Testing of the narrow crowned Norway spruce ideotype ( Picea abies [629555]

Testing of the narrow crowned Norway spruce ideotype ( Picea abies
f. pendula ) and the hybrids with normal crown form ( pyramidalis )i n
multisite comparative trials
Marius Budeanu, Ecaterina Nicoleta Apostol ⁎, Flaviu Popescu, Drago șPostolache, Lucia Ioni ță
National Institute for Research and Development in Forestry “Marin Dr ăcea”, Voluntari, 128 Eroilor Boulevard, 077190, Ilfov County, Romania
HIGHLIGHTS
•The narrow crown Norway spruce
ideotype presents superior stabilitycompared to normal crown form.
•42% of the consanguineous trees sur-
vived and most of them are dominatedand will disappear in the next years
•Breeding strategy could begin with the
selection of pendula after trees slender-ness, trait with the highest heritabilityGRAPHICAL ABSTRACT
abstract article info
Article history:
Received 2 March 2019Received in revised form 29 May 2019Accepted 29 June 2019
Available online 2 July 2019
Editor: Alessandra De MarcoThe study aims to analyse the stability of the narrow crowned Norway spruce ( pendula form) compared to the
normal spruce form ( pyramidalis form) and the hybrids of the two forms, in 5 field trials (Comand ău, Lep șa
1&2, Ilva Mic ăand Voineasa) located in the Romanian Carpathians. Trees height (Th), breast height diameter
(Dbh), height growth of the last year, crown diameter (Cd), number of branches per whorl (Nbw) and dominant
branch diameter (Dbd) traits were measured and survival rate (Sr) was determined, at 20 years old. Also,
branches finesse (Bf), trees volume (Tv) and trees slenderness (Ts) were calculated. In order to compare the
wood density (Cwd) there were collected cores. In all trials ANOVA revealed signi ficant (p b0.05) differences be-
tween the two forms of spruce and the hybrids (mainly between those that have a different crown form mother),
especially for the stability and quality traits. Factorial ANOVA revealed a high in fluence (p b0.001) of the locality
and also a signi ficant in fluence (p b0.05) of the locality × spruce form interaction. The factor “form ”was signif-
icant for some traits involved in Norway spruce stability (Ts, Cd, Nbw). The pendula trees present higher values
for Sr, Dbh and Tv, and lower values for Ts, Cd, Nbw, Dbd and Bf, compared to pyramidalis spruce form, which
showed a higher stability. Heritability was in generally low (h2b0.4), with exceptions of Ts which presents a me-
dium rate of heritage. For the same trait, different heritability was registered in different environmental condi-
tions. The Cwd was higher only with 2% for the pendula form in Lep șa trial, while in Comand ău trial the
pyramidalis registered a higher value (7%). In the new breeding programme, the selection strategy may be pur-
sued with the pendula trees selection based on Ts and branches traits.
© 2019 Elsevier B.V. All rights reserved.Keywords:
AdaptabilityComparative trials
Diallel design
Phenotypic traitsTree breedingWood densityScience of the Total Environment 689 (2019) 980 –990
⁎Corresponding author.
E-mail addresses: [anonimizat] (E.N. Apostol), flaviu[anonimizat] (F. Popescu).
https://doi.org/10.1016/j.scitotenv.2019.06.518
0048-9697/© 2019 Elsevier B.V. All rights reserved.
Contents lists available at ScienceDirect
Science of the Total Environment
journal homepage: www.elsevier.com/locate/scitotenv

1. Introduction
Norway spruce [ Picea abies (L.) Karst.] is considered one of the
most important tree species in Europe, and, in the last years, negative
influence on trees growth and stability has been registered, caused
most probably by climatic perturbations ( Kauppi et al., 2014 ;
Rosner et al., 2014 ;Keenan, 2015 ;Steffenrem et al., 2016 ;Pretzsch
et al., 2018 ).
In Romania, Norway spruce covers an area of 1.488.000 ha ( INS,
2016 ), which represents 23.2% of the country forests, being the most
dominant species of the Romanian Carpathians ( Șofletea and Curtu
2007 ,Feurdean et al., 2011 ). In this country, 89 Forest Genetic Resources
(FGR) were established for in situ conservation of the species ( Pârnu ță
et al., 2011, 2012 ;Budeanu et al., 2019 ), but the preservation of valuable
gene funds is performed also ex situ , in 14 seed orchards and 34 compar-
ative trials. The concept of the tree ideotype was introduced in Finland
(Karki, 1985 ) and adopted in Romania by Enescu (1987) andPârnu ță
(2001, 2008) , with the meaning of a model tree which is distinguished
by its high bioaccumulation capacity, superior wood quality and high
adaptability. The genetic determinism of the crown form was investi-
gated in Germany ( Lehner et al., 1995 ) and the vegetative (especially
by cuttings) and generative multiplications research was made in
Finland ( Zubizarreta Gerendiain et al., 2008 ). Also, the pendula spruce
descendants and the hybrids between narrow crown spruce and the
normal crown ones ( pyramidalis ) were tested in Finland ( Karki, 1985 ;
Pulkkinen and Pöykkö, 1990 ;Mikola, 1993 ;Pöykkö, 1993 ;Geburek
et al., 2008 ). These studies suggested some particularities of narrow
crown spruce as low and irregular flowering, bending the character of
branches since the age of 6 years, and in case of generative multiplica-
tion, only half of the descendants are inheriting the character ( Lepistö,
1985 ). Numerous studies ( Kuuluvainen, 1988 ;Zubizarreta Gerendiain
et al., 2008, 2009 ;Rodriguez et al., 2013 ) recommended for pendula
form a dense planting scheme, 1 m × 1 m and 1 m × 1.5 m, while the
pyramidalis form has grown much better at larger planting scheme.
Lehner et al. (1995) have shown that the narrow crown type is con-
trolled by a single dominant gene, while other authors previously con-
sidered that there was a poligenic control ( Lepistö, 1985 ;Pulkkinen,
1991 ).Schmidt-Vogt (1977) considers that the crown form of the
Norway spruce is a variable character and a large part of the variation
can be explained by the environmental factors. In Romania, the researchfocused on 455 spruce trees with narrow crown ( pendula form and
columnaris variety) within 25 populations located in their natural
range in the Romanian Carpathians, most of them being located in the
Apuseni Mountains ( Pârnu ță,2 0 0 8 ). In 1994, two half-sib comparative
trials were installed, testing 24 families with narrow and 24 with nor-
mal crown forms, as well as a collection of narrow-crowned spruce
genitors composed of 125 families. In 1996, five full-sib comparative tri-
als were established for testing narrow and normal crown spruce forms
and also their hybrids. A full diallel mating design and a factorial mating
design allows all possible combinations between parents ( Schlegel,
2010 ), including self, and provides various genetic information
(Acquaah, 2012 ).
The aim of the present study was to evaluate the stability of the nar-
row crowned spruce ( Picea abies f. pendula ) compared to the classical
form (pyramidal and wide crown) and also the adaptability of the full
diallel combinations of the two crown forms in these five trials. The fol-
lowing issues were evaluated:
✓Comparative analysis of bioaccumulation capacity, wood quality
and adaptability of trees for the two spruce forms and their
hybrids,
✓Determination of wood density for the two spruce forms,
✓Behaviour of the two forms of spruce and their hybrids under dif-
ferent environmental conditions of the testing sites,
✓Determining the correlations between the analysed traits for de-
veloping a breeding strategy.2. Material and methods
Controlled pollination was performed in a representative popula-
tion of the pendula form (Stâna de Vale, Bihor County) in a full diallel
mating design between three trees with narrow crown and three
trees with normal spruce form ( pyramidalis ). The population was
chosen, in 1990, of a total of 48 populations identi fied and evaluated
within the species range in Romania. The selection criterion was the
phenotypic character of the crown form ( pendula )a n dt h en u m b e ro f
trees with this form within the population ( Pârnu ță, 2008 ). The
pendula trees are reported to the sixth crown type while pyramidalis
are reported to the first type, according to Schmidt-Vogt (1977) .T h e
crown diameter of pendula is lower than pyramidalis and the first
order branches are pendent, descends along the trunk. Usually, the
branches of the pendula trees are thinner. Stâna de Vale population
is located in the Western Romanian Carpathians, in forest district
Sfânta Maria, unit II Stâna de Vale, plot 21B, on a surface of 6.5 ha.
It is a natural forest with superior quality and productivity, installed
on typical eutricambosoil, on a slope with north-eastern exposure
and 21°declivity, at an altitude of 1150 m ( Anonymous, 2010 ). The
selected spruce trees had an average age of 145 years old, the breast
height diameter of 56 cm, the trees height of 33 m, the pruning
height was about 50%, and the crown diameter was 3.1 m for pendula
form and 5.2 m for the pyramidalis form. The distance among the six
mother trees was of up to 50 m, in order to minimize the possibility
of selecting related individuals ( Bruschi et al., 2000 ;Franji ćet al.,
2006 ;Apostol et al., 2017 ).
