Abstract: In this study, the phenotypic correlation between several seed source and seedling characters on wild service tree (Sorbus torminalis L. Crantz.) was investigated by a half-sib progeny test in North of Iran. Seeds of Ashak and Sangdeh Provenances (2100-2300 and 1600-1800 m a.s.l., respectively) were collected from the individual parent trees (20 single trees from each provenance) and sown in a mountain nursery (Orimelk, 1550 m a.s.l) based on a randomized complete block design with three replications. Both provenances showed a high and positive significant correlation between the characters measured on progenies. Although Sangdeh Provenance was selected as a proper site for seed collection and seedling production in this nursery, because of limitation of genetic variation and conservation of species, the selection of well-formed parent wild service trees of Ashak Provenance could be a good measure for this target. It can be concluded that at selection stage, if genotypes with greater collar diameter and height and less branch angles be selected as the parent trees, the seedlings (progenies) with higher survival and growth would be achieved.
INTRODUCTION
Wild service tree (Sorbus torminalis L. Crantz) is one of the valuable economical (Piagnani Bassi, 2000; Demesure et al., 2000; Eriksson, 2001) and medicinal plant species in the North forests of Iran, distributed from Astara to Golidaghi more or less as small groups in mixed beech, hornbeam and oak stands (Espahbodi, 2005; Tabandeh et al., 2007). It is a tall tree and can get to 34 m in height and 100 cm in diameter at breast height (Espahbodi, 2005; Tabandeh et al., 2007). It enhances the biological values of forest and has the especial importance in afforestation and tree breeding (Demesure et al., 2000). Due to its high value and the hazards threatening it (Bednorz et al., 2006), seedling production of this species is promoted for afforestation programs. It should be noted that from point of view of the morphological characteristics for successful implementation of such programs, selection of better quality seed source (parent trees) by specific desirable morphological characters are required for seed collection and other practices, such as breeding.
It seems the populations of Sorbus torminalis tend to have different genetically characters along altitudinal gradients in response to natural selection (Saenz-Romero et al., 2006). In general, populations from mild environments at low altitudes have larger growth potential, longer period of shoot elongation and less resistance to frost damage, than population originated from colder environments at higher altitudes. It is expected that during provenance-progeny test of Sorbus torminalis, significant variation due to provenances and especially due to families within provenances is found in growth and wood density (Sotelo Montes et al., 2006). Likewise phenotypic and genotypic correlations may indicate that larger trees tend to have denser wood.
Tree growth is often evaluated by height, diameter at breast height and standing volume. In tree breeding, height of the tree is taken more important than the other growth characters. It is due to the ease of measurement (Adams et al., 2001), its high heritability (Espahbodi, 2005; Tabandeh et al., 2007) and its high correlation with diameter (Yanchuk and Kiss, 1993; Wu et al., 1995) is a common measure of wood production and marketing in forestry.
The decisive goal of the tree breeding projects is to improve tree species in terms of the quality and the quantity of wood production. This can be achieved through the selection of superior seed source and genotypes so that an indirect selection is performed (Chaturvedi and Pandey, 2004).
Correlation analysis, though frequently used in agricultural crops, has only recently been used on tree species (Gera et al., 1999). Due to long rotation of tree life span, the analysis of seedling characters is an important technique (Chaturvedi and Pandey, 2004). Present study aims to determine the better parent trees (seed origin) of wild service tree (Sorbus torminalis L. Crantz.), having the higher values in economical characteristics, based on the correlation between characters measured on parent trees and their progenies.
MATERIALS AND METHODS
In October 2002 seeds were collected on 40 parent trees (genotype) located in Ashak (53 ° 18` 0" N longitude, 36 °7` 10" E latitude, 2100-2300 m a.s.l.) and Sangdeh forest (53 ° 10` 30" N longitude, 35 °58` 30" E latitude, 1600-1800 m a.s.l.) North of Iran and sown in nursery bed of Orimelk (1550 m a.s.l.). Nursery climatic data based on Tabandeh et al. (2007) was as Table 1.
The grown seedlings were replanted in the second year in the field experiment based on a randomized complete blocks design with 3 replications. According to Cornelius et al. (1996) in each replication from each genotype 10 seedlings (progenies) were replanted with distance of 0.5 m from each other. During the progeny test (years 2 and 3) weeds were removed four times per year and seedlings irrigated once a week. At the end of the third growing season (in October 2005) collar diameter (in mm), height and length of longest branch (cm), angle of the longest branch (degree) and number of branches of the progenies were recorded. Vitality of seedlings was recorded based on Savolainen et al. (2004) as given in Table 2.
| Table 1: | Nursery climatic parameters |
| Table 2: | Classification of leaf vitality (Savolainen et al., 2004) |
PROCE MEANS (SAS Institute, 1989) was used for determining the means characteristic. Then, correlation coefficient analysis was conducted for studying the effect of provenance on progenies characteristics.
