Diversity of Local and Wild Pear Germplasm in Central Regions of Iran

ABSTRACT

This study was conducted to identify the local and wild pear species and select superior genotypes in some central regions of Iran. For this purpose, some morphological and pomological traits of selected genotypes from central regions of Iran were evaluated based on the DUS guideline for three years at five stages. Nine genotypes namely Hosseini (KB2), Hendi (KB3), Torsheh (KB4), local Barzouk (KB5), local late-ripening (SV1) and Armon (SV3), as well as wild species including Amroud (KB1), Kalanjir (SV2) and Khoj (SV4), were distinguished. The weight, length, and diameter of fruit and the width of the leaf showed a higher coefficient of variation than other traits. The range of the fruit length was recorded between 18 and 100 mm. The fruit weight varied between 25 and 325 g and SV2 and SV1 genotypes had the lowest and the highest values, respectively. Fruit TSS was between 7 and %13. KB1 genotype had the highest leaf length and the lowest leaf width. Based on cluster analysis, genotypes were divided into four separate groups. The three genotypes of KB5, SV1, and KB3 that were placed into one group are advisable as the promising genotypes because of their fruit weight, length, diameter and volume as well as optimum edible quality. These genotypes can be cultivated in orchards after examining regional adaptations and likely have genes related to different resistance that can be used for breeding programs.

KEYWORDS:

• Fruit traits
• genotype
• grouping
• morphology
• promising genotype

Introduction

Common pear (Pyrus communis L.) has been scattered in the Caucasus region of Russia, Asia, Malaya, South, and Central Europe (Mozafari, 2009). Jookovskiy (1950) believed that the Caucasus and western parts of Iran are the first centers of pear species diversity. In the pear genus, 22 species have been identified in the world that 12 species of them are in Iran.

Pear production in today’s world depends on a limited number of cultivars, such as ‘Williams’, ‘Anjo’, ‘Beurre Bosc’, ‘Comice’ and ‘Conference’ that all of them were selected in the 18^th and 19^th centuries. There are very few available options to choose the new pear cultivars according to today’s market needs of the world. For this reason, there is an urgent need for the management of germplasm and the identification of protected pear genotypes to achieve a sufficient level of genetic capital in the breeding programs (Wunsch and Hormaza, 2007).

The first program of pear germplasm collection in Iran began in 1954 with the collection and the introduction of various pear cultivars (Maniee, 1994). Sharifani et al. (2005) collected and evaluated wild pear species in Iran. In other research, wild pear species were identified and collected from some parts of West Azarbaijan province. These species showed high genetic diversity and different characteristics in the nursery (Henareh and Hasani, 2013). Khorshidi et al. (2016) compared fruit morphological traits of 23 ‘Dargazi’ genotypes and seven local and imported pear cultivars and showed that ‘Dargazi’ genotypes had significant differences with each other and with other cultivars in all traits.

Some of the morphological traits of pear cultivars were investigated by Tahzibihagh et al. (2012). The results showed that fruit size, diameter, color, firmness, and the ripening time of the fruit were the distinguishing traits in these cultivars. Germplasm collection has been carried out on some fruit trees in Iran such as walnut (Mousavi et al., 2015), jujube (Ghazaeian, 2015; Zeraatger, 2013) and grape (Doulati Baneh et al., 2013; Nejatian, 2007).

Paganova (2003) showed that morphological characteristics are very effective in assessing fruit diversity. In another study, they used the morphological characteristics of leaves and fruits to the identification of wild pear in Slovakia (Paganova, 2009). In another investigation, the pear germplasm bank was studied in northeastern Spain (Pereira-Lorenzo et al., 2012). Collected pears from England showed the highest fruit weight (109.95 g), fruit length (66.24 mm) and fruit width (56.15 mm), and the Peru group showed the lowest fruit weight (62.38 g), fruit length (53.06 mm) and fruit width (47.16 mm) (Pereira-Lorenzo et al., 2012). It was also found that similar to apple (Pereira-Lorenzo et al., 2003) and chestnut (Pereira-Lorenzo et al., 2006) fruit size was an important factor in the classification of pear cultivars.

