UVOD
The genus Pyrus belongs to the family Rosaceae, subtribe Pyrinae, and contains at least 22 widely recognized primary species, all indigenous to Asia, Europe, and the mountainous area of North America (Bennici et al., 2018). Recently, in the face of global climate changes, more attention has been paid in many European countries to rare woody species that can adapt to dry climatic conditions, and one of these is the wild pear (Drvodelić et al., 2012). In addition, in many European countries the wild pear is in the focus of interest of landscape designers and foresters, as this species has an aesthetic impact on the landscape, a good growth rate and provides valuable wood (Paganová, 2009).
The wild pear is widespread through the Europe. In north Europe it reaches the Baltic sea in Poland, Lithuania and northern Germany, in west Europe it reaches the Atlantic Ocean in France and occupies the southern parts of the United Kingdom, in south Europe it occupies most parts of Italy and Greece, but not the islands (with the exception of Sardinia) and Iberian Peninsula, and to the east partly it occupies Latvia, Belarus, Ukraine, Russia and Caucasus mountains and eastern Black sea shore (EUFORGEN, 2023). It is mostly found on the forest edges, in hedges on farmland or in very extreme locations where stronger competitors do not survive (Kleinschmit et al., 1998). In Slovakia, the wild pear is one of the rare woody plants and is often found in the scattered vegetation of the landscape, but also on forest edges, especially in warm oak-tree communities (Paganová, 2009). In Serbia, only a small population of wild pear has remained (Mikić et al., 2009), which indicates its threatened existence. According to EUFORGEN (2023), the wild pear is extremely light-demanding and has low competitive abilities. It has high morphological heterogeneity, which may be due to the effect of balancing selection caused by a self-incompatibility mechanism (Wolko et al., 2015) or due to the hybridization between P. communis and native P. pyraster, e.g. the gene flow from cultivars (Wagner and Büttner, 2019). Moreover, due to its economic insignificance and inconspicuous appearance, in combination with present-day intensive forest and agricultural management systems, a further decrease of this species is expected (Reim et al., 2017).
Wild pear trees grow slowly, but under very favourable habitat conditions they grow up to 20 m tall and reach a diameter of 0.5 m (Drvodelić et al., 2012). The species’ tolerance to drought can be attributed to its tap roots (Mikić et al., 2009). It has astringent fruit, with many stone cells that make it difficult to separate the seeds (Cerović et al., 2015). Its fruits are rarely suitable for consumption, and are usually dried or processed into brandy (Šebek, 2019) or marmalade and jam (Drvodelić et al., 2012). However, they are excellent food for wild animals, such as deer, roe deer etc. (Drvodelić et al., 2012). Its seeds may be used for generative propagation. It is known that wild pear seedlings are used as rootstock, and even today they are preferred on calcareous, poor sandy soils, where Fe-chlorosis is a serious problem (Cerović et. al., 2015; Šebek 2019, Bennici et al., 2018 according to Continella et al., 2006; Condello et al., 2009; Mišić, 1984). Seedlings can also be used for afforestation since its wood is highly valued on the market, and because it has high ecological importance, being part of natural threatened ecosystems (Camerano et al., 2013; Drvodelić et al., 2012; Kleinschmit et al., 1998). Moreover, it can be used as a very valuable breeding material (Mikić et al., 2009). The seeds of the wild pear are separated from the fleshy aril immediately after the collection in autumn. Sowing takes place in early spring after stratification for three to four months at a temperature of 3-5°C. Due to the extended time of planting in the field, climate changes and minor shocks due to transplanting seedlings, it is best to grow them in multi-containers made of solid plastic.
Due to its heterogeneity, there is great variability in fruit morphology. Therefore, the current study aims to evaluate the effects of fruit size of wild pear (Pyrus pyraster (L.) Burgsd.) on the morphological and physiological traits of its fruit and seed, as well to determine the variability of the studied traits in agroecological conditions of Žumberačko gorje, Croatia. Moreover, this study presents a continuation of two previous, similar studies (Drvodelić et al., 2018, 2015).
