Development of Somatic Embryo Maturation and Growing Techniques of Norway Spruce Emblings towards Large-Scale Field Testing

Article Development of Somatic Embryo Maturation and Growing Techniques of Norway Spruce Emblings towards Large-Scale Field Testing Mikko Tikkinen * ID , Saila Varis and Tuija Aronen Natural Resources Institute Finland, FI-58450 Punkaharju, Finland; [email protected] (S.V.); [email protected] (T.A.) *Correspondence: [email protected]; Tel.: +358-29-532-8475 Received: 23 March 2018; Accepted: 12 May 2018; Published: 4 June 2018   Abstract: The possibility to utilize non-additive genetic gain in planting stock has increased the interest towards vegetative propagation. In Finland, the increased planting of Norway spruce combined with fluctuant seed yields has resulted in shortages of improved regeneration material. Somatic embryogenesis is an attractive method to rapidly facilitate breeding results, not in the least, because juvenile propagation material can be cryostored for decades. Further development of technology for the somatic embryogenesis of Norway spruce is essential, as the high cost of somatic embryo plants (emblings) limits deployment. We examined the effects of maturation media varying in abscisic acid (20, 30 or 60 µ M) and polyethylene glycol 4000 (PEG) concentrations, as well as the effect of cryopreservation cycles on embryo production, and the effects of two growing techniques on embling survival and growth. Embryo production and nursery performance of 712 genotypes from 12 full-sib families were evaluated. Most embryos per gram of fresh embryogenic mass (296 ± 31) were obtained by using 30 µ M abscisic acid without PEG in the maturation media. Transplanting the emblings into nursery after one-week in vitro germination resulted in 77% survival and the tallest emblings after the first growing season. Genotypes with good production properties were found in all families. Keywords: Norway spruce; Picea abies L. Karst.; somatic embryogenesis; forest biotechnology; forest regeneration material; cryopreservation; maturation; embling production 1. Introduction In Finland, the increased planting of Norway spruce (Picea abies L. Karst.) seedlings and difficulties in seed production has resulted in intermittent shortages of regeneration material of a high breeding value [ 1 ]. One solution to this problem is to use vegetative propagation, e.g., somatic embryogenesis (SE), which was observed in Norway spruce for the first time in 1985 [ 2 , 3 ]. Vegetative propagation enables more efficient tree improvement e.g., by capturing non-additive genetic gain [4,5]. Additionally, the cryopreservation of embryogenic tissue (ET) in liquid nitrogen (LN) enables long-term storage of regeneration material in its juvenile state [ 6 – 8 ]. Cryopreservation techniques are available for several conifer species, based on either applying cryoprotectant before freezing, or either on drying embryos or embryogenic tissues in different developmental stages [ 9 – 14 ]. Additionally, for Norway spruce, reliable cryopreservation protocols applicable for large number of samples have been developed [ 8 , 15 ]. As a result, acceptable recovery rates together with high morphological and genetic fidelity have been observed [8,15]. The commercial scale production of conifer emblings, i.e., somatic embryo plants, has been achieved in Denmark, Ireland and France (Abies,Picea and Pinus species) and is being piloted in Sweden Forests 2018,9, 325; doi:10.3390/f9060325 www.mdpi.com/journal/forests Forests 2018,9, 325 2 of 15 (Picea abies). Companies producing conifers (e.g., Pseudotsuga menziesii,Picea glauca engelmannii complex, Picea glauca (Moench) Voss and Picea sitchensis (Bong.) Carr., Pinus radiata D. Don, Pinus taeda L. and Pinus elliottii Engelm.) with SE for planting stock exists in North America and New Zealand [ 6 , 16 , 17 ]. The main reason limiting the commercial application of SE in forestry is the high cost of emblings compared to seedlings [ 17 , 18 ]. Several efforts to reduce costs have been made e.g., producing emblings as donor plants for rooted shoot cuttings, thus fragmenting the high cost of emblings to several hundred rooted shoot cuttings [ 17 , 19 ]. Despite all the efforts, emblings are still rather expensive compared to seedlings, which limits their deployment especially in Nordic conditions [ 20 ]. Additionally, the loss of genetic material during SE is a major challenge [ 21 ]. However, this could be mitigated by improving production methods in the post cryopreservation phases and in the laboratory-nursery interface [ 20 – 23 ]. Abscisic acid (ABA) is a relatively expensive plant hormone, widely used in conifer SE to promote embryo maturation [ 24 ]. It has a positive effect in promoting the maturation of embryogenic tissues, but it can also inhibit the germination and height growth of emblings for several growing seasons after exposure [ 7 , 25 – 27 ]. The type and amount of ABA concentration significantly affects the maturation results, and the optimal concentration varies between species and genotypes [27–29]. Similar, species and genotype specific, responses in embryo maturation have been reported when various amounts and types of polyethylene glycol has been added to the maturation medium in different conifer species [ 30 – 33 ]. Polyethylene glycol is added to the maturation media to reduce the moisture content of somatic embryos, thus increasing the content of the storage materials in later phases of maturation compared to ABA [ 31 , 32 , 34 ]. In Norway spruce, polyethylene glycol 4000 (PEG), when added to the maturation media, is known to increase the number of somatic embryos but is also known to have a negative effect on the later growth and development of the embryos [ 25 , 35 , 36 ]. PEG has also been found to speed up somatic embryo maturation by several weeks [31,36]. The aim of this work was to improve the efficiency of Norway spruce embling production in order to enable large scale testing of numerous SE lines. To achieve this we studied, (I) the effects of different levels of ABA and PEG in maturation media and the effect of an additional cryopreservation cycle on embryo production capacity; and (II) the effects of two different growing techniques on embling yield and early performance. Furthermore, the embryo production capacity and survival rate in the nursery were tested for a wide range of genotypes (712) originating from 12 full-sib families, with the aim to initiate field testing with rooted cuttings. This was done to improve the properties and yield of emblings (I and II) and to evaluate embling production schemes needed for large-scale field testing and variation among full-sib families affecting them (III). 2. Materials and Methods 2.1. Origin of Embryogenic Lines The embryogenic lines used in this study were initiated in 2014 and 2015 from immature zygotic embryos of full-sibling families of progeny tested plus trees from Southern Finland. The medium and methods developed by Klimaszewska et al. [ 37 ], as described by Varis et al. [ 8 ] were used for culture establishment. Zygotic embryos without megametophytes, were placed on a modified Litvay’s medium (mLM) containing half-strength macroelements [ 37 , 38 ], 10 µ M 2,4-dicholophenoxyacetic acid (2,4-D), and 5 µM 6-benzyladenine (BA). The sucrose concentration of the medium was 1% (w/v) and the pH was adjusted to 5.8 prior to adding gellan gum (4 g/L, Phytagel ™ , Sigma-Aldrich, Saint Louis, MO, USA) and autoclaving. The cultures were kept in the dark (at 24 ◦ C) for two to eight weeks without subculturing, until embryogenic tissue (ET) started to grow. Established ETs were subcultured bi-weekly, on a fresh Petri dish of the same medium. Cryopreservation of ETs was done according to Varis et al. [ 8 ]. From each genotype one to four samples were cryopreserved right after initiation. Slowly growing ETs were not cryopreserved. The number of samples per genotype was kept low to increase the number of genotypes in cryostorage. Forests 2018,9, 325 3 of 15 For maturation in all experiments, from 150 to 200 milligrams of fresh embryogenic mass was weighed and absorbed on filter paper (Whatman # 2), using a Buchner funnel as done by Varis et al. [8]. 