 12.2 12.2
I m 201 16.0 18.6 16.6
m 261 16.2 16.2 16.0
I. 660 10.0 12.0 16.2
I!!! 669 12.9 19.2 13.6
J 100 202 12.6 10.9 10.6
I m 212 16.0 10.0 16.0
m 221 11.2 11.6 11.6
I m 269 11.0 11.6 16.0
I 296 219 16 .2 16.0 11.6
"A 219 11.2 11.6 16.6
"ent-Hen IOOI 11.66 11.26 12.66
1 e ve
Ieeree e! verieune Degreee. of Ilene equere Coupon-.111 of verlenee ee
freedoe e percent of totel verlence
2166666106 2 00 .0893] 2 .1
leed eeeree 16 266.920 10.9
Seed eeeree l
l1181““°' so 10.05-cg/ °°°
216 6.00 23.0
N - Sig-"Ice“ It the ele percent 1eve1.

1 - yeIIeIeeH 19 - derheet green.
"‘1‘, alchedenhie, mo... Gabe-y, "Huger-y, “Me, NOR-e},

‘ ey, m1 Hound-e, ”Berle, ”Ail. men. 100ee1ev1e.
g/ e1..1!1..... eetehllehed by t-teet.

Ieeed en eeed eenree I plenutien Internet“- en errer ten.
heed en eeed eeuree l replleete within-pinnacle. ee errer ten.

55

Scandinavian sources are more yellow at Higgins Lake than

at any other plantation. Source MSFG 521 from southern

Sweden does show a distinctly different color at Higgins Lake.
Furthermore, the components of variance expressed as a

percent of total variance differ by less than two percent

between the laboratory scoring and the combined plantation

analyses of the same three plantation scorings.

APPLICABILITY OF RESULTS TO OTHER AREAS

 

Results of other Scotch pine provenance tests indicate
that the results of this study, although based on very young
trees, are applicable over a large part of north-central and
northeastern United States as well as central Europe and
southern Sweden.

Seventeen-year results in New Hampshire (Wright and
Baldwin, 1957) and eighteen—year results in New York
(Schreiner S£.El°’ 1962) show that heightgrowth superiority
of the German and Belgium sources continues in this area as
the trees mature. Wiedemann's (1930) summary of the 1907
International Union of Forest Research Organizations test
based on unreplicated Scotch pine plantings in Germany,
Belgium, Sweden and the Netherlands also indicate the height
growth superiority of the central EurOpean sources through-
out central Europe.

The New Hampshire and New York plantations, which had
thirty-one sources in common, when subjected to a combined
analysis (see page 11), showed a seed source x plantation
interaction component that accounted for about four percent
of the total variation or about one-sixth the variation due
to seed source differences. Thus even the size of the
height growth interaction estimates may apply to mature trees
in the northeastern United States.

Needle length measurements made on four-year-old trees

of the New Hampshire study are also in general agreement

56

57

with the results of this study. Sources from Belgium and
Germany had the longest needles while sources from southern
Sweden were slightly shorter. Although there were no
needle length measurements of northern Scandinavian sources
in this present study, both the three-year Michigan nursery
study (Wright and Bull, 1963) and the New Hampshire study
indicated that sources from northern Sweden had shorter
needles than sources from southern Sweden.

Color results are probably applicable over the widest
range of sites and regions. The present study, the Michigan
nursery study (Wright and Bull, 1963) and the New Hampshire
study (Wright and Baldwin, 1957) all reported the same
pattern of color differences. Moreover, both the New Hamp-
shire study and a Swedish study (Langlet, 1936-—see below)
reported the same differences between sources from northern
and southern Sweden.

Further evidence of the constancy of Scotch pine seed
source color may be inferred from advertisements of tree
seed dealers. Many dealers have long advertised "French
green" or "golden yellow" strains of Scotch pine in
nationally distributed publications. Such advertising would
not be profitable for very long unless the seed produced the
color advertised wherever the buyer grew them.

However, there is evidence that seed source x plantation
interactions become more severe at the extreme northern

portion of the species range.

58

Johnsson (1955) reported on the lS-year growth of
progeny from 79 trees located in 20 stands in Sweden and
grown at 3 widely separated Swedish locations. The latitude
of the seed sources ranged from 57°35' north to 65°39! north,
and the three plantings were located at Boxholm (58°10' N.
Lat.), Dalfors (61°18' N. Lat.), and Rorstrom (64°13' N. Lat.).
The southern Scandinavian sources were the fastest growing
in the southernmost planting, but did not retain their
superiority in either northern planting.

Langlet (1936) reported on 26 Swedish Scotch pine seed
sources ranging in latitude from 56°07' N. to 70°0' N. The
sources were grown at three Swedish locations: Tannersjbheden
(56°00' N. Lat.); Kulbacksliden (60°10' N. Lat.); and Gailli-
vare (67°8' N. Lat.). At the southernmost plantation the
northern sources all had much shorter needles than the
southern sources. At Kulbacksliden, the intermediate site,
there were only slight differences in needle length between
all seed sources, while at the northernmost planting the
northern sources had the longest needles.