In recent years, the genetic heritage of the Pyrus has eroded in the world due to the destruction of local cultivars (Olivera et al., 1999). Free entry of foreign varieties, especially Bartlett leads to paying less attention to internal figures so that removing the native varieties and replacing them with foreign varieties or the other fruits, constantly increasing and there is a fear that in the not far future the species of native varieties destroyed and disappear from nature. Isfahan province is one of the areas of pear distribution and diversity in the world. Climate changes, drought and lack of precipitation in recent years have led to the destruction of orchards and genetic resources of pear in this province., so the identification and evaluation of local genotypes and wild species are important. Achievement of promising genotypes with desirable characteristics of the fruit, environmental adaptations, and tolerance to biotic and abiotic stresses are also useful. The aims of this research are the identification and evaluation of pear germplasm in Semirom and Kashan regions in Isfahan province, and their classification according to the vegetative and fruit characteristics to conserve them and use in subsequent programs.

Materials and Methods

Collection Site

In this research, local pear genotypes in some areas of Isfahan province as well as wild pear species in natural habitats were investigated for three years from 2014 to 2016. For this, some areas of the province, including Kashan (Ghahrood and Barzouk) and Semirom (Vanak, Sivar, and Padena) were considered.

Sampling Method

The experiment was conducted to identification, collection, and evaluation of local pear genotypes and wild pear species in Isfahan province. In the first year, 5 to 10 orchards were selected in each site and local cultivars that had a great difference with each other were identified. Generally, more than 300 pear trees were marked in these areas that were in their root. Then the pear trees of these orchards were visited in five stages, including physiological dormancy of the tree, flowering and leaf emergence, active vegetative growth of the tree, ripening, harvesting time of yield and leaf abscission stages. After studying the morphological characteristics, trees with similar characteristics were deleted from the study process during the first year. Afterward, the trees that were completely different from each other were selected, then, the precise location of trees was recorded, and further studies were carried out on these genotypes during 2015 and 2016. For identification of the wild pear species in some areas of Isfahan province, the distribution points of these species were determined (Khatamsaz, 1992; Sabeti, 1995). Then, these wild pear species were identified at the altitudes of Vanak, Padena and Fathabad regions of Semirom as well as Natanz and Borzouk regions of Kashan, and were studied in five stages. The pear in selected regions was grown mainly as a seedling in some orchards. The geographic location of selected genotypes, including longitude, latitude and altitude were recorded with the GPS device (Table 1). Some of the climatic characteristics of the collection site of genotypes are shown in Table 2.

Table 1. Local names, scientific name, code and characteristics of collection site of pear genotypes

Local name	Scientific name	Code	Collection site	Latitude	Longitude	Altitude (m)
Amroud	P. syriaca	KB1	Kashan	33 47ʹ 39.9ʹ’ N	51 13ʹ 44ʹ’ E	2020
Hoseini	P. communis	KB2	Kashan	31 36ʹ 44.3ʹ’ N	51 33ʹ 15.9ʹ’ E	2464
Hendi	P. communis	KB3	Kashan	32 07ʹ 55.4ʹ’ N	51 38ʹ 56.3ʹ’ E	1819
Torsheh	P. communis	KB4	Kashan	32 28ʹ 22.8ʹ’ N	51 41ʹ 53.5ʹ’ E	1599
Barzouk	P. communis	KB5	Kashan	32 28ʹ 22.8ʹ’ N	51 41ʹ 53.5ʹ’ E	1599
Late ripening	P. communis	SV1	Semirom	31 31ʹ 56.6ʹ’ N	51 19ʹ 56.6ʹ’ E	2250
Kalanjir	P. glabra	SV2	Semirom	32 36ʹ 56.6ʹ’ N	51 27ʹ 56.6ʹ’ E	1631
Armoun	P. communis	SV3	Semirom	31 29ʹ 48.6ʹ’ N	51 35ʹ 43.5ʹ’ E	1620
Khoj	P. communis	SV4	Semirom	32 36ʹ 42.3ʹ’ N	51 17ʹ 37.1ʹ’ E	1650

Table 2. Some climatic characteristics of collection site of genotypes

Collection site	The average annual rainfall (mm)	Minimumtemperature (°C)	Averagetemperature (°C)	Maximumtemperature (°C)
Kashan	95	−5	20	46
Semirom	544	−17	12	33

Morphological and Pomological Traits

For nine selected pear genotypes, 28 traits were recorded including morphological and pomological traits. Traits, abbreviations and measurement methods of these traits are listed in Table 3. Qualitative traits (16 traits) were recorded based on the national guideline for distinctness, uniformity, and stability (DUS) in pear. Fruit firmness was measured by a penetrometer (model EFFEGI, Italy, plunger diameter 11.1 mm, depth 7.9 mm), at opposite peeled sides and expressed as kg/cm². Total soluble solid contents (TSS) were determined by extracting and mixing two drops of juice from the two cut ends of each fruit into a digital refractometer (ATAGO N-1α, Japan) at 22°C. The total titratable acidity was determined by titration with sodium hydroxide (0.1 N) and expressed as a percent of malic acid.