2.2. Experiment I The effect of the ABA concentration on embryo production and plant viability was studied using three different trials (1, 2 and 3). In Trial 1, six continuously subcultured Norway spruce embryogenic lines from five families initiated in 2014 were matured in December 2015, and seven lines from four families initiated in 2015 were matured in February 2016. Each line was matured on six filter papers (Whatman # 2) which were placed in petri dishes filled with 28 mL of LM-media containing 60 or 20 µ M ABA (Later referred to 60ABA and 20ABA), and three maturations of each treatment (Table 1). Filter papers on the latter media were moved to fresh media two times at one-week intervals at the beginning of maturation. Cotyledonary embryos with visible initial shoot and root meristems and at least four cotyledons were manually counted after eight weeks maturation in the dark at 24 ◦ C room temperature, as was done in previous studies [39,40]. Table 1. Schematic description of experiments and treatments used in different phases of SE production in Experiments I to III. Exp./Trial Treatment Maturation Germination 1st Growing Period/Growing Season 2nd Growing Season I/1; 2; 3 60ABA 8 weeks I/1 3* 20ABA 8 weeks I/2 60ABA + PEG 8 weeks I/2 30ABA + PEG 8 weeks I/2; 3 30ABA 8 weeks II GT-I 8 weeks 18:6 Day-night Transplanted to Miniplugs Grown outside since 60 µM1 week (1) Controlled environment June 2017 30 µM ABA Transplanted to Plantek 81f Winterized and cold stored GT-II 8 weeks 18:6 Day-night Transplanted to Plantek 81f 60 µM1 week (1) Nursery greenhouse in March 2017 30 µM ABA Grown outside since June 2017 III Thawing lots A to D 8 weeks 18:6 Day-night Transplanted to Miniplugs Grown outside since 60 µM ABA 1 week (1) Controlled environment June 2017 Winterized and cold stored Thawing lot E 8 weeks 18:6 Day-night Transplanted to Miniplugs 30 µM ABA 1 week (1) Controlled environment Transplanted to Plantek 81f in March 2017 Grown outside since June 2017 Thawing lot F 8 weeks 18:6 Day-night Transplanted to Plantek 81f 30 µM ABA 1 week (1) Nursery greenhouse in March 2017 Grown outside since June 2017 In Experiment I, the effect of different concentrations/combinations of abscisic acid (ABA) and polyethylene glycol 4000 (PEG) on the yield of cotyledonary embryos was tested. 3* means a transfer of ET twice to fresh media. In Experiment II, two ex vitro growing techniques for emblings were tested (GT-I and GT-II). In Experiment III, large number of samples from 12 full-sib families was thawed from cryopreservation in lots A to F, to produce emblings (cutting donors) for clone testing. (1) three days in five µ mol/m −2 /s −1 , two days in 50 µ mol/m −2 /s −1 and two days in 150 µmol/m−2/s−1. Trial 2 consisted of seven genotypes (from different full-sib families) which were matured in May 2016. Two of the lines were initiated in 2015 and maintained in a subculture. Five lines initiated in 2014 were thawed from LN. Four different ABA (60 and 30 µ M) and PEG (4.75% concentration in media) combinations were used in mLM-media (later referred to 60ABA, 30ABA, 60ABA + PEG and 30ABA + PEG) (Table 1). The filters were kept in the original petri dishes for eight weeks under the same conditions as in Trial 1, after which cotyledonary embryos were counted. Trial 3 consisted of 120 cryopreserved genotypes from 12 families (10 genotypes per