Five comparative trials have been established using seedlings with
both known parents and seedlings obtained by open pollination, in
1996. The aim was to compare the narrow-crown spruce trees with
the normal crown form and also to investigate the adaptability of thehybrids between the two forms of spruce. The comparative trials are
managed by forest districts of Comand ău, Lep șa (2), Ilva Mic ăand
Voineasa. The first three trials are located in the Curvature Carpathians
(one inner and two outside of the curvature), the forth trial is located
in the Eastern Carpathians and the last one is situated in the Southern
Carpathians ( Fig. 1 ). All comparative trials were established in the opti-
mum ecological distribution range of Norway spruce: two in the moun-
tain beech forests, two in the mountain mixed forests and one in a
mountain spruce forest ( Table 1 ). The established comparative trials
are situated in following provenance regions: B120, B230 (2), A120
and C210 ( Pârnu țăet al., 2012 ), where A is Eastern Carpathians, B is Cur-
vature Carpathians and C is Southern Carpathians. An incompletely bal-
anced design was implemented in all comparative trials, with 4
replications and 3 –9 seedlings per subdivided plot, planted at 2.5 by
2.5 m spacing, in which are presented all of the six families consist of de-
scendants of open or controlled pollinated seeds, collected from pendula
and common spruce trees in the Stâna de Vale population. In order to
provide more detailed information, one replication of the Comand ău
trial is presented in Supplement 1, as well as the mathematical model
of the full diallel design.
Measurements were taken after twenty years, for main traits of
growth, quality and adaptability of trees, such as:
– Diameter at breast height (Dbh) and diameter at 2.2 m (D
2.2), were
measure using forest calliper;
– Trees height (Th), was taken using a Vertex IV instrument;
– Crown diameter (Cd) projection, was taken using a telemeter;
– Number of branches per whorl (Nbw) in the whorl situated at 2.2 m
from the ground;
– Dominant branch diameter (Dbd), in whorl situated at 2.2 m from
the ground, was measured with electronic callipers. Dry branches
below the height of 2.2 m were removed to ensure accessibility in
the comparative trials;
– Survival rate for each family is reported by current number to the
initial ones.981 M. Budeanu et al. / Science of the Total Environment 689 (2019) 980 –990

Next traits were calculated based on field measurements:
– Tree slenderness was calculated using the formula: Ts ¼Th
Dbh;
– Trees volume was calculated using the regression equation method
elaborated by Giurgiu et al. (2004) ;
– Branches finesse was calculated with next formula: Bf ¼Dbd
D2:2/C2100
(%).
In two comparative trials, Comand ău and Lep șa 2, 12 trees (3 in each
replication) were selected from each family, belonging to the average
population diameter, for collecting core samples using the Presslerincrement borer, at breast height (1.3 m above ground). For the deter-
mination of conventional wood density, the method proposed by
Dumitriu-T ătăranu et al. (1983) was used, respectively:
1. From 10 trees uniformly distributed in the comparative trial cores
were extracted with the Pressler increment borer, which were used
to determine the momentary humidity.
2. Apparent density was determined at a humidity of 17% (average hu-
midity of the 10 cores). All cores were weighed using an electronic
balance with the accuracy of 0.01 mg. The diameter and length of
the cores were measured with an electronic calliper.
3. Based on the mass, the diameter and the length of the core, the density
was determined. The method involves determining the density of an-
hydrous wood ( ρ0) depending on the density at 17% humidity ( ρ17%):
Fig. 1. The locations of the field trials and provenance origin.
Table 1
Locations and environmental conditions of the trials.
Trials N latitude/E longitude/altitude (a.s.l.) Exposition/slope Soil type/site productivity MAT(°C)/SAP(mm) Carpathian divisions
Comand ău 45°42 ′/26°18 ′/1020 m North/16° Disticambosoil/S 4.8/880 Curvature (inner)
Lepșa 1 45°58 ′/26°34 ′/620 m –/plane surface Eutricambosoil/S 7.6/730 Curvature (outside)
Lepșa 2 45°57 ′/26°27 ′/780 m –/plane surface Disticambosoil/S 6.4/815 Curvature (outside)
Ilva Mic ă 47°20 ′/25°00 ′/720 m –/plane surface Disticambosoil/S 5.2/850 Eastern
Voineasa 45°14 ′/23°55 ′/1400 m Northeast/24° Podzol/M 4.0/1000 Southern
Site productivity (superior = S, middle = M, inferior); MAT = mean annual temperature, SAP = sum of annual precipitations.982 M. Budeanu et al. / Science of the Total Environment 689 (2019) 980 –990

ρ0¼ρ17%
1:1−0:84∙ρ17%
4. The total volume swelling coef ficient was calculated as: βv=2 8×
ρ0,
5. Conventional density was calculated as follows: ρc¼ρ0100
100țβv:
The data obtained for all traits were processed using the STATISTICA
10.0 ( Statistica, 2010 )a n dRs o f t w a r e( R Core Team, 2018 ). The
Kolmogorov-Smirnov test was applied to check normal distribution
and the assumptions of ANOVA were veri fied using Levene's test. For
the open pollinated trees of the two crown forms total variance was
separated in the variance due to replications, forms, families and the
variance of error (residual) by applying the ANOVA test ( Nanson, 2004 ).
Xijk¼mțαițβjțγkțεijk
where: mis overall average value, αiis component of ith replications ( i
=1…a),βjis component of jth groups (crown forms) ( j=1…b),γkis
component of kth families, and εijkis random error affecting ijkplots.
The crown form × locality interaction and the in fluence of locality
were determined with bifactorial ANOVA model ( Nanson, 2004 ;
White et al., 2007 ).
Xijk¼mțαițβjțαβijțεij
where: m,a n d βjare as above, αiis component of ith locations ( i=1…
a),αβijis interaction of ith locations with jth groups and εijis random
error.
Concerning the hybrids of the two crown forms, the factorial ANOVA
was used to determine the in fluence of locality, combinations and their
interaction. For the full diallel design the mathematical model used was
the Grif fing 3th method, a modi fied diallel mating design ( Griffing,
1956 ) with the following variation sources: replications, general and
speci fic combining abilities (GCA, SCA), maternal (m) and reciprocal
(R) effects and environmental effects ( Nduwumuremyi et al., 2013 ).
Thefixed effect (Model I) was attributed to all variance components,
this model being the most suitable for this experimental design (Sup-
plement 1). The signi ficance level of the differences was established
with the Fisher (F) test (5%, 1% and 0.1%).
The heritability was calculated using the formula: h2
A¼
2σ2
GCA
2σ2
GCAțσ2
SCAțσ2
Mțσ2
Rțσ2
e
kwhere σ2
GCA,σ2
SCAare the variances of
the general and speci fic combining abilities, and σ2
M,σ2
Randσ2
eare the
variances attributed to maternal, reciprocal and environmental effects
(error), while k is the number of replications ( Pârnu ță,2 0 0 8 ).
In addition, Pearson correlations between traits were also
determined.
3. Results
The average values of the analysed traits, both for the open polli-
nated families and the reciprocal hybrids between the two crown
forms, were calculated ( Table 2 ) and also the in fluence of testing sites
and families/combinations are highlighted ( Table 3 ).
3.1. Correlations
Highly signi ficant correlations between analysed traits (in most
cases) were observed ( Table 4 ) without noticing signi ficant differencesdepending on the crown form or the hybrids. A direct and highly signif-
icant correlation (r = 0.77***) was registered between Th and Dbh,
more intense (with 5.3%) for pendula trees compared to pyramidalis
(Table 4 ). A direct and very signi ficant correlation resulted between
the growth traits and the branches characteristics (Nbw, Dbd), with
negative effects on the quality of wood (for timber). Also, direct and
very signi ficant correlation resulted between growth traits and crown
diameter ( Table 4 ), with negative effects on the stability of Norway
spruce forests. Branches finesse was in fluenced only by Dbh (negative
and very signi ficant), and not by the thickness of the branches (Dbd,
Table 4 ). The same level of correlation was registered for open polli-
nated trees and for hybrids of the two crown forms (full diallel mating
design: P × P and N × N; factorial mating design: P × N and N × P).