Table 3. Abbreviation and measurement method for recorded traits

Trait	Abb.	Unit	Measurement method
Tree vigor	TV	Code	Extremely weak (1), Weak (3), Intermediate (5), Vigorous (7), Extremely vigorous (9)
Tree growth habit	TGH	Code	Upright (1), Semi upright (3), Semi spreading (5), Spreading (7), Drooping (9)
Tree size	TS	Code	Very small (1), Small (3), Medium (5), Large (7), Very large (9)
Foliage Density	FD	Code	Very low (1), Low (3), Intermediate (5), High (7), very high (9)
Tree height	TH	Code	Very short (1), Short (3), Medium (5), Long (7), Very long (9)
Bearing Habit	BH	Code	Spur (1), Annual shoot (3), Annual shoot and spur (5)
Biennial Bearing	BB	Code	Without (1), Very low (3), Low (5), High (7), Very high (9)
Suckering Tendency	ST	Code	Without sucker (1), Very low (3), Low (5), High (7), Very high (9)
Peduncle Length	PL	mm	Ruler (average of 10 peduncles)
Diameter of Flower	DF	mm	Caliper (average of 10 flowers)
Flowering time	FT	Code	Very early (1), Early (3), Intermediate (5), Late (7), Very late (9)
Flower density	FD	Code	Very low (1), Low (3), Medium (5), High (7), Very high (9)
Leaf length	LL	mm	Ruler (average of 10 leaves)
Leaf width	LW	mm	Ruler (average of 10 leaves)
Leaf shape	LS	Code	Narrow (1), Elliptical (3), Wide oval (5), Oval (7), Oval-stretch (9)
Petiole length	PTL	mm	Ruler (average of 10 petioles)
Leaf Density	LD	Code	Very low (1), Low (3), Medium (5), High (7), Very high (9)
Fruit Ripening Time	FRT	Code	Very soon (1), Soon (2), Middle (3), Late (4), Very late (5)
Fruit Size	FS	Code	Very small (1), Small (3), Medium (5), Large (7), Very large (9)
Fruit weight	FW	gr	Digital scale (average of 10 fruits)
Fruit length	FL	mm	Caliper (average of 10 fruits)
Fruit diameter	FDI	mm	Caliper (average of 10 fruits)
Fruit Volume	FV	mm	Graduated cylinder
Color of Fruit	CF	Code	Light yellow (1), Dark yellow (3), Bright green (5), Dark green (7), Yellowish green (9)
Fruit Firmness	FF	Kg/cm^2	Pentameter (average of 10 fruits)
Total Soluble Solid	TSS	%	Refractometer (average of 10 fruits)
Titrable Acidity	TA	%	Titration
Edible Quality	EQ	Code	Very weak (1), Weak (3), Medium (5), Good (7), Very good (9)

The scion was prepared from identified promising genotypes and was grafted on pear rootstocks at the nursery for future research.

Data Analysis

Analysis of variance and comparison of means for quantitative traits was performed using SAS (version 9.1). Descriptive statistics, simple correlation of traits (Pearson method) and cluster analysis and grouping of genotypes (Ward method and based on squared Euclidean distance) were carried out using SPSS (version 15).

Results

The Analysis of Variance

According to the results, six pear genotypes, including Hosseini (KB2), Hendi (KB3), Torshe (KB4), local Barzouk (KB5), local late-ripening (SV1) and Armon (SV3), as well as three wild pear species, including Amroud (KB1), Kalanjir or Anchuchak (SV2) and Khoj (SV4) were identified and collected from different regions of Kashan and Semirom (Table 1).

The mean of traits in different genotypes, as well as the coefficient of variation in each trait, is presented in Table 4. The coefficient of variation of the traits varied between 16.9 and 79.04. The diversity coefficient for fruit weight, length, and diameter as well as leaf width traits were more than 40%.