family), which were thawed from LN and matured using LM media with two different ABA concentrations (Table 1). The first lot of 120 genotypes was thawed and matured in 2016. From these genotypes, Forests 2018,9, 325 4 of 15 samples were cryopreserved again and one sample per genotype was thawed and matured in 2017. From the first lot, samples from 65 genotypes were matured on media containing 60 µ M of ABA, and 55 genotypes were matured on media with a 30 µ M ABA concentration. In the second lot, all 120 genotypes were matured on media containing 30 µ M ABA. The filters were kept in the original Petri dishes for eight weeks in the same conditions as Trial 1, after which cotyledonary embryos were counted. 2.3. Experiment II Two different growing techniques (later referred to GT-I and GT-II) in a nursery were evaluated by germinating cotyledonary embryos from 18 genotypes (9 families) according to the 1w-filter protocol described by Tikkinen et al. [ 23 ]. In short: cotyledonary embryos, cold stored at +3 ◦ C (for at least four weeks in a large refrigerator unit) on the same filter papers in Petri dishes where the maturation was carried out, were germinated one-week in vitro under LED (Light emitting diode) lights (at a temperature from 20 to 23 ◦ C inside the Petri dishes). The emblings were transplanted to a peat-based growth media after germination in vitro , using the ‘pricking out’ method as described by Landis et al. [23,41], in which forceps were used to transfer the emblings and to place them in peat. Peat was gently compressed around the embling to provide sufficient edaphic conditions for developing roots, as demonstrated by Landis et al. [23,41]. In GT-I, 36 emblings from each genotype were transplanted in small containers (Preforma 126/JIF, ViVi Pak, ViVi, Burgh Haamstede, Netherlands) with 126 plugs per container (plug volume 3.4 mL) (Miniplugs), and grown in a controlled environment for 50 days, until the temperature sum reached 1300 degree days (d.d.) (later referred to as the growing period). The controlled environment refers to a growth room, where the light period, temperature and humidity were adjusted to obey suggested levels for the different stages of growth of Norway spruce seedlings [ 42 ]. After this artificial growing season, the emblings were transplanted into Plantek 81f containers, winterized and cold stored in a large cooler unit. After cold storage the emblings were transferred outside together with a large lot of seedlings (Table 1). In GT-II, 81 emblings from each genotype were transplanted straight into Plantek 81f containers (81 separate ventilated compartments of 85 cm 3 size) and were grown in a greenhouse as described by Tikkinen et al. [ 23 ]. These emblings were grown together with the emblings from thawing lots A to D of Experiment III (Table 1). 2.4. Experiment III To initiate the field testing of SE-lines, emblings were produced for donor plants for shoot cuttings. This was done with cryopreserved genotypes from 12 full-sib families, initiated in 2014 (Table 2). The aim was to produce emblings from 20 genotypes from each full-sib family to initiate field testing with rooted shoot cuttings. ETs were thawed from 712 genotypes at six different times, in thawing lots A to F. Thawings were carried out during 2016 (A to E) and 2017 (F) (Table 2). ETs were thawed, subcultured bi-weekly, cryopreserved again and matured (three Petri dishes each) in a five or ten weeks production cycle [ 8 , 37 , 38 ]. If the ET did not proliferate enough for cryopreservation and maturation in ten weeks it was discarded. The ETs were cryopreserved again to increase the number of samples from each genotype for future use. Cryopreservation was prioritized, so that maturation was delayed for five weeks, if necessary. In the cases of poorly proliferating ETs, they were matured first and cryopreserved only if enough