Thus while these results may apply over a wide area of
intermediate latitudes, the height growth and needle length

results probably do not apply north of about 60° N. latitude.

DISCUSSION

 

There was no indication that poor planting practice
affected the performance of individual seedlots at indivi-
dual plantings. Using first-year mortality as a measure of
planting technique, seed sources with an abnormally low
growth rate (for their geographic group) were compared with
mortality figures (not shown) for the same planting. There
was no apparent relationship.

Transplanting effects probably contributed to the
years x seed source height growth interaction. Wright and
Bull (1963) noted at the time of lifting and transplanting
that the Spanish sources were noticeably more tap-rooted.
The improved height growth of the Spanish sources in 1963
is probably due to the fact that their root systems were
relatively more damaged in transplanting and, therefore, the
effects of transplanting dissipate more slowly in these
sources.

In 1962 a combined analysis of the Higgins Lake and
Houghton data showed that six percent of the height growth
variance and two percent of the needle length variance was
accounted for by the seed source x plantation interaction.
However, in 1963 the same analyses showed a non-significant
interaction in height growth while the needle length inter-
action jumped to 6.6 percent of the total variance. Other

combined analyses show similar fluctuations (Tables 9 and

59

60

10) in the seed source x plantation interaction. This
suggests that soil and photoperiod (which obviously remain
constant from year to year) are not causal factors in the
height growth and needle length interactions.

This does not imply that photoperiod could not cause
seed source x plantation interactions. It only indicates that
the range of photoperiods covered by these sites produced
no interaction. This is not surprising since the difference
in maximum daylength between the northernmost and southern-
most Michigan plantations is only about thirty-five minutes.
Furthermore, as indicated by the controlled environment
tests of Perry (1962) and Irgens-Moller (1962), temperature
may mask the effects of photoperiod.

The lack of correlation between seed source location,
plantation location, and growth indicates that, within the
area covered by this study, one cannot use the simple rela-
tionships discovered by early authors to predict the
performance of a seed source at a particular site from its
latitude or average temperature. The relationships are
more complex.

It should be kept in mind that the seed source x year
interaction in this study differs somewhat in definition
from the seed source (or family) x year interaction of the
agronomy literature. In this study the measurements were
repeated on the same trees. Thus a combination of yearly

climatic fluctuations, physiological maturation of the plant,

61

and among-origin differences in transplanting recovery rate
all contribute to this interaction. It is expected then
that this component will diminish as the trees mature and
recover from transplanting.

The year x plantation height growth interaction agrees
with what is commonly observed on standard forest tree site
index curves. These curves show that for many tree species
both growth rate and change in growth rate with time vary
from site to site.

An exceptionally large portion of the total needle length
variance was due to the years x plantation interaction. There
was a late spring frost in 1963. All four Michigan plantations
had below freezing temperatures on May 23. This suggests
the following explanation for the large year x plantation
needle length interaction.

Wright and Bull (1963) reported that all sources began
growth together in early May at East Lansing. If we assume
the southernmost plantation would begin growth first and
the northernmost plantation begin last, then there would be
large between-plantation, but little within-plantation
(among-seed source) variation in foliage condition at the
time of the late frost. The needles at the southernmost
plantation would be elongating and exposed while the needle
fascicles at the northern plantation would still be protected
by the bud scales. As a result there would be little effect

on among—seed-source needle reduction, large between-planting

62

differences in needle reduction, and the amount of needle
length reduction would decrease as we move north.

This is exactly the pattern that was found. The seed
source interactions show little variation between plantings
while Kellogg Forest (the southernmost plantation) shows the
,greatest needle length reduction, Allegan the next greatest
reduction, Higgins Lake next, and Houghton (the northernmost
plantation) shows no reduction in needle length.

The amount of interaction was about the same (five percent
of the total) for all three traits. It is expected that for
the same variety of seed sources the interaction would be of
the same relative magnitude for other complex traits--diameter
growth, form, wood density, etc.

Replication in time and space seems to offer little
increase in precision of estimates of genetic variance. A
researcher studying genetic variation in Scotch pine could
make very reliable conclusions regarding ecotypes or clines

on the basis of only a single set of test conditions.

PRACTICAL APPLICATION OE RESULTS

 

There is no direct way of equating the size of the
interaction variance and the advantages of testing at sev-
eral locations rather than a single one. However, the
following examples show the practical consequences of this
interaction.

Considering 1963 height growth at Higgins Lake and
Kellogg Forest, the seed source x plantation interaction is
about 5 percent of the total variance, and the seed source
component is about 35 percent of the total. Both components
are statistically significant. If one wanted to breed a
faster growing Scotch pine for Higgins Lake but had data
only from Kellogg Forest, he would necessarily choose seed
sources 201, 251, 253, 310, and 318 because they are the
fastestgrowing at Kellogg. At Higgins Lake those sources
grew 126 percent as fast as the plantation mean, compared
with 128 percent for the 5 sources that were fastest growing
at Higgins Lake.