MATERIJALI I METODE
Initially, this study was designed to last for three consecutive years, from 2021 to 2023, on 10 wild pear trees (Pyrus pyraster (L.) Burgsd.) in the Žumberačko gorje area (45°50¨33.10˝˝N; 15°35¨08.65˝˝E). However, due to frost damage and the fact that the wild pear is prone to alternative bearing, the fruit was collected only in the fall of 2022. One hundred and fifty fruits were randomly collected from the ground and divided into three groups according to their weight (small (<10 g), medium (10–15 g) and large (>15 g)). Consequently, the small, medium and large fruit consisted of 24, 37 and 89 fruit per sample, respectively.
All analyses were conducted at the Laboratory of Forest Seed and Nursery Production of the Institute of Ecology and Silviculture of the University of Zagreb, Faculty of Forestry and Wood Technology (Figures 1 and 2). Fruit length and width (mm) were measured with a digital scrolling scale (Sylvac Pro) with an accuracy of 0.01 mm. Fruit width was measured at the widest part of the fruit. For each fruit, two opposite lengths and widths were measured, from which the average values were calculated. Fruit weight (g) was measured using a digital analytical balance (Kern Pls, Kern&Sohn GmbH, Germany) with an accuracy of 0.01 g. Then all fruits were submitted to the manual seed extraction process with small knife. At the same time, the number of filled and empty seeds per each fruit was recorded. The weight of the full fresh seed per fruit (g) was recorded on a RADWAG balance (Model AS 220.3Y, S/N: 565960, Max 220 g, Min 10 mg, d=0.1 mg, Bracka, Poland), while the weight of 1 fresh seed was calculated as the ratio between the weight of the full fresh seed per fruit and the number of full seeds per fruit. Afterwards, all seeds were placed on the sterilised surface at room temperature to reduce their moisture content to 8-14%, which took several days. The weight of full air-dried seeds per fruit was then measured using the aforementioned balance and the weight of 1-air dried seed was calculated as the ratio between the weight of full air-dried seed per fruit and the number of full seeds per fruit. The share of seed dry matter was calculated as a difference in % between fresh and air-dried seed mass.
Figure 1. Fruit of the wild pear (left) and transverse and longitudinal section through the fruit (right)
Slika 1. Plodovi divlje kruške (lijevo) i poprečni i uzdužni presjek ploda (desno)
Figure 2. Wild pear seeds
Slika 2. Sjeme divlje kruške
Data was statistically analysed using XLSTAT add-on for Microsoft Office 2016 by ANOVA and Tukey's HSD test (P < 0.05). Pearson’s correlation analysis was also performed in the same software. Prior to principal components analysis (PCA), Kaiser-Meyer-Olkin measure of sampling adequacy was used to examine if it was correct to perform PCA and gave a value of 0.56, which was acceptable (Kaiser, 1974) for continuation to the PCA analysis
REZULTATI
According to the ANOVA, fruit size had a significant effect on the following morphological wild pear traits: fruit weight, fruit length, fruit width, the number of filled seeds, the weight of fresh seeds per fruit, the weight of 1 fresh seed and the weight of air-dried seeds per fruit. However, for fruit shape index, the number of empty seeds, the weight of 1 air dried seed and the share of seed dry matter no significant difference was recorded (Table 1). Large fruit had significantly the highest and small fruit significantly the lowest values of fruit weight, fruit length, fruit width, the number of filled seeds, the weight of fresh seeds per fruit and the weight of air-dried seeds per fruit (Table 1). The weight of 1 fresh seed was significantly higher in large fruit in contrast to that in small and medium fruit, where no significant difference was recorded (Table 1).