Table 4. Mean, Range and coefficient of variation in the studied traits

Characteristics	Coefficient of variation (%)	Standard deviation	Minimum	Mean	Maximum
Peduncle Length (PL)	24.5	5.58	20	22.77	37
Diameter of Flower (DF)	17.64	5	20	28.33	35
Leaf length (LL)	16.9	8.9	42	52.66	70
Leaf width (LW)	42.86	14.62	18	34.11	69
Petiole Length (PTL)	26.95	9.67	26	35.88	53
Fruit weight (FW)	79.04	103.99	25	131.55	325
Fruit length (FL)	41.85	22.88	18	54.66	95
Fruit diameter (FDI)	41.8	21.04	13	50.33	81
Fruit Volume (FV)	39.57	66.22	40	167.33	260
Fruit Firmness (FF)	32.92	1.86	3.4	5.65	8
Total Soluble Solid (TSS)	19.11	2.08	8	10.88	13
Titrable Acidity (TA)	17.05	0.73	3.2	4.28	5.5

The analysis of variance showed that the genotypes were significantly different in more the studied traits (Table 5). The range of leaf length was recorded between 65 and 42 mm for KB1 and SV3 genotypes, respectively. The leaf width varied between 41 and 17 mm for KB2 and KB1 genotypes, respectively. In general, Amroud wild pear (KB1) has the highest leaf length and the lowest leaf width (Table 5).

Table 5. Mean of quantitative registered traits for nine genotypes of pear

Pear genotype	PL	DF	LL	LW	PTL	FW	FL	FDI	FV	FF	TSS	TA
KB1	22.33bc	40a	65a	17 c	50a	62d	55 c	36d	96e	5bc	9b	3.2a
KB2	22.33bc	40a	49abcd	41a	25d	74 cd	55 c	56b	172d	3 c	13a	4.6a
KB3	20.66 c	37.33a	44 cd	31abc	35b	226b	100a	81a	210 c	4 c	12a	3.7a
KB4	21.66bc	25bbc	46bcd	31abc	34bc	70 cd	55 c	47bc	165d	3 c	12a	3.5a
KB5	22bc	35ab	59ab	35ab	53a	240b	58 c	51bc	225b	5bc	13a	4.1a
SV1	22bc	30abc	58ab	29abc	37b	325a	80b	75a	260a	5bc	12.5a	4.5a
SV2	21 c	20 c	56abc	28abc	26d	25e	18e	13e	40 f	7.66a	7 c	4a
SV3	37a	30abc	42d	23bc	33bc	78 cd	36d	42 cd	166d	7a	9b	5a
SV4	25b	30abc	50abcd	30abc	29 cd	84 c	40d	42 cd	172d	6.66ab	8.5bc	4.5a

The highest (53 mm) and lowest (25 mm) petiole length were belonged to KB5 and KB2 genotypes, respectively. The highest diameter of the flower belonged to KB1 and KB2 (40 mm) and the lowest of it was related to SV2 genotype (20 mm). The peduncle length varied from 37 mm in SV3 genotype and 20.66 mm in KB3 genotype. The highest length of fruit belonged to KB3 (100 mm) and the highest fruit diameter was related to SV1 genotype (75 mm). The SV2 genotype showed the lowest fruit length and diameter, with averages of 18 and 13 mm, respectively (Table 5).

The average fruit weight was varied from 325 g in SV1 and 25 g in SV2. These two genotypes also had the highest and the lowest fruit volumes with averages of 260 and 40 ml, respectively. The SV2 genotype showed the most firmness and KB2 and KB4 genotypes showed the least firmness with mean values of 7.66 and 3 kg/cm², respectively (Table 5).

The TSS was varied between 13% in KB2 and KB5 genotypes and 7% in the SV2 genotype. No significant difference was found between genotypes in the titrable acid.

The SV2 genotype had weak growth and the other genotypes were intermediate or vigorous. The growth habit of all the examined pear genotypes was upright and semi-upright. The size of the trees was large and so large. Foliage density was lower in SV2 genotype and intermediate or high in other genotypes. The height of the trees was also medium and long. Bearing habit in KB2 and KB3 genotypes were on the spur and in other genotypes were on both spur and annual shoot. KB1 genotype was without biennial bearing and other genotypes had a low biennial bearing. The KB1 genotype was without sucker, while the SV3 and SV4 genotypes had a lot of suckers and other genotypes produced a few suckers (Table 6).