Considering 1963 height growth at Kellogg Forest and
Houghton, the sizes of the seed source and interaction com-
ponents are about the same as between Kellogg Forest and
Higgins Lake. However, the consequences of selection at
Kellogg for planting at Houghton would be much more serious.
At Houghton the 5 seed sources growing the fastest at Kellogg

grew only 118 percent as fast as the plantation mean, whereas

63

60

there are 5 sources at Houghton capable of growing 129
percent as fast as the plantation mean.

Similar comparisons using 1962 height growth and the
combined 1962 and 1963 height growth indicate that the 5
best selected at Kellogg Forest may be as much as 18 percent
lower than the 5 best growing at either Houghton or Higgins
Lake.

The ten fastest growing seed sources at Kellogg Forest,
Higgins Lake, and Houghton are all in groups G and H (Belgium-
Germany-Czechoslovakia). This indicates that in the present
study height growth measurements of individual seed sources
should be confined to those sources in groups G and H.

In tests of other species, the tree breeder should
establish only a single large outplanting, measure it for
two or three years, and then replicate in space only those
ecotypes, or portions of clines, that do best in the large
planting. Such a procedure would also allow the tree breeder
to test a larger number of seed sources with the same amount
of effort thus increasing the amount of genetic gain per
dollar spent.

That there was a significant amount of seed source x
plantation interaction would have little practical signifi-
cance in a breeding program designed to obtain a tree with
darker green foliage because the interaction involved only
the medium and yellower sources. If one selected the
,greenest seed sources in table 15, he would choose sources

218, 219, 238, 239, 205, and 207. In other words, one could

65

select for color almost any place in Michigan for planting

at almost any other place.

SUMMARY.

Scotch pine seed, collected from 122 native stands
throughout the species'range, was sown in the Michigan State
University forest tree nursery in the spring of 1959. Each
seed lot consisted of seed from about ten trees per stand.

In 1961 two-year-old stock was used to establish per-
manent test plantations throughout Michigan and the central
United States. The plantings follow a randomized block
design with seven to ten replications. The number of seed
sources per plantation varies from 50 to 100.

In 1962 following the second growing season after out-
planting, one plantation in central Illinois was measured
for height growth while in Michigan seven plantings were
measured for height growth, five for needle length, and
seven for color. In 1963 four Michigan plantations were
measured for height.growth and needle length.

Analyses of variance of each character was made for
each test plantation using plot totals as items. The indi-
vidual plantation analyses were grouped into various combin-
ations and the mean squares used to compute the resulting
variance components. These components were then expressed
as a percent of the total variance for comparative purposes.

Seed sources from Belgium, Germany, and Czechoslovakia
made the most height growth at all plantings in 1962 and

1963. Sources from Spain made the least. However in 1963

66

67

there were indications that the Spanish sources might outgrow
the Scandinavian sources as the effects of transplanting
dissipated.

The planting sites also showed marked differences in
growth rate. They differed in the amount of growth per year
and in the change in growth rate with time.

Sources from central Europe had the longest needles and
sources from either the northern or southern extremity of
the Scotch pine range had the shortest needles. However,
this trend differed more by plantation than did height
growth. While the Scandinavian sources were relatively
short—needled in the northernmost Michigan plantation, they
had relatively long needles in the southernmost plantation.
The Spanish sources on the other hand all had relatively
longer needles at the northernmost plantation.

The mean needle length for all sources combined differed
sharply with site and year. In 1963 the mean needle length
decreased by twenty-five percent at Kellogg while it increased
by ten percent at Houghton. It is suggested that this year
x plantation interaction was the result of a late frost in
May 1963.

The Spanish, Greek-Turkish, and south France sources
were the darkest green and sources from the Ural mountains
and Scandinavia were the most yellow. Sources from Scandi-
navia showed more yellowing at Higgins Lake than at any

other plantation but were still not as yellow as the Ural

68

mountain sources.

The seed source x plantation interaction of the indivi-
dual seed sources showed no relation to seed source location
or plantation location. Differences in performance of
sources between plantations seems more a result of tempera-
ture and moisture variations than between-planting differences
in soil or photoperiod. I

The component of variance resulting from seed source x
plantation interaction never accounted for more than six
percent of the total variation encountered in all plantings
in either 1962 or 1963. This interaction component was about
1/6 the seed source component for height growth; 1/5 the seed
source component for needle length; and about 1/12 the seed
source component for color.

The effect of yearly fluctuations of climate on seed
source differences (year x seed source interaction) was also
very small in relation to the seed source differences.

Under the conditions of pest infestation encountered in
this test (very low), the researcher could gain little more
than a two or three percentgain in precision by replicating
measurements in time and space. From a practical standpoint
thesedata indicate that the most geneticgain per dollar
spent would be obtained by testing an increased number of
seed sources in a single planting and then replicating in
time and space only a few of the best sources.

Data from other tests indicates that these results should
also be applicable in mature trees throughout the North-

central and Northeastern United States and central Europe.