Table 1. Morphological and physiological traits of wild pear (Pyrus pyraster (L.) Burgsd.) fruit (mean ± SD)
Tablica 1. Morfološka i fiziološka svojstva ploda divlje kruške (Pyrus pyraster (L.) Burgsd.) (srednja vrijednost ± SD)
Note: Values with the same letter are not significant according to Tukey’s HSD test at P < 0.05 level
Bilješka: vrijednosti obilježene istim slovima nisu značajne prema Tukeyjevom HSD testu pri P < 0,05
In all samples significant and very high positive correlation was recorded between fruit weight and fruit width (r = 0.93). Significant and high positive correlations were recorded between fruit weight and fruit length (r = 0.87) and between fruit length and fruit width (r = 0.87). Significant and moderate positive correlations were recorded between fruit weight and the weight of air-dried seeds per fruit (r = 0.62); between the weight of fresh seeds per fruit (r = 0.61) or the number of filled seeds (r = 0.51); between fruit length and the weight of air-dried seeds per fruit (r = 0.50) and between fruit width and the weight of air-dried seeds per fruit (r = 0.67), the weight of fresh seeds per fruit (r = 0.61) or the number of filled seeds (r = 0.54) (Table 2).
Table 2. Correlation coefficients for the morphological and physiological traits evaluated on all wild pear (Pyrus pyraster (L.) Burgsd.) fruit
Tablica 2. Korelacijski koeficijenti za morfološka i fiziološka svojstva svih plodova divlje kruške (Pyrus pyraster (L.) Burgsd.)
*All bold values present significant correlations at P < 0.05 *Sve podebljane vrijednosti predstavljaju značajnu korelaciju pri P < 0.05
In small wild pear fruit significant and high positive correlations were recorded between the weight of fresh seeds per fruit and the weight of air-dried seeds per fruit (r = 0.88); between the number of empty seeds and the weight of 1 fresh seed (r = 0.83); between fruit weight and fruit width (r = 0.73) and between the weight of 1 fresh seed and the weight of 1 air-dried seed (r = 0.73). Significant, positive and moderate correlations were recorded between fruit weight and fruit length (r = 0.52); between fruit length and fruit shape index (r = 0.53); between the number of filled seeds and the weight of fresh seeds per fruit (r = 0.66) and between the weight of air-dried seeds per fruit (r = 0.66); between the weight of fresh seeds per fruit and the weight of 1 fresh seed (r = 0.55). Significant, negative and high correlation was observed between the number of empty seeds and the share of seed dry matter (r = -0.79). Significant, negative and moderate correlations were observed between the number of filled seeds and the number of empty seeds (r = -0.58) and significant, negative and low correlation was observed between fruit width and fruit shape index (r = -0.49) (Table 3).
Table 3. Correlation coefficients for the morphological and physiological traits evaluated on small wild pear (Pyrus pyraster (L.) Burgsd.) fruit
Tablica 3. Korelacijski koeficijenti za morfološka i fiziološka svojstva malih plodova divlje kruške (Pyrus pyraster (L.) Burgsd.)
*All bold values present significant correlations at P < 0.05
*Sve podebljane vrijednosti predstavljaju značajnu korelaciju pri P < 0.05
In medium wild pear fruit significant and very high positive correlations were recorded between the number of filled seeds and the weight of air-dried seeds per fruit (r = 0.91), between the weight of fresh seeds per fruit and the weight of air-dried seeds per fruit (r = 0.99) and between the weight of 1 fresh seed and the weight of 1 air-dried seed (r = 0.90). Significant and high positive correlations were recorded between the number of filled seeds and the weight of fresh seeds per fruit (r = 0.80) and between fruit weight and fruit length (r = 0.73). Significant, positive and moderate correlations were recorded between fruit length and fruit width (r = 0.50) or fruit shape index (r = 0.58); between fruit width and the weight of 1 air-dried seed (r = 0.57) and between the weight of 1 air-dried seed and the share of seed dry matter (r = 0.51) (Table 4).