3.2. Breast height diameter (Dbh)
For open pollinated trees, the mean value for Dbh registered for the
narrow crown form, in all experiments was 16.0 cm, which is 1.3%
higher than the value registered in pyramidalis . Signi ficant differences
in Dbh between the two crown forms were recorded in two compara-
tive trials. In Comand ău trial the average Dbh in pyramidalis form was
higher (+9.4%), while in the Lep șa 2 trial, the narrow crown spruce
had a higher value ( Table 2 ). The Dbh of the pendula form was higher
in three out of the fivefield experiments. A superior Dbh for all descen-
dants were registered in Lep șa2 ,I l v aM i c ăand Lep șa 1 trials compared
to the Dbh values registered in Comand ău and Voineasa trials. We con-
sider that the most plausible explanation for the observed lower Dbh
values might be that Voineasa test is located at the highest altitude,
and in Comand ău may be related to repeated damages caused by late
frosts on the terminal buds with a large number of trees presenting
2–3 and even 4 stems. Interventions are required in the Comand ău
test, to remove the extra stems, which could improve the results in
the coming years. Reduced variation was observed between families
of spruce trees with normal crown, compared with the families of
pendula form, which will be taken into account for selecting the most
vigorous trees (P6 descendants) for passing to the next generation.
The replication has a signi ficant in fluence (for different traits) in the
three trials located in the Curvature Carpathians (especially in Lep șat r i –
als), which strengthens the need to establish comparative trials with
replications.
The factorial ANOVA analysis ( Table 3 )s h o w sav e r ys i g n i ficant in-
fluence of the testing site and a distinctly signi ficant in fluence of the
locality × form interaction, the fact that both forms of spruce react to
the testing site. For the ensemble of all experiments, no signi ficant dif-
ferences were found between the two spruce forms. Also the hybrids
react differently to changes in the environmental conditions ( Fig. 2 ).
For the ensemble of all 5 trials, the hybrids of the two crown forms
registered superior (P × P, N × P) or inferior (P × N, N × N) results com-
pared to open pollinated pendula trees with highly signi ficant differ-
ences (p = 0.0000) between them ( Table 2 andFig. 2 ). In generally,
the pendula (P) and P × P registered superior Dbh while normal
crown form (N) and especially N × N presents lower Dbh values. The
lowest Dbh values registered by N × N in almost all trials (except
Comand ău trial) might be the consequence of a high number of consan-
guineous trees that survived until this age (96 trees, especially belong-ing to 41 × 41 combination), with a mean Dbh inferior by 24% to the
N open pollinated trees. However, consanguineous trees have reduced
the N × N average only with 3% both for Dbh and Th. Only 19 consan-
guineous pendula trees (belonging especially to 6 × 6 combination)
survived.
3.3. Trees height (Th)
For open pollinated trees there were no signi ficant differences for
the Th between the average values of the two crown forms ( Table 2 ).
The normal crown Norway spruce registered a plus of 2.4% and983 M. Budeanu et al. / Science of the Total Environment 689 (2019) 980 –990

superiority in 4 out of the 5 trials, with signi ficant values observed in
two of the trials, Comand ău and Lep șa 1 (p = 0.008, respectively p =
0.048). Only in the Ilva Mica trial was observed a higher Th mean
value for narrow crown spruce (+4.2%). At the family level, the P6
trees registered highest value, with a mean Th of 12.8 m, being exceeded
only by the descendants of the N40 tree, with 0.8%.
Also for this trait, the factorial ANOVA ( Table 3 ) indicates a very sig-
nificant in fluence (p b0.001) of the testing site, but also a distinctly sig-
nificant in fluence (p b0.01) of locality × form interaction. The both
forms have registered different results from one experiment to another.
For the ensemble of all 5 trials, the hybrids of the two crown forms
recorded average Th value close to the open pollinated trees, standingout those with mother N (N × P, Table 2 ).
3.4. Trees volume (Tv)
For Tv the descendants of pendula open pollinated spruce registered
a mean value (0.15 m
3) slightly higher (2.7%) compared to the normal
crown spruce families, but is statistically insigni ficant ( Table 2 ). The
families of pendula spruce registered superior Tv values in three out of
five trials, with the biggest difference being registered in the experiment
located in the Eastern Carpathians, Ilva Mic ă(+19.4%). The largest vol-
ume increases were recorded by descendants of the P6 (0.169 m3). Fac-
torial ANOVA analysis ( Table 3 ) shows a very signi ficant in fluence of the
testing site and, in the same time, signi ficant locality × form interaction.
The hybrids of the two crown forms were slightly superior to the
open pollinated trees, especially the trees belonging to N × P combina-tion ( Table 2 ).
3.5. Annual height increment (Ahi)
The annual height increment (Ahi) was higher for open pollinated
spruce trees with normal crown with mean value for all tests of
68 cm, being superior with 6.3% (signi ficant differences presented in
Table 2 ). At the trial level only in Comand ău we found signi ficant differ-
ences between the two crown forms and also the lowest values for Ahi
for all experiments were recorded. It is worth to mention that highest
values of Ahi were recorded in the Voineasa test, located at the highestaltitude, which can be explained by the high temperatures recorded in
one of the warmest years of meteorological measurements (year
2015). The factorial ANOVA ( Table 3 ) deviates from the trend previously
presented, this time, besides the very signi ficant in fluence of the test
site there is also a distinctly signi ficant in fluence of the crown form,
the fact that spruce trees with normal crown have higher growths
than pendula trees in all trials.
The hybrids of the two crown forms registered intermediate results
between the two open pollinated crown form trees, standing out again
the N × P combination ( Table 2 ).
3.6. Trees slenderness (Ts)
Trees slenderness (Ts) trait is expressed by the fraction of tree height
and Dbh and is of particular importance in order to assess the stability of
Norway spruce forests, estimating that a value of Ts higher than 80
could endanger the stability of stands against wind and snow damages
(Popa, 2005 ).
For open pollinated trees, the average value of Ts observed in all
pendula trees was 80, which is signi ficantly lower (p = 0.011) than
the average value registered for the trees of pyramidalis spruce form.
Lower values for Ts were observed for the pendula trees (favourable to
wind and snow resistance) in 4 out of the 5 comparative trials, with sig-
nificant differences observed in Lep șa 2 and Voineasa trials ( Table 2 ). In
three trials pendula spruce presented Ts b80, and in the other two tests
the Ts values were below 90, showing a high resistance of pendula
spruce trees to wind. The normal crown trees presented Ts value
lower than 90 in all experiments and in two tests the Ts values were
below the threshold of 80. The in fluence of the test site and the
locality × form interaction were statistically signi ficant and additionally
the form of the crown signi ficantly in fluenced the Ts, with those of the
pendula form exhibiting superior resistance to wind ( Table 3 ). At the
family level, again the P6 presented a Ts value of 76, being 11% lower
than the family P1. Thus, taking into account this trait, special attention
should be given to family P6 in the selection towards the pendula form.
The superiority of pendula trees for this important stability traits is
strengthened by P × P hybrids, which present lowest (favourable) aver-
age value of Ts (76) for the ensemble of 5 trials. In all trials it resulted aTable 2
Mean values (±standard deviation) of spruce forms and reciprocal hybrids and the signi ficant differences between forms.