LITERATURE CITED

 

Arend, J. L., Smith, N. F., Spurr, S. H., and Wright, J. W.
1961. Jack pine geographic variation--five year results
from Lower Michigan. Mich. Acad. Sci. Papers 1961.
06:219-38.

Baron, Frank J., and Schubert, Gilbert H. 1963. Seed origin
and size of ponderosa pine planting stock grown at
several California nurseries. Pacific Northwest Forest
and Range Expt. Sta. Res. Note PSW-9. 11 pp.

Callaham, R. Z. and Liddicoet, A. R. 1961. Altitudinal
variation at 20 years in ponderosa and Jeffrey pines.
J. For. 59: 810-820.

Callaham, R. Z. 1962. Geographic variability in growth,

p. 311-325. In T. T. Kozlowski, (ed.), Tree growth.
Ronald Press, New York.

Cochran, William G., and Cox, Gertrude M. 1957. Experi-
mental designs. 2nd ed. John Wiley 8 Sons, New York.
611 pp.

Comstock, R. E. and Moll, R. H. 1963. Genotype-environment
interactions, p. 160-196. In W. D. Hanson and H. F.
Robinson, (ed.), Statistical genetics and plant breed-
ing. Nat. Acad. Sci. Publ.982.

Gardner, C. O. 1963. Estimates of genetic parameters in

cross—fertilizing plants and their implications in

69

70

plant breeding, p. 225-252. In w. D. Hanson and H. F.
Robinson, (ed.), Statistical genetics and plant breed-
ing. Nat. Acad. Sci. Publ. 982.

Genys, John B. Variation in EurOpean larch. Fox. Res.
Forest Bull. 13. 96 pp.

Henry, B. W. 1959. Disease and insects in the southwide
pine seed source study plantations during the first
five years. South. Conf. on Forest Tree Impr. Proc.
5:12—17.

Holst, M. 1963. Growth of Norway spruce (Piggg_gbig§_(L.)
Karst.) provenances in eastern North America. Canada
Dept. of Forestry Publ. 1022. 15 pp.

Irgens-Moller, H. 1962. Genotypic variation in photo-
periodic response of douglas-fir seedlings. Forest
Sci. 8:360562.

Johnsson, H. 1955. Utvecklingen i 15- ariga f6r56ksodlingan
av tall i relation till proveniens och odlingsort.
Svenska SkogstBren. Tidskr. 53:58-88.

Kalela, A. 1937. Zur Synthese der experimentallen Unter-
suchungen fiber Klimarassen der Holzarten. Helsinki.
030 pp.

Karshcon, R. 1909. Untersuchen fiber die physiologische
Variabilitat von Fohrenkumlingen antochthoner Popula-
tionen. Schweiz. Anstalt forstl. Versuchsw. Mitt.

26: 205-200.

Kriebel, H. B., and Wang, Chi-Wu. 1962. The interaction

71

between provenance and degree of chilling in bud-break
of sugar maple. Silvae Genetica, 11:125-130.

Langlet, Olof. 1936. Studier 6ver tallens fysiologiska
variabilitet och des samband med klimatet. Meddelanden
frgn Statens SkogszrsBksanstalt 29:219-070.

Langlet, Olof. 1963. The Norway spruce provenance experi-
ments at Danelt and Hjuleberg. Stbckholm, 1963
(Mimeographed).

Matzinger, D. F. 1963. Experimental estimates of genetic
parameters and their applications in self-fertilizing
plants, p. 253-279. In W. D. Hanson and H. F. Robin-
son, (ed.), Statistical genetics and plant breeding.
Nat. Acad. Sci. Publ. 982.

Mirov, N. T., Duffield, J. W., Liddicoet, A. R. 1952.

Altitudinal races of Pinus ponderosa--a 12-year

 

progress report. J. Forestry 50:825-831.

Munger, T. T., and Morris, W. G. 1936. Growth of Douglas-
fir trees of known seed source. U. 8. Dept. Agr. Tech.
Bull. 537. no p.

Munger, T. T. 1907. Growth of ten regional races of ponderosa
pine in six plantations. Pacific Northwest Forest Expt.
Sta. Forest Res. Note 39. H p.

Perry, Thomas O. 1962. Racial variation in the day and
night temperature requirements of red maple and loblolly
pine. Forest Sci. 8:336-340.

Rubner, K. 1957. Ergebnisse einer heute 20- Jahrigen

Fichtenherkunfst-versuches. 1. Teil. Die Flache in

72

Bayern. Silvae Genetica, 6:65-70.

Rudolph, T. D. 1962. Lammas growth and prolepsis in jack
pine in the Lake States. Ph.D. Thesis. Univ. of
Minn. (L. C. Card No. Mic 61-5866) 325 p. Univ.
Microfilms. Ann Arbor, Mich. (Dissertation Abstr. 22:
2156-2157)

Rycroft, H. B. and Wicht, C. L. 1907. Field trials of
geographical races of Einug pinaster in South Africa.
Pretoria, Dept. of Forestry. Brit. Empire Forestry
Conf. 12 p.