Table 4. Correlation coefficients for the morphological and physiological traits evaluated on medium wild pear (Pyrus pyraster (L.) Burgsd.) fruit
Tablica 4. Korelacijski koeficijenti za morfološka i fiziološka svojstva srednje velikih plodova divlje kruške (Pyrus pyraster (L.) Burgsd.)
*All bold values present significant correlations at P < 0.05
*Sve podebljane vrijednosti predstavljaju značajnu korelaciju pri P < 0.05
For large wild pear fruit significant and very high positive correlations were recorded between the weight of fresh seeds per fruit and the weight of air-dried seeds per fruit (r = 0.98) and between the weight of 1 fresh seed and the weight of 1 air dried seed (r = 0.93). Significant and high positive correlations were recorded between fruit weight and fruit width (r = 0.83); between the number of filled seeds and the weight of fresh seeds per fruit (r=0.80) and between the number of filled seeds and the weight of 1 air-dried seed per fruit (r = 0.74). Significant, positive and moderate correlations were recorded between fruit weight and fruit length (r = 0.69); between fruit length and fruit width (r = 0.69) or fruit shape index (r=0.55) (Table 5).
Table 5. Correlation coefficients for the morphological and physiological traits evaluated on large wild pear (Pyrus pyraster (L.) Burgsd.) fruit
Tablica 5. Korelacijski koeficijenti za morfološka i fiziološka svojstva velikih plodova divlje kruške (Pyrus pyraster (L.) Burgsd.)
*All bold values present significant correlations at P < 0.05
*Sve podebljane vrijednosti predstavljaju značajnu korelaciju pri P < 0.05
PCA revealed 3 significant principal components having eigenvalues higher than 1, which explained 75.88% of the total variability. Higher factor loading and squared cosine values of the variables meant that there is a tighter association with the same principal component. Factor 1 (F1) described 44.73% of the total variability and included fruit weight, fruit length, fruit width, the number of filled seeds, the weight of fresh seeds per fruit and the weight of air dried seeds. Factor 2 (F2) described 17.84% of the total variability and included the weight of 1 fresh seed and the weight of 1 air dried seed. Factor 3 (F3) described 13.31% of the total variability and fruit shape index was the only trait associated with this factor (Table 6).
Table 6. PCA factor loadings and squared cosines of the variables for the morphological and physiological traits of wild pear (Pyrus pyraster (L.) Burgsd.) fruit
Table 6. Faktorska opterećenja i kosinus kvadrata u analizi glavnih komponenti za morfološka i fiziološka svojstva plodova divlje kruške (Pyrus pyraster (L.) Burgsd.)
*All bold values present significant correlations at P < 0.05
*Sve podebljane vrijednosti predstavljaju značajnu korelaciju pri P < 0.05
Figure 3 shows that large wild pear fruit differ from small wild pear fruit and partly from medium wild pear fruit by F1 axis. F2 and F3 axes did not produce a clear separation (data not shown for F3 axis).
Figure 3. Canonical plot for a principal component analyses in relation to PC1 and PC2 with respect to wild pear fruit size (Pyrus pyraster (L.) Burgsd.) fruit
Slika 3. Grafički prikaz uzorka u odnosu na PC1 i PC2 s obzirom na veličinu ploda divlje kruške (Pyrus pyraster (L.) Burgsd.). Legenda: mali plodovi (zelena boja), srednje veliki plodovi (tamno roza boja), veliki plodovi (smeđa boja)
RASPRAVA
In the north western part of Bosnia and Herzegovina, Zavišić et al.(2018) reported that the fruit weight of Pyrus communis var. pyraster in 2012 and 2013 was 11.82 and 9.64 g (respectively) for Banja Luka locality, 13.61 and 14.74 g (respectively) for Manjača locality, 11.05 and 11.30 g (respectively) for Potkozarje locality and 11.96 and 13.49 g (respectively) for Kozara locality. In Montenegro, in the area of Polimlje, Šebek (2019) reported that the average three-year Pyrus pyraster fruit weight of 11 studied genotypes ranged from 7.95 to 27.06 g. In this study (Table 1) variations in fruit weight are in accordance with the available literature data.