Trial Forms and hybrids Dbh (cm) Th (m) Tv (m3) Ts (index) Ahi (cm) Cd (m) Nbw Dbd (mm) Bf (%)
Comand ău Normal (N) 14.0 ± 3.4a11.9 ± 2.1a0.107 ± 0.05 88 ± 14 52 ± 16a3.1 ± 0.5 6.6 ± 0.9 19.5 ± 5 15.0 ± 4a
Pendula (P) 12.8 ± 3.3 11.1 ± 1.4 0.090 ± 0.05 90 ± 15 47 ± 12 2.8 ± 0.6a5.9 ± 0.9a18.5 ± 4 16.2 ± 2
P × N 13.1 ± 3.5 11.2 ± 1.9 0.093 ± 0.05 89 ± 14 52 ± 15 2.8 ± 0.4 6.0 ± 0.9 18.4 ± 4.0 15.3 ± 2.6N × P 15.8 ± 3.7 12.2 ± 1.6 0.140 ± 0.07 81 ± 14 53 ± 14 3.1 ± 0.5 6.7 ± 1.0 20.9 ± 3.7 14.9 ± 3.1
Ilva Mic ă Normal 16.5 ± 3.5 14.4 ± 1.7 0.175 ± 0.09 90 ± 13 68 ± 15 3.4 ± 0.5 6.2 ± 0.4 19.9 ± 3 13.0 ± 2
Pendula 17.8 ± 3.3 15.0 ± 1.4 0.209 ± 0.08 87 ± 13 67 ± 16 3.1 ± 0.3
a5.9 ± 0.4a20.0 ± 3 11.9 ± 2a
P × N 16.6 ± 3.7 14.1 ± 1.7 0.165 ± 0.08 91 ± 13 75 ± 22 3.2 ± 0.4 5.9 ± 0.4 19.2 ± 3.8 12.9 ± 1.5
N × P 17.5 ± 3.8 14.8 ± 1.8 0.200 ± 0.09 88 ± 14 75 ± 21 3.2 ± 0.5 6.1 ± 0.5 20.4 ± 3.9 12.5 ± 1.1
Lepșa 1 Normal 16.8 ± 3.2 13.4 ± 1.5a0.168 ± 0.07 82 ± 8 76 ± 14 3.4 ± 0.6 6.3 ± 0.4 19.0 ± 3 11.8 ± 1
Pendula 16.3 ± 3.5 12.8 ± 1.5 0.154 ± 0.8 80 ± 12 72 ± 15 3.4 ± 0.6 5.9 ± 0.5a17.5 ± 4a11.1 ± 1a
P × N 17.8 ± 4.9 13.4 ± 1.8 0.187 ± 0.11 81 ± 17 76 ± 17 3.6 ± 0.9 6.1 ± 0.7 20.8 ± 5.2 12.7 ± 0.9
N × P 16.8 ± 4.2 13.3 ± 1.8 0.172 ± 0.10 83 ± 14 73 ± 16 3.5 ± 0.8 6.1 ± 0.5 19.9 ± 4.4 12.5 ± 1.0
Lepșa 2 Normal 17.7 ± 3.1 13.5 ± 1.3 0.182 ± 0.07 78 ± 9 63 ± 14 3.1 ± 0.4 8.1 ± 1.0 20.5 ± 3 12.4 ± 2
Pendula 18.7 ± 2.8a13.3 ± 1.3 0.200 ± 0.06 72 ± 9a59 ± 17 3.1 ± 0.4 7.1 ± 1.1a21.2 ± 3 12.3 ± 2
P × N 18.6 ± 2.7 13.6 ± 1.2 0.197 ± 0.07 75 ± 9 68 ± 16 3.2 ± 0.4 7.8 ± 1.0 20.9 ± 3.4 12.3 ± 1.6N × P 18.7 ± 3.2 13.7 ± 1.1 0.204 ± 0.07 75 ± 11 63 ± 18 3.1 ± 0.4 7.9 ± 1.1 21.4 ± 3.4 12.4 ± 1.7
Voineasa Normal 14.3 ± 2.1 10.6 ± 1.1 0.097 ± 0.03 76 ± 7 80 ± 13 3.1 ± 0.6 6.0 ± 0.9 17.9 ± 3 13.8 ± 2
Pendula 14.8 ± 2.7 10.4 ± 1.3 0.099 ± 0.04 72 ± 10
a75 ± 18 3.1 ± 0.5 6.0 ± 0.9 18.5 ± 3 13.6 ± 3
P × N 14.3 ± 2.5 10.0 ± 1.2 0.085 ± 0.03 75 ± 9 80 ± 17 3.2 ± 0.5 6.2 ± 1.0 17.7 ± 3.4 14.4 ± 2.6N × P 14.6 ± 2.6 10.6 ± 1.1 0.099 ± 0.04 75 ± 10 80 ± 16 3.1 ± 0.6 6.0 ± 0.9 18.1 ± 3.2 13.6 ± 2.5
All trials Normal 15.8 12.8 0.146 83 68
a3.2 6.6 19.4 13.2
Pendula 16.0 12.5 0.150 80a64 3.1a6.2a19.1 13.0
P × N 15.9 12.4 0.146 81 66 3.1 6.4 19.5 13.5N × P 16.6 12.8 0.160 79 68 3.2 6.6 20.2 13.3
Dbh = breast height diameter, Th = trees' height, Tv = trees' volume, Ts = trees' slenderness index, Ahi = annual height increment, Cd = crown diameter, N bw = number of branches
per whorl, Dbd = dominant branch diameter, Bf = branches' finesse.
a=s i g n i ficant differences between forms (p b0.05) revealed by ANOVA.984 M. Budeanu et al. / Science of the Total Environment 689 (2019) 980 –990

highly signi ficant difference between combinations and a highly signif-
icant in fluence of locality. The average Th value of consanguineous trees
(85), with 90 Ts average in Comand ău, is a consequence of low in-
creases, the respective trees being dominated and will be naturally
eliminated from trials in coming years.
3.7. Crown diameter (Cd)
For the ensemble of all trials Cd was lower for the pendula open pol-
linated spruce with 3.1% (signi ficant differences) compared to the pyra-
midal spruce. If in two trials (Comand ău and Ilva Mic ă) the differences
were signi ficantly and favourable to pendula spruce ( Table 2 ), in the
other three trials the results were similar for both crown forms of
Norway spruce. At the family level, P1 descendants highlighted the low-
est Cd values in 4 of the 5 experiments. The factorial ANOVA ( Table 3 )
indicated a highly signi ficant in fluence (p b0.001) of the testing site,
and also a distinctly signi ficant in fluence (p b0.01) of the ideotype.
High homogeneity was observed in the pyramidal spruce families
(reflected by standard deviation), all of them registering the same aver-
age value for the ensemble of all experiments, 3.2 m.
The hybrids P × P and P × N registered the same average value as P,
lower values compared to N and N × P average Cd values. Highly signif-icant differences (p = 0.0004) resulted between combinations for the
ensemble of all trials and, at the trial level, in 3 of them (except Lep șa
1a n dL e p șa 2 trials) have appeared the signi ficant differences between
combinations. The lowest Cd average values were registered by P, P × P,
P × N, in Comand ău trial, the maternal component playing an important
role again.
3.8. Branches characteristics
Number of branches per whorl (Nbw): the average value for Nbw at
the height of 2.2 m was 6.2 for narrow crown open pollinated spruce
trees, which is 6% lower than values observed for spruce trees with nor-
mal crown (signi ficant difference). In all trials, trees of pendula form
presented in generally smaller Nbw values compared to the normal
spruce crown form, differences being statistically signi ficant in 4 of the
five trials ( Table 2 ). The factorial ANOVA ( Table 3 ) indicated a highly
significant in fluence (p b0.001) of the testing site, of the crown form,
and also of the locality × form interaction. At the family level wereTable 3
The in fluence of the testing site and spruce forms/combinations, revealed by ANOVA.
Traits Factors Open pollinated trees Hybrids
DF MS p DF MS p
Dbh Locality (L) 4 537.2 0.0000 4 2329.8 0.0000
Forms (F)/combinations
(C)1 6.6 0.4314 34 137.4 0.0000
L × F/C 4 36.4 0.0086 136 25.1 0.0000
Error (trees) 632 10.6 2797 12.4
Th Locality 4 330.41 0.0000 4 1918.3 0.0000
F/C 1 8.18 0.0707 34 41.5 0.0000
L × F/C 4 8.44 0.0095 136 7.6 0.0000
Error (trees) 632 2.50 2797 2.5
Tv Locality 4 0.2901 0.0000 4 1.49495 0.0000
F/C 1 0.0029 0.4216 34 0.0406 0.0000
L × F/C 4 0.0129 0.0219 136 0.01191 0.0000
Error (trees) 632 0.0045 2797 0.00535
Ts Locality 4 6519 0.0000 4 21,742 0.0000
F/C 1 872 0.0110 34 2048 0.0000
L × F/C 4 355 0.0328 136 285 0.0000
Error (trees) 632 134 2797 161
Ahi Locality 4 17,393 0.0000 4 85,279 0.0000
F/C 1 2083 0.0027 34 1458 0.0000
L × F/C 4 78 0.8497 136 537 0.0000
Error (trees) 632 229 2797 316
Cd Locality 4 3.64 0.0000 4 18.74 0.0000
F/C 1 2.35 0.0035 34 2.07 0.0000
L × F/C 4 0.63 0.0576 136 0.74 0.0000
Error (trees) 632 0.27 2797 0.31
Nbw Locality 4 62.01 0.0000 4 310.3 0.0000
F/C 1 30.27 0.0000 34 5.1 0.0000
L × F/C 4 5.35 0.0000 136 1.5 0.0000
Error (trees) 632 0.75 2797 0.8
Dbd Locality 4 180.1 0.0000 4 1246 0.0000
F/C 1 9.3 0.3916 34 144 0.0000
L × F/C 4 31.4 0.0420 136 32 0.0000
Error (trees) 632 12.6 2797 15
Bf Locality 4 343.8 0.0000 4 1010.4 0.0000
F/C 1 3.6 0.4371 34 42.4 0.0000
L × F/C 4 23.7 0.0032 136 12.3 0.0000
Error (trees) 632 5.9 2797 6.4
Dbh = breast height diameter, Th = Trees' height, Tv = Trees' volume, Ts = Trees' slen-
derness index, Ahi = Annual height increment, Cd = crown diameter, Nbw = number ofbranches per whorl, Dbd = dominant branch diameter, Bf = branches' finesse. DF = de-
grees of freedom, MS = mean square. Signi ficant differences, at p b0.05, are highlighted
with bold.
Table 4
Pearson correlation coef ficients between analysed traits.