Santamour, Frank S. Jr. 1960. Seasonal growth in white pine
seedlings from different provenances. Northeast.Forest
Expt. Sta. Forest Res. Note 105. u p.

Schbnbach, H. Ergebuisse eine heute 20- jahrigen Fichten-
herkunftsversuches. II. Teil Die Flachen in Thuringen
und Sachsen. Silvae Genetica 6:70-91.

Schreiner, Ernst J., Littlefield, E. W., and Eliason, E. J.
1962. Results of 1938 IUFRO Scotch pine provenance test
in New York. Northeast. Forest Expt. Sta. Pap. 166.

23 p.

Sluder, Earl R. 1963. A white pine provenance study in the
southern Appalachians. Southeastern Forest Expt. Sta.
Res. Pap. SE—2. 16 p.

Snyder, E. B. and Allen, R. M. 1963. Sampling, nursery, and
year-replication effects in a longleaf pine progeny test.

p. 26-27. Proc. Forest Genet. Workshop, Macon, Georgia.

73

Oct. 1962. Southern Forest Tree Impr. Con£.Publ. 22.

Squillace, A. E., and Silen, Roy R. 1962. Racial variation
in ponderosa pine. Forest Sci. Monog. 2. 27 p.

Vaartaja, O. 1950. PhotOperiodic ecotypes of trees. Canad.
J. Bot. 32:392-399.

Vaartaja, O. 1959. Evidence of photoperiodic ecotypes in
trees. Ecol. Monog. 29:91-111.

Vins, B. 1963. Report on the state and preliminary evalua-
tion of Czechoslovak provenance trial plots of Norway
spruce in the international series from the year 1938.
World Consultation of Forest Genet. and Tree Impr.,
Stockholm, 1963. (Mimeographed).

Wakeley, Philip C. 1961. Results of the southwide pine seed
source study through 1960-1961. South. Conf. on Forest
Tree Impr. Proc. 6:10-20.

Wassink, E. C. and Wiersma, J. H. 1955. Daylength responses
of some forest trees. Acta Bot. Neerl. 0:657-670.
Wiedemann, Erband. 1930. Die Versuche fiber den Einfluss der

Herkunft des Kiefernsamens aus der preussischen
forstlichen Versuchsanstalt. Z. Forst- u. Jagdw., 62:
age-522, 809-836.

Wright, Jonathan W. 1962. Genetics of forest tree improve-
ment. FAO Forestry and Forest Prod. Studies 16, Rome.
399 p.

Wright, Jonathan, W., and Baldwin, Henry J. 1957. The 1938

International Union Scotch pine provenance test in New

7”:

Hampshire. Silvae Genetica 6:2-14.
Wright, Jonathan W., and Bull, W. Ira. 1963. Geographic
variation in Scotch pine. Silvae Genetica 12:1-25.
Wright, Jonathan W., Lemmien, Walter L., and Bright, John.
1963. Geographic variation in eastern white pine--
6-year results. Quart. Bull., Mich. Agr. Expt. Sta.,

East Lansing. 05:691-697.

APPENDIX A

METHODOLOGY

 

Height measurement

 

In the nursery phase of this test Wright and Bull (1963)
found no differences between progeny in the amount of within-
plot variability. If this constant within-plot variation
pattern continues in the field, it should be possible to
develop many measurement shortcuts based on measuring only
a few extreme individuals within each plot, i.e. the
tallest, shortest, longest needled, etc.

In 1962 the fastest growing tree on each plot was re-
corded separately for comparison with results from the four-
tree measurement. The following tabulation shows a comparison
of the coefficients of variation between four-tree totals and

the tallest tree measurement.

 

Coefficient of Variation

 

 

Plantation : u-tree total ' : Tallest tree
Percent
Kellogg Forest 6.9 7.7
Russ Forest 8.1 9.0
Newaygo 7.0 8.2
Higgins Lake 8.9 9.5
Allegan 6.3 7.6
Rose Lake 8.1 9.9
Edwards 9.3l/ 9.7

.

 

l/ Based on two-tree total.

Measuring all the trees on the plot reduces the coeffi-
cient of variation by about 1.5 percent. With ten replications
per planting, an increase in the coefficient of variation of

two percent will increase the size of the detectable

76

77

difference at the one percent significance level about six
percent (Cochran and Cox, 1957, Table 2.1).

Moreover, a combined analysis of the tallest-tree-per-
plot data for the 02 sources common to Kellogg Forest, Russ
Forest, Newaygo, Allegan, Higgins Lake, and Rose Lake shows
that 16.8 percent of the total variance is accounted for by
plantation differences, 28.2 percent of the variance is due
to seed source differences, and 9.6 percent of the variance
due to seed source x plantation interaction. These compon-
ents are all within two percent of the four-tree plot data
shown in Table 10.

Thus there seems to be little advantage gained from
measuring more than the tallest tree on each plot.