Paganova (2009) reported that in Slovakia Pyrus pyraster subsp. pyraster and Pyrus pyraster subsp. achras mean fruit length was 24.50 and 28.14 mm (respectively). In the north-western part of Bosnia and Herzegovina, Zavišić et al.(2018) reported that fruit weight of Pyrus communis var. pyraster in 2012 and 2013 was 24.80 and 22.50 mm (respectively) for Banja Luka locality, 25.00 and 25.00 mm (respectively) for Manjača locality, 25.30 and 25.00 mm (respectively) for Potkozarje locality and 24.60 and 25.10 mm (respectively) for Kozara locality. Fruit length per locality ranged from 20.20 to 27.00 mm for Banja Luka locality, 34.70 to 19.00 for Manjača locality, 20.30 to 30.00 for Potkozarje locality, 30.80 to 20.70 for Kozara locality (Zavišić et al., 2018). In Montenegro, in the Polimlje area, Šebek (2019) reported that the average three-year Pyrus pyraster fruit length of 11 studied genotypes varied from 20.12 to 36.21 mm. In this study (Table 1) variations in fruit length are higher than in aforementioned available literature data, although the average values are in agreement with previously reported data.
Paganova (2009) reported that in Slovakia Pyrus pyraster subsp. pyraster and Pyrus pyraster subsp. achras mean fruit maximum diameter was 26.43 and 25.89 mm (respectively). Zavišić et al. (2018) reported that Pyrus communis var. pyraster fruit maximum diameter in 2012 and 2013 was 27.20 and 25.60 mm (respectively) for Banja Luka locality, 28.50 and 29.40 mm (respectively) for Manjača locality, 26.80 and 28.20 mm (respectively) for Potkozarje locality and 26.80 and 25.10 mm (respectively) for Kozara locality. The maximum fruit diameter per locality was between 24.70 and 29.40 mm for Banja Luka locality, between 32.90 and 23.80 mm for Manjača locality, between 24.70 and 28.30 mm for Potkozarje, and between 35.10 and 19.90 mm for Kozara (Zavišić et al., 2018). In Germany, Wagner and Büttner (2019) reported that the mean fruit width of Pyrus pyraster was 24.2 mm. In Montenegro, in the area of Polimlje, Šebek (2019) reported that average three-year Pyrus pyraster fruit width of 11 studied genotypes varied from 12.31 to 30.06 mm. In this study (Table 1), variations in fruit width are in accordance with available literature data. It is also evident that bigger fruit have higher values of fruit width.
In their study, Paganova (2009) reported that in Slovakia Pyrus pyraster subsp. pyraster and Pyrus pyraster subsp. achras fruit length to diameter ratio was 0.93 and 1.09 (respectively). In Germany, Wagner and Büttner (2019) reported that the fruit length to width ratio of Pyrus pyraster was 0.93. Zavišić et al.(2018) reported that Pyrus communis var. pyraster fruit length to diameter ratio in 2012 and 2013 was 0.92 and 0.88 (respectively) for Banja Luka locality, 0.88 and 0.85 (respectively) for Manjača locality, 0.90 and 0.90 (respectively) for Potkozarje locality and 0.93 and 0.90 (respectively) for Kozara locality. Per locality, fruit length to diameter ratio ranged from 0.81 to 1.02 for Banja Luka locality, from 0.75 to 1.13 for Manjača locality, from 0.82 to 1.08 for Potkozarje locality, from 0.84 to 1.18 for Kozara locality (Zavišić et al., 2018). In this study, the average fruit shape index for small, medium and large fruit were 0.86 ± 0.064, 0.83 ± 0.056 and 0.83 ± 0.045 (respectively), while minimum and maximum fruit shape index of all fruit were 0.74 and 1.02 (data not shown). Such results are a little bit lower than in the available literature. When ratio is greater than 1 it means that the fruit is elongated, pear-shaped (Misimović et al., 2012). In this study (Table 1), almost all fruit had values below 1, indicating their squat shape (Brewer et al., 2006). Since there was no significant difference in fruit shape ratio between different fruit sizes, wild pear fruit does not change shape with size.