embryogenic tissue was available after 10 weeks of proliferation. This was done to increase the number of genotypes available for field testing. Forests 2018,9, 325 5 of 15 Table 2. Description of material included in Experiment III. Crossing Explants Initiated Cryostored Thawing Thawed Cryo + Maturation E/gFW Over 200 E/gFW Surv. % A to D Surv. % E to F E1551 ×E2229 99 67 61 III, IV and VI 48 45 94 (±13) 7 28 (±2) 72 (±2) E162 ×E81 200 101 66 II and VI 48 42 89 (±12) 5 14 (±2) 87 (±2) E18 ×E436 200 133 127 III and VI 47 30 110 (±18) 6 62 (±2) 85 (±2) E207 ×E1373 200 145 140 III, IV and VI 50 36 81 (±11) 3 54 (±2) 70 (±5) E207 ×E252 200 60 52 II, IV and VI 51 43 124 (±14) 6 30 (±2) 78 (±2) E2105c ×E2283 200 114 98 II, IV and VI 50 45 40 (±9) 1 14 (±2) 66 (±5) E212 ×E54 200 127 92 I, IV and V 92 61 45 (±7) 2 14 (±2) 80 (±2) E242 ×E222 400 187 106 I, IV, V and VI 106 49 36 (±7) 2 10 (±2) 82 (±2) E46 ×E3222 400 244 209 III and VI 78 53 89 (±12) 5 25 (±2) 88 (±2) E462 ×E64 200 160 137 I and VI 48 41 81 (±13) 5 54 (±2) 81 (±1) E799 ×E1366 45 42 36 II and IV 34 32 144 (±21) 7 24 (±2) E9 ×E3231 400 132 123 I and VI 60 52 68 (±10) 5 5 (±1) 71 (±2) Overall 2744 1512 1247 I to VI 712 529 79 (±4) 54 30 (±1) 80 (±1) Number of explants, genotypes initiated and cryopreserved in 2014. Thawing lot, number of genotypes thawed, cryopreserved again and matured. Yield of cotyledonary embryos (E/gFM) from three maturation dishes (mean ± standard error), as well as the number of genotypes producing over 200 E/gFW and the survival rates (Surv. %) of emblings (mean ± standard error) from different crossings from thawing lots A to D and E to F, in Experiment III. Forests 2018,9, 325 6 of 15 2.4.1. Thawing, Proliferation Maturation and Germination After eight weeks of maturation under the same conditions as Experiment I, cotyledonary embryos were manually counted and cold stored on filter papers, as described by Tikkinen et al. [ 23 ]. In the case of thawing lot B, the maturation dishes were moved to cold storage before counting. The maturation medium contained 60 µ M ABA, in thawing lots A to D, and 30 µ M ABA in thawing lots E and F (Table 1). Maturation media was changed between thawing lots D and E, because of the higher yield of cotyledonary embryos in media with a lowered ABA concentration observed in Experiment I. 2.4.2. Embling Production A one-week in vitro germination protocol was used in all cases following the methods described by Tikkinen et al. [ 23 ]. Because of limited resources, e.g., work force and growing space, for thawing lots A to E, only up to 36 cotyledonary embryos from each genotype were selected for cultivation depending on the availability of cotyledonary embryos. The cotyledonary embryos were grown as in GT-1 (Table 1). The emblings from thawing lot E were grown according to GT-I, with the exception that the emblings were transferred straight to the nursery in spring 2017, while they were still growing height (Table 1). Germinated emblings from thawing lot F were grown as in GT-II (Table 1). From thawing lots E and F, up to 81 cotyledonary embryos were selected for cultivation depending on the availability of cotyledonary embryos. 2.5. Measurements and Data Analysis Cotyledonary embryos were counted and the embryo productivity was calculated per one gram of fresh cell mass (E/gFM) in all experiments. Mean values are presented with their standard errors ( ± ). All measurements and inventory results were analyzed using the IBM SPSS Statistics 22 software package (International Business Machines Corporation, Ammonk, NY, USA). The level of confidence used was 5%. In Experiment I, nonparametric tests were used because normal distribution could not be assumed. In Trials 1 and 3 of Experiment I, differences between treatments were analyzed with Mann–Whitney U-test. In Trial 2 of Experiment I, the Kruskal–Wallis test (one-way