Normal crowned
Variables Dbh Th Tv Ts Ahi Cd Nbw Dbd Bf
Narrow crowned Dbh – 0.75 0.96 −0.72 0.17 0.64 0.39 0.71 −0.59
Th 0.79 – 0.82 −0.13 0.10 0.49 0.35 0.53 −0.52
Tv 0.95 0.84 – −0.57 0.15 0.60 0.36 0.67 −0.55
Ts −0.68 −0.14 −0.50 – −0.17 −0.48 −0.23 −0.54 0.39
Ahi 0.27 0.19 0.20 −0.28 – 0.16 −0.15 −0.07 −0.31
Cd 0.66 0.48 0.58 −0.54 0.35 – 0.08 0.51 −0.37
Nbw 0.41 0.26 0.35 −0.32 0.08 0.21 – 0.34 −0.20
Dbd 0.71 0.53 0.65 −0.55 0.06 0.49 0.27 – 0.08
Bf −0.58 −0.58 −0.50 0.30 −0.39 −0.42 −0.26 0.03 –
Narrow × normal Normal × narrow
Variables Dbh Th Tv Ts Ahi Cd Nbw Dbd Bf
Dbh – 0.73 0.96 −0.71 0.05 0.63 0.40 0.79 −0.51
Th 0.76 – 0.81 −0.08 0.07 0.36 0.31 0.54 −0.50
Tv 0.95 0.81 – −0.53 0.05 0.57 0.36 0.75 −0.44
Ts −0.70 −0.13 −0.53 – −0.04 −0.54 −0.25 −0.60 0.27
Ahi 0.20 0.19 0.17 −0.16 – 0.12 −0.12 −0.06 −0.20
Cd 0.68 0.47 0.61 −0.54 0.29 – 0.09 0.51 −0.35
Nbw 0.37 0.23 0.30 −0.33 −0.01 0.18 – 0.27 −0.26
Dbd 0.81 0.59 0.78 −0.61 0.07 0.58 0.27 – −0.02
Bf −0.57 −0.56 −0.48 0.35 −0.28 −0.37 −0.24 −0.06 –
Dbh = breast height diameter, Th = Trees' height, Tv = Trees' volume, Ts = Trees' slenderness index, Ahi = Annual height increment, Cd = crown diameter, N bw = number of branches
per whorl, Dbd = dominant branch diameter, Bf = branches' finesse. Signi ficant correlations (p b0.05) are bold it. Number of trees: Pendula = 321, Normal = 321, P × N = 607, N × P =
994.985 M. Budeanu et al. / Science of the Total Environment 689 (2019) 980 –990

remarked (smallest Nbw) the P7 (6.0 Nbw) followed by P1 (6.2 Nbw)
and by P6 (6.4 Nbw).
In each trial and also for the ensemble of the five trials, the hybrids P
× P and P × N, presents, as well as P open pollinated trees, the lowest
Nbw with signi ficant differences to the N × P and N × N combinations.
Dominant branch diameter (Dbd) and branches finesse (Bf): for the
ensemble of the 5 trials, average values for these traits were slightly
favourable to pendula open pollinated spruce (both Dbd and Bf were
smaller with 1.5%) but the differences between the two forms of spruce
were not signi ficant. Narrow crown Norway spruce presented thinner
branches in two trials (with signi ficant differences in Lep șa 1 trial),
and with lower values of Bf for descendants of pendula in 4 trials (except
Comand ău); in two of these trials differences being statistically signi fi-
cant ( Table 2 ). The factorial ANOVA ( Table 3 ) showed a highly signi fi-
cant in fluence (p b0.001) of the testing site and a signi ficant (p b
0.05) effect for the locality x form interaction. At the family level, thin-
ner branches present P1 (Dbd = 18 mm), while for Bf was remarked
P6 (Bf = 12.6%).
The hybrids with pendula mothers (P × N and P × P) were remarked
by thinning branches (except of Lep șa 1 trial), differences to the hybrids
with pyramidalis mothers being signi ficant in 4 trials. For the ensemble
of all 5 trials highly signi ficant differences (p = 0.0000) resulted be-
tween hybrids and the thinning branches presented the P × N, but
thicker than the pendula open pollinated trees.
3.9. Heritability
In all trials highest value for heritability, 0.44 (0.37 –0.56 amplitude),
resulted for Ts, one of the most important trees stability traits ( Table 5 ).
Next is Dbd, with an average heritability (0.22) which represent only a
half of the Ts heritability, but with important values in Comand ău( 0 . 4 0 )
and Lep șa 2 (0.39). For the same traits different values for genetic heri-
tability resulted from one trial to another. For example, Th registered a
very important rate of heritage in Lep șa 2 trial (0.46) while in three of
the other trials the heritability was lower than 0.06. For Cd and Dbd,
the highest heritability were registered in Comand ău trial while for
Nbw an important rate of heritage was obtained in Lep șa 2 trial. InIlva Mic ătrial, the trees presented highest rate of heritage for Bf
(Table 5 ). In generally, for heritability were registered low ( b0.4) to me-
dium (0.4 –0.7) values.
3.10. Survival rate
For the open pollinated progeny, survival rate was 84% for the nar-
row crown trees and 83% for the trees with normal crown. Apart from
the Lep șa 1 trial where the normal crown form of spruce resisted better
(+18%), the pendula trees displayed higher adaptability in the other 4
trials with biggest difference (13%) being registered in Ilva Mic ătrial
(Fig. 3 ). For the ensemble of all trials almost all hybrids (except of con-
sanguineous) registered superior survival rate than open pollinated
trees, higher values obtained for P × N and P × P (88% and 87%, respec-
tively). A highly signi ficant in fluence of testing site and not signi ficant
influence of combinations (p = 0.3966) were registered. Inside of the
trials, signi ficant differences between combinations were registered in
three of the five trials (except Comand ăua n dL e p șa 1). Only 42% of
the consanguineous trees were surveyed and many of them are domi-
nated and could dry up in the coming years. For the 41 × 41 consanguin-
eous survival rate was 69%, 2.4 times more than for the others (6 × 6 and
40 × 40).
Fig. 2. Mean Dbh values of pendula (P), normal (N) crown and their hybrids, in all trials. P = pendula and N = normal crown form.
Table 5
Heritability of the analysed traits in all 5 full sib trials.
Trials Traits
Dbh Th Tv Ts Ahi Cd Nbw Dbd Bf
Comand ău 0.21 0.00 0.19 0.42 0.18 0.29 0.16 0.40 0.05
Ilva Mic ă 0.08 0.06 0.00 0.39 0.00 0.14 0.06 0.06 0.29
Lep șa 1 0.17 0.04 0.12 0.44 0.00 0.00 0.00 0.07 0.24
Lep șa 2 0.20 0.46 0.06 0.56 0.00 0.00 0.37 0.39 0.08
Voineasa 0.25 0.25 0.21 0.37 0.00 0.11 0.20 0.19 0.18Average 0.18 0.16 0.12 0.44 0.04 0.11 0.16 0.22 0.17
Dbh = breast height diameter, Th = trees' height, Tv = trees' volume, Ts = trees' slender-
ness index, Ahi = annual height increment, Cd = crown diameter, Nbw = number ofbranches per whorl, Dbd = dominant branch diameter, Bf = branches' finesse.986 M. Budeanu et al. / Science of the Total Environment 689 (2019) 980 –990

3.11. Conventional wood density (Cwd)
Cores were sampled and the wood density was determined for open
pollinated trees of the two spruce forms in Comand ăua n dL e p șa 2 com-
parative trials because they displayed extreme results for growth traits
of trees. Signi ficant variations from one trial to another and even within
the same trial, between forms and even between families have beenregistered. Thus, Cwd was 0.371 g/cm
3in Comand ău trial, while in
Lepșa 2 trial the Cwd was lower by 9%. The pendula trees were superior
in Lep șa 2 trial and the pyramidalis trees at the Comand ău trial. A supe-
rior homogeneity for pendula trees was observed especially in Lep șa2
trial ( Fig. 4 ). By cumulating the data from all trees a slightly higher den-
sity (+2.9%) was registered for trees belonging to the normal spruce
crown form. A very high variability between the normal crown spruce
families in both tests resulted. Also, the same normal family displayed
the highest density in Comand ău test and the lowest in Lep șa 2 trial
(N39) and vice versa (N40). From the pendula families in both trials is
outstanding the P1 (highest Cwd), at the opposite pole being P6
(Fig. 4 ). Higher values of Cwd ensure superior resistance of tree stems
(Zeltin šet al., 2018 ).
4. Discussion
The continuously increasing demand in industry for wood and do-
mestic heating requires finding solutions to increase the production of
wood/ha. This could be achieved by: using in plantations the genetically
improved forest reproductive material, the adoption of optimal planting
schemes and timely execution of thinning, soil fertilization and the
adoption of optimum production cycles ( Routa et al., 2013 ;Pyörälä
et al., 2014 ;Kilpeläinen et al., 2016 ). In the same time, climate changeaffects the trees resistance and their ongoing physiological processes.
An increase in the time interval between two abundant fructi fications
and an increase frequency of strong winds that cause massive wind-
throw have been observed during the last decades especially in the
Norway spruce stands. In this context, the identi fication, selection and
promotion of ideotype trees, capable of exceeding the values described
in the production tables (for tree height/age), while presenting superiorquality wood and higher resistance to the action of perturbing factors,
are of great concerns to geneticists.