Needle length measurement

 

The needle length data presented in this study is based
on the measurement of one needle fascicle per tree. To
determine the increase in precision from measuring more
fascicles, ten to fifteen fascicles per tree were removed
from nineteen seed sources in five replications of three
plantations in 1962. The following tabulation shows the co-
efficient of variation for: ten to fifteen fascicles measured
in five replications; one fascicle per tree measured in the
same five replications; and one fascicle per tree measured

in all replications.

78

 

: Coefficient of variation
:10-15 fascicles: 1 fascicle l fascicle

 

per tree : per tree per tree

 

Plantation 5 reps. " : 5 repls. all reps._
Percent

Kellogg Forest 5.8 5.9 0.0

Higgins Lake 5.0 0.7 0.2

Allegan 0.5 5.7 0.5

 

1/ Kellogg Forest, Higgins Lake, and Allegan had respec-
tively 10, 7, and 8 measured replications.

As the tabulation indicates, measuring one fascicle
per tree was as precise as measuring many fascicles at
Kellogg Forest and Higgins Lake, but resulted in a small
loss of precision at Allegan.

Increasing the number of replications was more important

than increasing the number of fascicles.

APPENDIX B

fnhle V-l.

80

1963 conhlned Inllyeee of variance of height croeth.

 

llnntniinne in

Source 01 variation

 

 

 

 

 

cnnhined ennlyeie llanietinn Seed eonrce Seed enurce X Error
Agplenlellnn
I ll 1 ‘ Ue reee 0 freedom
0 org 010-
? b4 16? 1610
nip.tnn Lake .
All keen equere
efflfl
" h! 20081.03" 2005.530. 221.10.. 110.06
nuy 0n
nellngy Parr-L ungreee of freedo-
Hiyyine l‘ke 2 64 109 1160
"ought an Mean 8:1“. re
0010.50". 2150.3500 239.68" 131.11
K " r t Degroee of freedo-
e up: are-
I 50 54 610
'Hirlnelehe
keen eguere
8029.21“. 1550.619. 290.06" 130.01
”1,53“. 1‘3» aerreee «1 freedom
‘ Posh-111011 1 :14 5‘ 102
Mean enunre
7 . 1"“ 11H!.b0" 201.nu-- Inb.uu
‘"119L intv-l negleen of freedun
i-mlfhiou 1 50 b4 “'14
henn equere
.ns~.ut"' 1606.0 -- 100.61. 101.08

 

nu - \nn-ui'nlficnnl

0 - §|fn1r|Clu1 at the five percent level.

eI—i“.

.nhlt- 11- .

nifICnni n1. Um «one l-elcent level.

Inn! cunhinvd ennlyeee of veriencr of height yin-1h.

 

11-Iulni Inn»- in

.X'Il'ilcn 01' Vnr1fl1lnn

 

 

 

 

 

 

In!1|lil-I1 nah-lie llentntlun Viv—:11 lnlll‘co‘ Semi enurce X finer
ylnninlunn
‘n-llm far A! .b- [1‘0‘10 01. 11'!‘|‘:1¢)M
' “'"* "‘" e. In“ 1020
L11! RI
.‘Jn I'llllrl'
e
‘u " li'JuU.~H'0 .550.UQIO 411'I3ce 109.77
11:) 1' I'-1 . ll't‘l I11 1'I‘O‘l1l“
, n4 10~ 1114
'1 ins In ,.
‘vn-i ev'ufll'e'
' .1. 'x' .2;
flfuon.nl-- 50 1,0600 |U-.<3-' ‘7l.50
11:110‘ 1‘ I 51 ..I- lun‘fl Hf 1;--.-d~ua
.-.b——_
o ‘1"! lnlv 1 50 54 150
Mann euuqro
l'i‘ ‘0'_‘u'. 19".I”‘.. ‘¢“‘.‘,.)". .‘i1'1.13
I llv- 1A.-- c- [run “1 1l-'--«1(v-v
m. on. I 7.. {)4 70.‘
mrnu ntunve
l 71. ‘n- 04 -- n _‘l'mI 'Lu.\n
'1‘ 1- [v-1 . l .. .yf {random
1 1 >0 1 51 5‘. 91°
1 .-n n .r-unra-
\ 1,Ju'0 .l 3,61" fih‘,0029 1|].ln
>~ - -~;--| will. .4
' - 1 L|I1|Cun1h1 ".- ii\'-‘ ,vnunl luv-'1.
- \‘n I111!Mh1, at l'»- i-ue ~,vn‘c-nl lave-1.

TIL] a 11.1.

10a? contined nnelyeee of verience of bejiht yroelh.