RESULTS
The means of survival, vitality, collar diameter, height, number of branches, angle and length of the greatest branch in Ashak provenance are 69.67%, 4.41 (Score), 12.38 mm, 60.38 cm, 6.24 units, 60.22 ° and 17.58 cm, respectively (Table 3, 4). The means for the characters on Sangdeh provenance are 83.67%, 4.31 (Score), 13.96 mm, 74.65 cm, 9.78 units, 57.61 ° and 22.67 cm, respectively. The coefficient of variation for survival, vitality, collar diameter, height, number of branch, angle and length of the longest branch in Ashak provenance are 29.66%, 11.45 (score), 30.66 mm, 44.47 cm, 54.06 units, 28.94 ° and 53.04 cm, respectively. The coefficient of variation for these characters in Sangdeh provenance is 22.44%, 10.83 (score), 26.31 mm, 38.45 cm, 51.26 units, 34.20 ° and 46.85 cm, respectively.
In both provenances, there was mostly positive significant correlation between the characters measured on the progenies (Table 5, 6). Based on Ashak provenance a negatively significant phenotypic correlation was found between vitality and other characters measured (except branch number and angle of the longest branch), between survival and angle of the longest branch and between length of the longest branch and angle of the longest branch. Correlations were insignificant only between branch number with vitality and angle of the longest branch with vitality.
| Table 3: | Minimum, maximum, Mean ± SD of seedling characters of Ashak provenance |
| Table 4: | Minimum, maximum, Mean ± SD in characters of seedlings (progenies) Sangdeh provenance |
| Table 5: | Phenotypic correlation coefficients of the recorded characters, between progenies of Ashak provenance |
| Table 6: | Phenotypic correlation of characters estimated between progenies of Sangdeh provenance |
| Table 7: | Phenotypic correlation of characters estimated between seedling (progenies) and provenances |
| **Significant at 1% probability. *Significant at 5% probability. ns: Non-significant | |
In Sangdeh provenance a negatively significant phenotypic correlation was found between vitality and other characters (except angle of the longest branch), between survival and angle of the longest branch and between length of the longest branch and angle of the longest branch. Correlations were insignificant only between collar diameter, angle of the longest branch, length of the longest branch with vitality.
The values show a negatively significant correlation between the altitudes of the provenances and survival, collar diameter, height, branches number, length of longest branch. Provenance was positive and significantly correlated with vitality and angle of the longest branch (Table 7). In fact the seedlings from the lower elevation seed source (Sangdeh provenance), compared to those from higher elevation seed source (Ashak provenance), showed greater values in survival, collar diameter, height, number and length of branch and smaller value in vitality and angle of the longest branch. It can be stated that Sangdeh provenance is more suitable than Ashak provenance for seed collection and also seedling production at the nursery.
DISCUSSION
Results showed that quantitative and qualitative characteristics of the progenies (seedlings) originated from different seed sources are different. Most of the characters recorded on the progenies produced from lower elevation seed source, nearer to nursery location, were higher than those at higher elevation seed source. This finding is completely similar to those of the first and the second years reported by Espahbodi (2005).
Joyce et al. (2002), on Pinus strobus and Schmidtling (1994) showed the temperature of seed source area is the most decisive factor on qualitative and quantitative characters of progenies and in most cases the seedlings produced from warmer provenances grow faster than those from colder provenances. Likewise, Nienstaedt and Olson (1961), as well as Saenz-Romero et al. (2006), on Pinus occarpa), considered height growth is greater in progenies created from longer-growth season (warmer areas) populations than those from shorter-growth season (colder area) populations. These results also are in agreement with Sebbenn et al. (2003) who emphasis that growth performance of South provenances is generally better than other regions. As Schoenike (2002)` finding, in present study a significant correlation between seedlings survival and altitude of seed source (provenance) was found.
It can be emphasized that local provenances have higher priority for seedling plantation than provenances with further distances from planting area. The reasons might be due to the genetic adaptations of the seedlings produced from the local seeds with desired environmental conditions. It is basically accepted that the progenies produced from Southern (warmer) areas have a faster growth than those from Northern (colder) areas.
One of the important reasons could be the longer period of growth season in warmer areas compared with those in colder areas (higher altitudes and or Northern longitudes).
Present investigation showed that Sangdeh provenance, compared to Ashak provenance, is of better conditions for seedling production of wild service tree. On the other hand, it should be declared that Sorbus torminalis is not a species having high genetic variation in the area. It is generally regarded as an extinct risk species (Espahbodi, 2005). Therefore, it is necessary to use the seed sources of this species even those of further distances from the nursery location.
Based on these results there is a significant correlation between the seedling characters such as survival, collar diameter, height, length and number of branches in both of Ashak and Sangdeh provenances. Such results were also observed during the first and second years of the study (Espahbodi, 2005). In fact, increased collar diameter positively affects height growth and branches. Vargaz-Hernantez et al. (2003) conducted a progeny test of Douglas-fir, reported similar results.
Significant correlation between two characters could be due to several reasons. The most important reasons might be linkage between the controlling genes and pleiotropy function of the genes (Makarechian, 2002). In the case of linkage between the controlling genes of two or more characters, a few genes control the desired characters but all of them are located on the same chromosome near each other. So, the closer genes have the higher correlation. If the genes responsible for increasing the characters have a close linkage, the correlation of the characters would be positive. In contrast, if the gene responsible for increasing a character is linked with a decreasing gene, the correlation would be negative. In the case of pleiotropy, a gene controls several characters.
Generally, we concluded that in the selection time of parent trees for seedling production, if the genotypes with greater collar diameter and height with less angle of branches be used as the seed stands the higher survival and growth would be gained. By improving one characters conducted by the tree breeding programs other characters might be improved, provided a close correlation between the characters.