The narrow crown Norway spruce ideotype presents superior prop-
erties compared to the normal crown form of spruce, both in terms of
growth traits and wood quality and also demonstrate higher resistance
to disturbing factors (wind and snow). Cumulating data from 5 compar-
ative trials, the pendula spruce shows favourable values compared to
pyramidalis form of spruce for breast height diameter (Dbh), trees vol-
ume (Tv), trees slenderness (Ts, signi ficant), crown diameter (Cd, sig-
nificant) number of branches per whorl (Nbw, signi ficant), the
thickness of the branches (Dbd), branches finesse (Bf) and survival
rate (Sr), being slightly inferior for trees height (Th), annual height in-
crement (Ahi, signi ficant) and conventional wood density (Cwd).
Therefore, the traits used for the selection, Tv (quantitative), Ts and Cd
(resistance), Nbw and Dbd (qualitative and resistance) are favourable
to the pendula
ideotype. In general, the combinations of the two
crown forms follow the trend presented for the open pollinated with
the same mother (P × P and P × N the same as P, while N × P and N
× N as N). The effects of the General combining ability (GCA) and the
maternal effects (Supplement 2) show a superior GCA effect of pendula
1 parents in term of Dbh, Cd and Dbd, while pendula 6t r e e sp r e s e n t e d
superior GCA for Ts and Nbw. For maternal effects the results differs be-
tween trials. The important maternal effects were registered for the
Fig. 3. Survival rate of the pendula (P) and normal crown (N) forms of Norway spruce and of their hybrids. Trials: C = Comand ău, I = Ilva Mic ă,L 1=L e p șa1 ,L 2=L e p șa 2, V = Voineasa. P
= pendula and N = normal crown form. Signi ficant differences (p b0.05) between the results registered by all crown forms in different trials were resulted. Inside of the trials, signi ficant
differences between crown forms were registered in I, L2 and V trials.987 M. Budeanu et al. / Science of the Total Environment 689 (2019) 980 –990

pendula 7 mothers in term of Ts (except of Comand ău trial). For all of the
analysed traits, ANOVA (Grif fing Multienvironment RCBD method 3,
model 1) show a highly signi ficant in fluence of environmental condi-
tions (testing sites and replications in trials), combinations and the
site x combination interactions. The GCA play a signi ficant role for Ts,
Nbw and Dbd while maternal factor are signi ficant for Nbw and Th.
The 115 consanguineous trees (survival rate 42%) registered an av-
erage Dbh and Th values inferior with 25% and 23% respectively to the
average of all trees. The differences would have been much greater if
not died a large part of the consanguineous, neighbours with the re-
maining ones. In this way, those left behind have bene fited from in-
creased light, mineral and organic substances from the soil. Even so,
most of them are dominated and probably will disappear in the next
years.
In the 5 analysed full sib trials, the pendula character began to be vis-
ible only for a few years (while in Finland appears at 6 years), so the dif-
ference from the pyramidalis (in terms of crown diameter) is very low
but it will certainly increase in the next years. In the origin population
(Stâna de Vale) at the age of 145, the average crown diameter of the se-
lected pendula spruce trees is 40% lower than the average of the
pyramidalis trees.
The branches' characteristics analysed in this paper are of particular
importance both for the quality of wood and for the stability of thespruce forests. Thus, the number of branches per whorl and the thick-
ness of branches affects both the quality of wood by the number of
wooden knots and their size, and the strength of the tree by the amount
of snow retained in the canopy, directly proportional to the two traits
(Isik and Isik, 1999 ,Mäkinen and Hein, 2006 ,Moberg, 2006 ,Kantola
et al., 2007 ,Hein et al., 2008 ,Barszcz et al., 2010 ).
In the next generations of the breeding programme, the selection
should be directed towards choosing pendula trees with fewer and thin-
ner branches and having the Ts closest to the optimal value, 80. From
the narrow crown Norway spruce open pollinated families, it would
be necessary to promote the P1 family which, although it displays
values below the P6 family in terms of Dbh and Th, it demonstrated a
higher density of wood, a smaller crown diameter, number of branches
in whorls and thinner branches. Although the survival rate and the trees
slenderness are less favourable at P1, when compared with P6 descen-
dants, nevertheless the values recorded by the P1 family (Sr = 81%, Ts
= 85) are within the limits that ensure the stability of the spruce stands
(Barbu, 2004 ). Another option could be the selection of pendula at the
tree-level, establishing an order of importance of the traits, for example:
the thickness of the branches, number of branches per whorl, crown di-
ameter, wood density, the trees slenderness, etc.The same selection
procedure could be applied also to the most valuable trees of P × P
and P × N combinations. Special attention should be given to combina-
tions 1 × 6 and 6 × 1.
For the analysed traits, heritability was in generally low (h
2b0.4),
with exception of Ts which presents a medium rate of heritage. Forthe same trait, different heritability was registered in different environ-
mental conditions. For the growth traits (Dbh and Th) a medium herita-
bility was resulted for Th in Lep șa 2 trial (0.46), but the others trials
registered values lower than 0.25, meaning a low heritability, a similar
result to the research conducted by Pârnu ță(2008) . Also, the qualitative
traits, especially Ts and Dbd presents a superior rate of heritage compar-
ing to growth traits, comparable with the results of the research con-
ducted by Pârnu ță. For Cd, heritability has recorded an appreciable
value in Comand ău trial (0.29) while in the other two trials established
also in Curvature Carpathians (Lep șa 1 and 2) the heritage rate was al-
most zero. In Comand ău trial, and also in the Ilva Mic ătrial (where
was registered the second heritage rate for Cd) the pendula trees and
the P × P and P × N combinations (diallel pendula and mother pendula )
presented signi ficantly lower (favourable) Cd values than pyramidalis
trees and the combinations N × N and those with have pyramidalis as
mother (N × P).
The correlations between the analysed phenotypic traits are in line
with previous studies and indicate the need to implement a two-steps
selection strategy ( Budeanu et al., 2014 ). The classical correlation be-
tween the growth traits, reported in numerous studies previously car-
ried out for Norway spruce ( Ujvari and Ujvari, 2006 ;Kowalczyk et al.,
2007 ;Pârnu ță,2 0 0 8 ;Mihai, 2009 ;Levkoev et al., 2017 ), was obtained
(Th and Dbh, r = 0.77***). Negative correlation was observed betweennumber of branches per whorl and trees slenderness (r = −0.25***, for
all trees), with stronger negative correlation for narrow crown spruce (r
=−0.32***), which was also obtained in Germany ( Hein et al., 2008 ).
The pendula spruce form is recommended to dense planting
schemes, from 1 × 1 m, up to 2 × 2 m, which provides the increased pro-
duction per hectare without being necessary to apply too many inter-
ventions for thinning the stand ( Zubizarreta Gerendiain et al., 2008,
2009 ). Unfortunately, all fivefield tests analysed in this paper have
been placed at a less dense schemes, 2.5 × 2.5 m, the planting distance
at which, in Finland, the pendula ideotype has been inferior to the clas-
sical form of spruce, results opposite to that recorded in much more
dense trials, established at 1 × 1 m ( Zubizarreta Gerendiain et al.,
2007, 2008, 2009 ). The signi ficant differences between pendula
and
pyramidalis , in terms of number of branches per whorl were unexpected
and have not been reported before in other research studies. Certainly
there are particularities of the Carpathian pendula spruce compared
with the one from Finland. Investigations will continue at the DNA
level and it will be installed a clonal seed orchard (obtained by grafting)
forpendula spruce. All this to determine how the character is transmit-
ted and to increase the transmission rate.
5. Conclusions
Norway spruce narrow crown ideotype ( Picea abies pendula form)
and the combinations which have pendula as mother registered supe-
rior results than normal crown form of spruce ( pyramidalis form and
Fig. 4. Conventional wood density of the two forms of spruce in Comand ău (left) and Lep șa 2 (right) trials (1,6,7 = pendula , 39,40,41 = pyramidalis ). Highly signi ficant differences (p b
0.001) between families only in Comand ău trial. No signi ficant differences between crown forms, in both trials. Distinctly signi ficant in fluence (p = 0.002) of locality.988 M. Budeanu et al. / Science of the Total Environment 689 (2019) 980 –990

combinations which have pyramidalis as mother) for the main traits in-
volved in the selection strategy: growth traits (tree's volume), wood
quality (number of branches per whorl and branches finesse) and resis-
tance to wind and snow perturbations (survival rate, trees slenderness
and crown diameter). Conventional wood density, as an important
trait for the quality and resistance of wood, is slightly superior in
pendula trees in one of the comparative trials (Lep șa 2) but inferior in
the other (Comand ău).