 

llnninlinne in

Snurce of verielinn

 

 

 

 

 

 

 

 

 

 

 

conhlned ently-19 llenieiinn fieed enurce Seed eource I Error
ijflnntetion
Kallop- Farr-i Uefreee of freedo-
Huee Vale-l 6 01 306 3009
Allunnn Moan enuere
2...,39 [7015,4200 0900.98" 213.11-0 1:2.05
-i(.lne Lake
Blue 1:10
Fouihton
k-Ilegu lnru-l Degree. of freeGOIJ
Inee Ferrel 0 01 100 1163
\eenyyn New" enunre
Allnlen S0011.R0" 4111.189. 200.61" 118.21
”anyhlon
Inc-110;; an'i'll W
“on Pale-1 u 01 123 1101
Alluyun Menu eguere
HnnyLlon 20Hrn.930' 130?.R000 "30.150I 121.1:
tellorp Pole-t Degrees of freedo-
Luee Porn-l 2 01 82 1394
Alleren Mean 'Qfllfli
26961.5420 211£.RR" 210.00.. 125.15
helloy. {ore-l varren of {rundo-
Hi;;ine Lake 2 01 92 002
zuuphtnn keen enunre
4:23.13". ""50.ro-- 610.1000 129.90
'«Ilu v Forest BflzLSLLJflLlI££i22
Allnynn 1 11 01 13h
Lean enunre
5¢.ud."1'° In n.9o-n 66.00"' 93.14
lpllngg Fore-9 .mgrve- nf freedom
.nev Lake 1 11 . 01 656
keen enunre
11600.39I' 1072.11" 310.02.. 111.91
hellogg Intent perroee of {tendon
Iuee Parent 1 41 01 138
Mean equnre
anin.o9"‘ 24.1.1... 220.33" 100.65
Lelloug Fare-i ueIrFee of freedo-
Houghlon 1 01 4| 050
Mean eueere '
2211.91“' 1610.69“. 168.90”. 133.34
Kollogr Pore-t Uegreee of frezggg
\oeeyuo 1 01 01 735
Menu sauere
38066.00" 1686.60.” 183.10.. 109.2
Kellogg Fore-6 22‘£:2£_2£_£L33322
Higgine Lake 1 01 91 615
leen eauere
6609.16'. 1613.60'0 213.36" 128.61

 

ne - Ion-eiuniticent

0 - Significen! n! the five percent level.

00 - Significant .1 the one percent level.

 

81

feble 9-0. 1992 ee-Mned enelyeee e! eerlenee e! needle leegu.

 

Pleeutiene in

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

knee :2 nrieun
eenhlned eeelyele llentetiee eeee eenree leed eeeree x in..-
plenlellen
lelleu Pereet Dareee :1 1“ng _
Ilgglne Lnte 3 60 199 15. r
Allegen Ieen 53"
log»... 3666.91" 111.10“ 11.61- 0.0:
Kellogg lereet Qg‘reee 3! free!!!
Allegen 1 60 60 6910
Ieen gar:
seamen-o 130.06“ 10.19“ 0.99
lelleg; Parent E‘I’OOI et {pede- I
Ilgglne the l 94 64 910 '
Ieen ggnnrg
1066.68" 181.99.. 10.06. 1.99
lelle“ 1'eree0 a‘reee ef {reeds
leeghten l 00 09 094
Ieen ”ICE.
6629.90" 199.99“ 29.01“ 9.61
reele H. 1963 eenbined enelyeee e! enrlenee e2 needle leg“.
Higgine Lnie De‘reee e! {reeggg
11...»... 1 u u m "mm": 1. M
u”- eer eenelned enelyele Ilenletlene Ieed eeeree lend eeeeee I Iver
2412.14» 14.54“ o.ae- 1.34 ""1““.
1.11.“ For-at Wing:
ne - \oe-eignlfiennl» Ilulne late I 90 I99 1m
2 - Significene e0 the five percent level. “l...- m .
00 - Si‘nlllcnnl It the one percent level. Ill‘hill 1109.8... 190.9... 9.0... 0.09
“2”.“ '2'.“ W
Illegen 1 .6 09 919
mm.
1199.990. 99.... 9.99" 9.99
1.11... form W
Iiulne late 1 66 99 199
w
909.11.. 99‘900' 11.1... 9.99
Yehle H. Co-blned enelyeee e! verienee fer eeler eeerle‘.
“Hon hm. W
llanteiione in Ieerce e! verlellen leeglten 1 .4 Id 919
ceehined eeelyeie Ileneeiiep leed eeeree leed eeeree I Irrer
plenutlel "Lee” nude” 19..” 0.99
De r f ! ede-
Ielleu fer-e1. b ‘5:- 3 re '. n“ “a“. m. ree fr
knee Pereel h...‘.. _ 1 .9 “ g.
um.- ” M
O . .0 8.30.. 0...
flgggn. 111.70. 113 .1 1'...“ 99.” ..‘.. ..“
Ilgglne Lake
u” 1". ne - lee elgnlfleent
u n 1 y. g, u f tr"... e o ligeltieene e0 the flee pereenl level.
0 o o I BIL—L1.“—
R". :"n ‘ M 220 2362 0: - lie-Hines e1. the ene rerun level.
Allegen 3222.2322L1
..u”. 39.1¢ee 161.96“ 3.660e 1.01
Higrine Lele
hHm 10ml We
Inee fereet 0 69 200 2000
.11.... 52-44122
«um 1111.31" 102.11» 0.22" 0-16
linee late
leeeygo Ezlreee g! {53:20-“
0
Ilgglne lake 3 0. 1" '93.
Ieee late u"
5._ogee 92.89.. 3.0.‘. 1",
1.11.“ r"... W
Inee Pbreet ' .. '1. "I.
I...“. man
".900. 111.0... 1.50. 1.10
Degreee I! freedo-
Iellm Yereet 1 66 u 10“
en
'1'” ”n“ 0.69" 111.11“ 1.02" 0.12

 

nn - lee elpnlflcent
0 - lignlflcenl e0 the flee percent level,
00 - Ii‘nificenl e0 "be ene percent level.