ANOVA on ranks) was used to analyze differences between treatments. To compare the two growing techniques in Experiment II, the survival of the emblings was inventoried from GT-I after the first growing period in a controlled environment, i.e., before cold storage. The survival and height measurements for the 2017 growing season, were obtained for both growing techniques. Logistic regression was used to examine the differences in survival between growing techniques, after the first growing period and after the growing season of 2017. A non-parametric test (Mann–Whitney U) was used to test the differences in embling height between the two growing techniques, because a normal distribution could not be assumed. In Experiment III, the differences in embling survival between families were examined with logistic regression after the first growing period and after the growing season of 2017. A crossing covariate was used to investigate a possible parental effect. A thawing covariate was used to distinguish differences between different thawing lots. Among thawing lots, variation occurs in the date of thawing, the ABA concentration in the maturation media, the growing method and the lenght of the cold storage period. The effect of the thawing lot had no effect on the percentage of cases predicted correctly and was left out from the final models. The effect of location inside the containers was studied by using row and column covariates, which defined the location of a single embling inside a container. This effect was significant, but explained only 0.1% of the correctly predicted cases; hence row and column covariates were excluded from the final models. Embryo production between full-sib families was analyzed with Kruskal-Wallis and Mann–Whitney U nonparametric tests, because normal distributions could not be assumed. Forests 2018,9, 325 7 of 15 3. Results 3.1. Effect of ABA Concentration and PEG in the Maturation Media In Trial 1, with the first group of genotypes (Experiment 1), the mean yield of cotyledonary embryos was 80 ± 23 E/gFM when the media contained 60 µ M ABA, and 185 ± 24 E/gFM when the ETs on filters were twice transferred to fresh media containing 20 µ M ABA (Figure 1). Reducing the ABA concentration in the maturation media enhanced embryo production by 131% (p< 0.001). Figure 1. Yield of cotyledonary embryos in different treatments in three Trials in Experiment I. Mean values are presented with standard error bars. In Trial 3, values are presented for genotypes for which maturation with both treatments was available. In Trial 2, reducing the amount of ABA enhanced the productivity even though the cell mass was not transferred to fresh media. The mean productivity on 30 µ M ABA was 296 ± 31 E/gFM while on 60 µ M ABA it was 139 ± 28 E/gFM, with the increase being 113% (p= 0.001) (Figure 1). Adding PEG increased the productivity only when combined with 60 µ M ABA (211 ± 24 E/gFM). With 30 µ M ABA the number of cotyledonary embryos decreased to 179 ±22 E/gFM when PEG was included in the media (Figure 1). In Trial 3, the average yield of cotyledonary embryos among genotypes in the first lot was 180 ±9 E/gFM. For the 65 genotypes matured on 60 µ M ABA, the average embryo yield was 191 ±1 E/gFM. For the 55 genotypes matured on 30 µ M ABA, the embryo yield was 166 ±14 E/gFM . In the second lot, following cryopreservation, with 30 µ M ABA used in the maturation media for all genotypes, the overall embryo yield was 206 ± 19 E/gFM. For the genotypes matured with 60 or 30 µ M ABA in the first lot, the average embryo yields in the second lot were 240 ± 28 E/gFM and 167 ± 24 E/gFM, respectively. In the second lot, 116 genotypes were successfully regenerated from cryostorage and 114 were matured. No significant change in the average embryo production was found between lots when separately examining the genotypes matured on media