Consanguineous trees registered only a 42% of survival rate, a 25%
lower Dbh, and 23% lower Th than all trees, and most of them are dom-
inated and probably will disappear in the next years.
For the ensemble of all 5 trials, the results were signi ficantly in flu-
enced by the testing site (p b0.001) and by the interaction between
the testing site and the spruce crown form (p b0.05), while the crown
form has in fluenced signi ficantly the trees slenderness, crown diameter,
number of branches per whorl and the annual height increment. Herita-
bility was in generally low (h2b0.4), with better result for Ts, which
presents a medium rate of heritage (0.44, reaching even the 0.56 value
in Lep șa 2 trial). For the same trait, different heritability was registered
in different environmental conditions. The results of GCA and maternal
effects together with the heritability are of great importance in the se-
lection strategy.
The correlations between phenotypic traits converged towards the
adoption of a two-steps breeding strategy, starting from the selection
ofpendula trees by stability traits, especially for Ts.
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.scitotenv.2019.06.518 .
Acknowledgements
This paper was financed by the Romanian Ministry of Research and
Innovation, in the frame of Nucleu Programme contracted with National
Institute for Research and Development in Forestry “Marin Dr ăcea”
(Projects PN18040203 and PN19070302). We wish to thank to our de-
voted colleagues: Dan Pepelea, Ivan Robert, Cristiana Dinu, Ștefan
Tănasie, Ioana Ple șca and Gabriela Lupu, for their help in the field mea-
surements and to Adrian Loren țfor the elaboration in GIS of Fig. 1 .T h e
authors express full consideration to the anonymous reviewers for their
important contribution to the manuscript improvement.
References
Acquaah, G., 2012. Principles of Plant Genetics and Breeding. 2nd ed. Wiley-Blackwell,
Oxford.
Anonymous, 2010. Amenajamentul O.S.P. Sfânta Maria, UP II Stâna de Vale (Management
planning of Sfânta Maria forest district, unit 2 Stâna de Vale) (in Romanian).
Apostol, E.N., Curtu, A.L., Daia, L.M., Apostol, B., Dinu, C.G., Șofletea, N., 2017. Leaf morpho-
logical variability and intraspeci fic taxonomic units for pedunculate oak and grayish
oak (genus Quercus L., series Pedunculatae Schwz.) in Southern Carpathian Region
(Romania). Sci. Total Environ. 609, 497 –505. https://doi.org/10.1016/j.
scitotenv.2017.05.274 .
Barbu, I., 2004. Gospod ărirea culturilor de r ășinoase instalate în afara arealului natural
(Management of resinous cultures installed outside of the natural distributionrange). Annals of “Ștefan Cel Mare ”University, Forestry Series, Suceava, pp. 37 –66
(in Romanian).
Barszcz, A., Sandalak, A., Sandalak, J., 2010. Knottiness of spruce stems from the Dolomites
as the basis for distinguishing quality zones in round wood. Folia For. Pol., ser. A 52,
89–97.
Bruschi, P., Vendramin, G.G., Bussotti, F., Grossoni, P., 2000. Morphological and molecular
differentiation between Quercus petraea (Matt.) Liebl. and Quercus pubescens Willd.
(Fagaceae ) in Northern and Central Italy. Ann. Bot. 85, 325 –333. https://doi.org/
10.1006/anbo.1999.1046 .
Budeanu, M., Șofletea, N., Petri țan, I.C., 2014. Among-population variation in quality traits
in two Romanian provenance trials with Picea abies L. Balt. For. 20, 37 –47.
Budeanu, M., Popescu, F., Șofletea, N., 2019. In situ conservation of forest genetic resources
in Romania. In: Šijačić-Nikoli ć, M., Milovanovi ć, J., Nonic, M. (Eds.), Forests of South-
east Europe under a Changing Climate. Conservation of Genetic Resources. Springer
International Publishing, Switzerland, pp. 195 –205.
Dumitriu-T ătăranu, I., Ghelmeziu, N., Florescu, I., Milea, I., Mo ș, V., Tocan, M., 1983.
Estimarea calit
ății lemnului prin metoda carotelor de sondaj (Estimation of Wood
Quality Using the Method of Cores Sampling). Tehnic ăPublishing House, Bucharest
(348 p. (in Romanian).Enescu, V., 1987. Crearea ideotipurilor de arbori cu constela ții optime de caractere și
însu șiri valoroase (Creating trees ideotype with optimal constellations of valuable
characters and traits). Revista P ădurilor 102 (1), 14 –18 (in Romanian).
Feurdean, A., Tan țău, I., F ărcaș, S., 2011. Holocene variability in the range distribution and
abundance of Pinus ,Picea abies ,a n d Quercus in Romania; implications for their cur-
rent status. Quat. Sci. Rev. 30, 3060 –3075. https://doi.org/10.1016/j.
quascirev.2011.07.005 .
Franji ć,J . ,L i b e r ,Z . , Škvorc, Ž., Idžojtić,M . , Šoštarić, R., Stan čić, Z., 2006. Morphological and
molecular differentiation of the Croatian populations of Quercus pubescens Willd.
(Fagaceae ). Acta Soc. Bot. Pol. 75, 123 –130. https://doi.org/ https://doi.org/10.5586/
asbp.2006.015 .
Geburek, T., Robitschek, K., Milasowszky, N., 2008. A tree of many faces: why are there
different crown types in Norway spruce ( Picea abies [L.] Karst.)? Flora 203,
126–133. https://doi.org/ https://doi.org/10.1016/j. flora.2007.01.003 .
Giurgiu, V., Decei, I., Dr ăghiciu, D., 2004. Metode și tabele dendrometrice (Methods and
Dendrometrical Tables). Ceres Publishing House, Bucharest (575 p. (in Romanian).
Griffing, B., 1956. Concept of general and speci fic crossing systems. Australian J. Sci. 9,
463–493.
Hein, S., Weiskittel, A.R., Kohnle, U., 2008. Branch characteristics of widely spaced
Douglas- fir in south-western Germany: comparations of modelling approaches and
geographic regions. For. Ecol. Manag. 256, 1064 –1079. https://doi.org/10.1016/j.
foreco.2008.06.009 .
I.N.S, 2016. Institutul Na țional de Statistic ă, statistica activit ăților din Silvicultur ă
(Romanian National Institute of Statistics, Statistics of Forestry Activities), Bucharest
(27 p. (in Romanian).
Isik, K., Isik, F., 1999. Genetic variation in Pinus brutia Ten. in Turkey II. Branching and
crown traits. Silvae Genet 48, 293 –302.
Kantola, A., Mäkinen, H., Makela, A., 2007. Stem form and branchiness of Norway spruce
as a sawn timber – predicted by a process based model. For. Ecol. Manag. 241,
209–222. https://doi.org/10.1016/j.foreco.2007.01.013 .
Karki, L., 1985. Genetically narrow-crowned trees combine high timber quality and high
stem wood production at low cost. In: Landsberg, J.J. (Ed.), Crop Physiology of Forest
Trees. Tigerstedt Publishing House, Helsinki, pp. 245 –256.
Kauppi, P.E., Posch, M., Pirinen, P., 2014. Large impacts of climatic warming on growth of
boreal forests since 1960. PLoS One 9, e111340. https://doi.org/10.1371/journal.
pone.0111340 .
Keenan, R.J., 2015. Climate change impacts and adaptation in forest management: a re-
view. Ann. For. Sci. 72, 145 –167. https://doi.org/10.1007/s13595-014-0446-5 .
Kilpeläinen, A., Alam, A., Torssonen, P., Ruusuvuori, H., Kellomäki, S., Peltola, H., 2016. Ef-
fects of intensive forest management on net climate impact of energy biomass
utilisation from final felling of Norway spruce. Biomass Bioenergy 87, 1 –8.https://
doi.org/10.1016/j.biombioe.2016.02.011 .
Kowalczyk, J., Nowakowska, J., Sulkowska, M., 2007. Norway spruce in the conservation of
forest ecosystems in Europe. IUFRO W.P.S. 2.02.11: Norway Spruce Provenances and
Breeding, Warsaw (21 p).
Kuuluvainen, T., 1988. Crown architecture and stemwood production in Norway spruce
(Picea abies (L.) Karst). Tree Physiol. 4, 337 –346.
Lehner, A., Campbell, M.A., Wheeler, N.C., Pöykkö, T., Glössl, J., Kreike, J., Neale, D.B., 1995.
Identi fication of a RAPD marker linked to the pendula gene in Norway spruce ( Picea
abies (L.) Karst. f. pendula ). Theor. Appl. Genet. 91, 1092 –1094. https://doi.org/
https://doi.org/10.1007/BF00223924 .
Lepistö, M., 1985. The inheritance of pendula spruce ( P. abies f.pendula ) and utilization of
the narrow – crowned type in spruce breeding. The Foundation for Forest Tree Breed-
ing Information. vol. 1, pp. 1 –8 (in Finnish with English abstract).