82

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

' \1
1 5555 5-5 . 55 5555-. 51 5555555.: ene-enu- 5555
.1
‘ *1
‘ 1 55.55.51- _ I M“ 2.
p .ee. .4 . 5.1 59 nude-55 51 lame-51 pleeuuee 1955
. 551 5 M. ‘ ' . 1 ”he“. 9559 eeeree her
M m an... m m
”the“. 9555 eeelee tree . u ‘
a...
5555,: 5555.55» 555.55
1.11... m... m
‘ 9 99 0“
, I I I III! 5115.55 .1.”
555.51 555.55» 155.55 | 55 555
u
“be“ m m-11 551.51» 51.01
5 55 555
“ neu- u. w
155.15 ‘ 151.55» 55.55 . 55 .5
, m ,
513.1.- m. m . 555.55 1555.15» 115.15
5 55 555
‘1‘ Inch- “.5111.
115.55 550.51» 151.15 1 55 515
hens
Iain... M 1.... “an"...{i 1...!"
1 55 515
m '0 Mullen“ 55 55. ene pereene level.
555.55 155.55» 155.15
Kellogg Peren6 m ' fie H e m.“ m .' 0‘1m' '1‘.“_ 1“
5 51 555 . ' "" '
m 1155555555 w
555.15 1555.55» 155.55
m 5555 eeeeee .5551
mm-
"h'" 1 a u- ' Wm M» mne-
u 0 55 .5
215.15 555.55» 55.15 m
. L 55.55 155.15» 5.51
111.1155 uh m H‘ ‘u I I I I
5 51 555 "n .
m ' “ 515
225.55 555.55» 151.55 , m
15.55 55.55» 5.55
Ilen:'lnl.en W
1 51 5111 55m.- ueo. ‘ m -
w ‘ 9 90 990
525.51 125.15» 155.55 m
1 12.99 55.51» 5..
1.... r..... W
9 01 969 leeclten m
“ 1 50 '515
1511.55 1555.11» 151.55 “
55.55 55.55» 1.51
a... me am '
1 1 51 551 5.11.“ fan» 9“.“
5 51 555
hen-n
451 .55 1:5 .55» 55 .55 “
~ 15.55 155.51» 5.55
11...... mm:- *
111.... m
5 51 555 1 51 501
m ..
551.5: 115.55» 55.15 m '
5.55 55.51» 5.15
W ‘ -
Meerde
l2 _ 01 092 ulggiee 1.555 W
I...“ . 01 650
155.51 115 .55» :5 .55 m
'0 Iigelfleeet 55 5115 ene percent level. 12.39 ”J.“ .J.
’ new
5.5.115.-
1 51 551‘
m .
55.55 ~ 51.55» 1.55
M- u- Mail-
1 51 551 -
m '
55.55 55.55» 5.55

 

 

oo
0)

9 ‘5 91-16 blue 56“ pt 15. 5555555551 5155555555
. . needle In“. .
55.55555. W
m 05‘ .5555 line

 

 

 

 

""1“ 'm“ m
9 99 999
a
99.99 91.99.. 95. 7
mon- m I
9 99 “9
m
15.55 55.55» 5.51 °
, 5...... 1.5. m
9 99 .99
m...
89.99 99.99.. 9.19
um... m
1 99 919
h...
.599 99..” 9.19

 

'0 21.11155“ 50 565 ene pereenl. level.

Inble 9-11. 1062 ennlyeln ef'vnrlenee reenl6e e2 plenentlee eel-r
eeerlngn.

 

1155555155 lonree e! vnrlellen

 

heplleetlen lend nonree Irror

 

[51155; rereel . 0 re {re en
0 66 600

9555 naerg
2.63 36.10" 1.11

lllegnn 23‘r555 :1 ireedgg
0 600

923 ”991'!
05.. 9.59... 0.“

 

 

ligglne Lnke De‘rege e! {regggg
6 66 339
M
1 .09 30.1000 1 .21

 

llnnn Pol-555 Degreee 32 freed:
’ 0 66 600

lee! ”are
3.26 00.03" 0.12

 

levnno ' Degreee e! Irgedg
9 69 909

g leen genre
0.60 61 .66” l .66

 

I»- Iel- We:
1 _ 55 555

lleen genre
6.69 11.26” 0.16

\

 

“613511155“ 55 665 ene pereent level.

3 _'

 

"'TITifitflmnIILIMIlfiflflfifljflfliflflfiflflflfi“