containing 30 µ M ABA in both lots, or between genotypes matured with 60 µM ABA in the first lot. 3.2. In Vitro Germination, Survival and Height Growth In Experiment II, all selected cotyledonary embryos germinated and were transplanted ex vitro in both growing techniques. In GT-I, the average survival rate was 93 ± 1%, after the growing period, before cold storage. After cold storage and the growing season of 2017 in the nursery, the average survival rate and height were 53 ± 2% and 5.8 ± 0.3 cm, respectively. In GT-II, the average survival rate and height of the emblings after the first growing season (2017) in the nursery were 77 ± 1% and 10.9 ±0.2 cm, respectively. The difference in the embling survival rates between GT-I (after first growing period) and GT-II (after the first growing season of 2017) was significant (p< 0.001) (Table 3). The survival and height of the emblings obtained from different growing techniques varied after the growing season of 2017 (p< 0.001, in both) (Table 3). Forests 2018,9, 325 8 of 15 Table 3. Logistic regression models used for analyzing binary response (living or dead) in different growing techniques (GT-I and GT-II), after first growing season in the nursery or artificial growing season in growth room (GP1) and after the growing season of 2017 in the nursery, for Experiments II and III. Experiment Measurement Model Variable Sig. Odds Ratio (95% Confidence Interval) Correct, % Experiment II After GP1 log(p/1−p) = 2.839 −1.372g1−0.076c2−0.437c4−0.748c6−0.820c7+ 0.470c8−0.748c9−0.163c10 −0.546c11 81 Growing technique <0.001 1 GT-I 0.254 (0.177–0.363) GT-II Family <0.001 1 E1551 ×E2229 0.927 (0.591–1.453) E18 ×E436 0.623 (0.341–1.138) E207 ×E252 0.473 (0.293–0.765) E212 ×E54 0.44 (0.25–0.775) E242 ×E222 0.473 (0.293–0.765) E462 ×E64 1.6 (0.956–2.676) E46 ×E3222 0.849 (0.509–1.419) E799 ×E1366 0.579 (0.324–1.035) E9 ×E3231 Experiment II Autumn 2017 log(p/1−p) = 0.386 + 1.210g1+ 0.150c2−0.437c4−0.588c6−0.654c7+ 0.491c8−1.248c9−0.899c10 −1.095c11 74.1 Growing technique <0.001 1 GT-I 3.354 (2.684–4.191) GT-II Family <0.001 1 E1551 ×E2229 1.162 (0.785–1.72) E18 ×E436 0.646 (0.368–1.134) E207 ×E252 0.556 (0.361–0.855) E212 ×E54 0.52 (0.31–0.873) E242 ×E222 0.287 (0.188–0.437) E462 ×E64 1.634 (1.06–2.52) E46 ×E3222 0.407 (0.266–0.622) E799 ×E1366 0.335 (0.203–0.55) E9 ×E3231 Experiment III After GP1 log(p/1−p) = 0.859 + 0.238c 1 + 0.945c 2 + 0.421c 3 + 0.018c 4 + 0.104c 5 + 0.405c 6 + 0.419c 7 + 1.206c 8 + 0.651c 9 + 0.574c 10 −0.003c11 78 Family <0.001 1 E1551 ×E2229 1.269 (1.019–1.58) E162 ×E81 2.573 (2.035–3.252) E18 ×E436 1.523 (1.188–1.954) E207 ×E1373 1.018 (0.83–1.25) E207 ×E252 1.11 (0.832–1.48) E2105c ×E2283 1.499 (1.183–1.899) E212 ×E54 1.52 (1.197–1.931) E242 ×E222 3.34 (2.601–4.289) E462 ×E64 1.918 (1.575–2.336) E46 ×E3222 1.775 (1.371–2.298) E799 ×E1366 0.997 (0.812–1.225) E9 ×E3231 Forests 2018,9, 325 9 of 15 Table 3. Cont. Experiment Measurement Model Variable Sig. Odds Ratio (95% Confidence Interval) Correct, % Experiment III Autumn 2017 log(p/1−p) = −2.652 + 0.613b1+ 1.186b2+ 0.915b3+ 3.753b4+ 3.334b5+ 0.632c1+ 1.186c2+ 0.886c3+ 0.912c4+ 0.108c5+ 0.365c6+ 0.401c7+ 1.494c8+ 0.831c9+ 0.840c10 + 0.126c11 76.5 Thawing <0.001 1 A 1.846 (1.326–2.571) B 6.66 (4.935–8.988) C 2.497 (1.714–3.638) D 42.655 (29.934–60.782) E 28.049 (21.653–36.334) F Family <0.001 1 E1551 ×E2229 1.881 (1.427–2.479) E162 ×E81 3.275 (2.659–4.034) E18 ×E436 2.425 (1.931–3.046) E207 ×E1373 2.49 (1.907–3.25) E207 ×E252 1.114 (0.776–1.601) E2105c ×E2283 1.441 (0.973–2.135) E212 ×E54 1.493 (1.049–2.123) E242 ×E222 4.456 (3.266–6.08) E462 ×E64 2.297 (1.905–2.769) E46 ×E3222 2.316 (1.659–3.232) E799 ×E1366 1.134 (0.88–1.462) E9 ×E3231 In the models g1 is a design variable for growing technique II, c1 to c11 are design variables for full-sib families (Family) and b1 to b5 are design variables for thawing lots (Thawing).