In Montenegro, in the area of Polimlje, Šebek (2019) reported that the average three-year mass of 100 dry seeds of 11 studied genotypes of Pyrus pyraster ranged from 2.66 g to 3.73 g. In this study (Table 1), variations in air-dried seed mass are higher in our study than in the available literature data, although the average values are in agreement with the previously reported data.
When the correlation coefficients of all wild pear fruits were analysed, fruit weight and width showed a significant moderate and fruit length a significant small positive correlation with the number of filled seeds (Table 2). This indicates that wild pear fruit with higher values for the size traits have a higher number of filled seeds. However, when correlation coefficients with regard to fruit size thresholds were examined, in small wild pear fruit no such significant correlations were observed. Meanwhile in medium and large wild pear fruit there was low, but significant and positive correlation between fruit weight or length and the number of filled seeds (Tables 4 and 5). The aforementioned indicates that bigger fruits are associated with more filled seeds. In addition, in all wild pear fruit, fruit width had higher correlation with the number of filled seeds than fruit length (Table 2). This is even more clearly visible in correlations with regard to fruit size groups, where fruit length did not achieve significant correlation with the number of filled seeds in any fruit size group (Tables 3, 4 and 5). Therefore, it is clear that fruit weight and width have impact on the number of filled seeds. In addition, only small wild pear fruit had significant and high positive correlation between the number of empty seeds and the weight of 1 fresh seed, and negative with the share of seed dry matter. Positive significant correlations in all cases between the number of filled seed and the weight of air-dried seeds per fruit is logical, since a higher number of filled seeds means a higher number of filled seed per fruit and thus a higher seed weight per fruit.
PCA analysis discrimination between large and small, and partly between medium wild pear fruit, by F1 axis, means that traits that mostly influenced F1 (fruit weight, fruit length, fruit width, the number of filled seeds, the weight of fresh seeds per fruit and the weight of air dried seed per fruit) are most important for majority of distinguishing between the aforementioned fruit classes (Figure 3). The correlations between the selected variables and factors are also shown in the biplot (Figure 4). According to the biplot, fruit length, fruit width, fruit weight, the weight of air-dried seeds per fruit and the weight of fresh seeds per fruit are positively correlated. It can be seen that there is a negative relationship between these variables and the number of empty seeds, the total dry matter and the fruit shape index.
Figure 4. Biplot, in relation to PC1 and PC2, for principal component analyses of the morphological and physiological traits evaluated on large, medium and small wild pear (Pyrus pyraster (L.) Burgsd.) fruit Slika 4. Grafički prikaz, u odnosu na PC1 i PC2 analize glavnih komponenti morfoloških i fizioloških svojstava dobivenih na malim, srednje velikim i velikim plodovima divlje kruške (Pyrus pyraster (L.) Burgsd.). Legenda: aktivne varijable (crvena boja), mali plodovi (smeđa boja) srednje veliki plodovi (tamno roza boja), veliki plodovi (zelena boja)
ZAKLJUČCI
According to the results obtained from this study, it can be concluded that for nursery production of wild pear seedlings bigger fruit should be used due to their higher number of filled seeds. In addition, when collecting wild pear fruit for seed extraction, more emphasis must be given to fruit weight and width then length, since they have higher effect on the number of full seeds. In future work focus will be put on seed germination rate and seedling quality from different fruit size groups.