Levkoev, E., Kilpeläinen, A., Luostarinen, K., Pulkkinen, P., Mehtätalo, L., Ikonen, V.-P.,
Jaatinen, R., Zhigunov, A., Kangas, J., Peltola, H., 2017. Differences in growth and
wood density in clones and provenance hybrid clones of Norway spruce. Can. J. For.
Res. 47, 389 –399. https://doi.org/10.1139/cjfr-2016-0285 .
Mäkinen, H., Hein, S., 2006. Effect of wide spacing on increment and branch properties of
young Norway spruce. Eur. J. For. Res. 125, 239 –248. https://doi.org/10.1007/s10342-
006-0115-9 .
Mihai, G., 2009. Surse de semin țe testate pentru principalele specii de arbori forestieri din
România (Tested Seed Sources for the Main Species of Forest Trees in Romania).
Silvic ăPublishing House, Bucharest (281 p. (in Romanian with English abstract).
Mikola, J., 1993. Breeding of Norway spruce in Finland: problems and remedies. Proceed-
ings of IUFRO (S2.2 –11) Norway Spruce Provenances and Breeding Symposium,
Latvia, pp. 231 –239.
Moberg, L., 2006. Predicting knot properties of Picea abies andPinus sylvestris from generic
tree descriptors. Scand. J. For. Res. 21 (S7), 49 –62.https://doi.org/10.1080/
14004080500487011 .
Nanson, A., 2004. Génétique et amélioration des arbres forestières (Genetic and forest
trees breeding). Les presses agronomiques de Gembloux, Gemboux (712 p. (in
French).
Nduwumuremyi, A., Tongoona, P., Habimana, S., 2013. Mating designs: helpful tool for
quantitative plant breeding analysis. J. Plant Breed. Genet. 1, 117 –129.
Pârnu ță, G., 2001. Variabilitatea genetic ăa unor caractere ale descenden țelor biparentale
de molid pendula și comun, testate în culturi multista ționale (Genetic variability of
some traits of full sib descendants of pendula and common spruce tested in multisite
trials). Analele ICAS 44, 23 –28 (in Romanian with English abstract).
Pârnu ță, G., 2008. Variabilitatea genetic ăși ameliorarea arborilor de molid cu coroan ă
îngust ăîn România (Genetic variability and breeding of narrow-crown spruce trees
in Romania). vol. 181. Silvic ăPublishing House, Bucharest (in Romanian with English
abstract).
Pârnu ță, G., Stuparu, E., Budeanu, M., Sc ărlătescu, V., Marica, F.-M., Lalu, I., Filat, M.,
Tudoroiu, M., Loren ț,A . ,D a i a ,M . , Șofletea, N., Curtu, A.L., Pârnu ță, P., Nica, M.-S.,
Teodosiu, M., Che șnoiu, E.-N., Marcu, C., 2011. Catalogul na ți o n a la lr e s u r s e l o r989 M. Budeanu et al. / Science of the Total Environment 689 (2019) 980 –990

genetice forestiere (National Catalogue of Forest Genetic Resources). vol. 525. Silvic ă
Publishing House, Bucharest (in Romanian).
Pârnu ță, G., Budeanu, M., Stuparu, E., Sc ărlătescu, V., Che șnoiu, E.-N., Tudoroiu, M., Filat,
M., Nica, M.-S., Teodosiu, M., Loren ț, A., Daia, M., Dinu, C., 2012. Catalogul na țional
al materialelor de baz ăpentru producerea materialelor forestiere de reproducere
(National catalogue of basic materials for the production of forest reproductive mate-
rials). Silvic ăPublishing House, Bucharest (304 p. (in Romanian).
Popa, I., 2005. Wind throw-risk factor in mountainous forest ecosystems. Ann. For. Res.
48, 3 –28.
Pöykkö, T., 1993. A short-term breeding programme applying the ideotype concept.
Proceedings- Nordic Group for Tree Breeding, pp. 110 –117 Edinburg.
Pretzsch, H., Biber, P., Schütze, G., Kemmerer, J., Uhl, E., 2018. Wood density reduced while
wood volume growth accelerated in Central European forests since 1870. For. Ecol.Manag. 429, 589 –616. https://doi.org/10.1016/j.foreco.2018.07.045 .
Pulkkinen, P., 1991. Crown form and harvest in pendulous Norway spruce. Silva Fenn 25,
207–214.
Pulkkinen, P., Pöykkö, T., 1990. Inherited narrow crown form, harvest index and stem bio-
mass production in Norway spruce, Picea abies . Tree Physiol. 6, 381 –391. https://doi.
org/10.1093/treephys/6.4.381 .
Pyörälä, P., Peltola, H., Strandman, H., Kilpeläinen, A., Asikainen, A., Jylhä, K., Kellomäki, S.,
2014. Effects of management on economic pro fitability of forest biomass production
and carbon neutrality of bioenergy use in Norway spruce stands under the changing
climate. BioEnergy Res 7, 279 –294. https://doi.org/10.1007/s12155-013-9372-x .
R Core Team, 2018. A Language and Environment for Statistical Computing. Available on-
line. http://www.R-project.org/ , Accessed date: 6 May 2019.
Rodriguez, Y.P., Morales, L., Willför, S., Pulkkinen, P., Peltola, H., Pappinen, A., 2013. Wood
decay caused by Heterobasidion parviporum in juvenile wood specimens from
normal- and narrow-crowned Norway spruce. Scand. J. For. Res. 28, 331 –339.
https://doi.org/ https://doi.org/10.1080/02827581.2012.746387 .
Rosner, S., Sv ětlík, J., Andreassen, K., Børja, I., Dalsgaard, L., Evans, R., Karlsson, B.,
Tollefsrud, M.M., Solberg, S., 2014. Wood density as a screening trait for drought sen-
sitivity in Norway spruce. Can. J. For. Res. 44, 154 –161. https://doi.org/10.1139/cjfr-
2013-0209 .Routa, J., Kellomäki, S., Strandman, H., Bergh, J., Pulkkinen, P., Peltola, H., 2013. The timber
and energy biomass potential of intensively managed cloned Norway spruce stands.
GCB Bioenergy 5, 43 –52.https://doi.org/10.1111/gcbb.12002 .
Schlegel, R.H.J., 2010. Dictionary of Plant Breeding. 2 ed. CRC Press, Taylor & Francis
Group, Boca Raton.
Schmidt-Vogt, H., 1977. Die Fichte. Ein Handbuch in zwei Bänden. Band I. Taxonomie,
Verbreitung, Morphologie, Okologie, Waldgesellschaft. Paul Parey, Hamburg und Ber-
lin (647 p. (in German).
Șofletea, N., Curtu, L., 2007. Dendrologie. Editura Universit ății Transilvania, Bra șov, p. 418.
STATISTICA 10.0. StatSoft Inc., Tulsa, OK, USA, 2010.
Steffenrem, A., Solheim, H., Skrøppa, T., 2016. Genetic parameters for wood quality traits
and resistance to the pathogens Heterobasidion parviporum and Endoconidiophora
polonica in a Norway spruce breeding population. Eur. J. Forest Res. 135, 815 –825.
https://doi.org/10.1007/s10342-016-0975-6 .
Ujvari, E., Ujvari, F., 2006. Adaptation of progenies of a Norway spruce provenance test
(IUFRO 1964/68) to local environment. Acta Silv. Lignaria Hung. 2, 47 –56.
White, T.W., Adams, W.T., Neale, D.B., 2007. Forest Genetics. CABI Publishing, Cambridge
(682 p).
Zeltin š, P., Katrevi čs, J., Gailis, A., Maaten, T., B āders, E., Jansons, A., 2018. Effect of stem di-
ameter, genetics, and wood properties on stem cracking in Norway spruce. Forests 9,
546, 10 p. https://doi.org/ https://doi.org/10.3390/f9090546 .
Zubizarreta Gerendiain, A., Peltola, H., Pulkkinen, P., Jaatinen, R., Pappinen, A., Kellomäki,
S., 2007. Differences in growth and wood property traits in cloned Norway spruce
(Picea abies ). Can. J. For. Res. 37, 2600 –2611. https://doi.org/10.1139/X07-113 .
Zubizarreta Gerendiain, A., Peltola, H., Pulkkinen, P., Ikonen, V.-P., Jaatinen, R., 2008. Dif-
ferences in growth and wood properties between narrow and normal crowned
types of Norway spruce grown at narrow spacing in Southern Finland. Silva Fenn.
42, 423 –437. https://doi.org/ 10.14214/sf.247 .
Zubizarreta Gerendiain, A., Peltola, H., Pulkkinen, P., 2009. Growth and wood property
traits in narrow crowned Norway spruce ( Picea abies f. pendula ) clones grown in
southern Finland. Silva Fenn 43, 369 –382. https://doi.org/10.14214/sf.194 .990 M. Budeanu et al. / Science of the Total Environment 689 (2019) 980 –990