Table 6. Qualitative registered traits for nine genotypes of pear

Pear genotype	TV	TGH	TS	FD	TH	BH	BB	ST
KB1	Vigorous	Upright	Large	High	Long	Annual shoot and spur	Without	Without sucker
KB2	Intermediate	Semi upright	Medium	Low	Medium	Spur	Very low	Very low
KB3	Intermediate	Semi spreading	Medium	Intermediate	Medium	Spur	Very low	Very low
KB4	Intermediate	Semi upright	Large	Intermediate	Long	Annual shoot and spur	Very low	Very low
KB5	Vigorous	Upright	Large	High	Long	Annual shoot and spur	Very low	Very low
SV1	Intermediate	Semi upright	Medium	Intermediate	Medium	Annual shoot and spur	Very low	Very low
SV2	Weak	Semi upright	Large	Intermediate	Long	Annual shoot and spur	Very low	Very low
SV3	Vigorous	Upright	Large	High	Long	Annual shoot and spur	Very low	High
SV4	Vigorous	Upright	Large	High	Long	Annual shoot and spur	Very low	High
Pear genotype	FT	FD	LS	LD	FRT	FS	CF	EQ
KB1	Early	High	Oval	High	Late	Small	Bright green	Weak
KB2	Intermediate	Medium	Oval	Medium	Soon	Medium	Dark yellow	Very good
KB3	Intermediate	High	Oval	Medium	Late	Large	Dark yellow	Good
KB4	Early	High	Oval	Medium	Late	Medium	Dark yellow	Very good
KB5	Intermediate	High	Oval-stretch	High	Late	Large	Dark yellow	Good
SV1	Intermediate	High	Oval-stretch	High	Very late	Very large	Bright green	Good
SV2	Intermediate	Very high	Narrow	Medium	Middle	Very small	Dark green	Very weak
SV3	Intermediate	Very high	Oval	Medium	Very late	Medium	Bright green	Weak
SV4	Intermediate	Very high	Wide oval	Medium	Late	Small	Bright green	Weak

The KB1 and KB4 genotypes were early flowerings and other genotypes were intermediate. Flower density was very high in SV2, SV3, and SV4 genotypes and others, it was medium and high. The time of fruit ripening in KB2 genotype was soon and in other genotypes was middle or very late.

The fruit size was very small in SV2, very large in SV1 and the medium in other genotypes. The fruit edible quality in SV2, KB1, SV3, and SV4 genotypes were very weak and weak and for other genotypes was good and very good (Table 6).

Correlation of Traits

There was a significant correlation between some of the measured traits (Table 7). According to the results, the tree growth vigor had a significant positive correlation with foliage density (r = 0.75). There was a significant negative correlation between fruit density with fruit edible quality (r = −0.816) and TSS (r = −0.833), and a positive correlation with fruit firmness (r = 0.922). Fruit size had a significant positive relationship with fruit edible quality (r = 0.689), fruit weight (r = 0.903), length (r = 0.812), diameter (r = 0.863), volume (r = 0.935) and TSS (r = 0.806). The fruit edible quality was positive relationship with fruit diameter (r = 0.747), fruit firmness (r = 0.881) and TSS (r = 0.938). No correlation was observed between leaf length and width. Fruit weight, length, diameter and volume had a positive and significant relationship with each other. Fruit firmness and TSS (r = −0.447) showed a significant negative correlation.

Table 7. Results of simple correlation analysis for some studied characteristics

	TV	BD	FT	FD	FS	EQ	FW	FL	FDI	FV	FF	TSS	TA
TV	1
BD	0.75*	1
FT	−0.134	−0.134	1
FD	0.088	0.619	0.189	1
FS	0.144	−0.144	0.231	−0.459	1
EQ	−0.08	−0.557	−0.106	−0.816**	0.689*	1
FW	0.075	−0.008	0.357	−0.271	0.903**	0.44	1
FL	0.054	−0.232	−0.008	−0.576	0.812**	0.652	0.768*	1
FDI	0.059	−0.386	0.238	−0.621	0.863**	0.747*	0.734*	0.911**	1
FV	0.312	−0.04	0.315	−0.364	0.935**	0.663	0.843**	0.731**	0.849**	1
FF	0.098	0.523	0.442	0.922**	−0.491	−0.881**	−0.282	−0.617	−0.573	−0.379	1
TSS	−0.056	−0.48	0.038	−0.833**	0.806**	0.938**	0.646	0.758*	0.806**	0.728*	−0.847**	1
TA	0.376	−0.16	0.722*	0.463	−0.263	−0.432	−0.173	−0.595	−0.344	−0.23	0.653	−0.411	1

Cluster Analysis

According to the results of cluster analysis, genotypes were classified into four groups at five squared Euclidean distance basis of the Ward method (Figure 1). In the first group, KB2, KB4, and SV3 genotypes were placed along with with the SV4 (Khoj) species. The KB5, KB3 and SV1 genotypes in the second group, KB1 wild pear species (Amroud) in the third group and SV2 wild pear species (Kalanjir or Anchuchek) in the fourth group were placed.

Figure 1. Grouping of pear genotypes based on measured characteristics by Ward method

Discussion

Investigating of the range and the coefficient of variation for each trait in selected pear genotypes showed that traits with a high coefficient of variation, have a wider range of trait quantity that provides a more choice range for that trait.

Fruit weight, length, diameter, and leaf width had a higher coefficient of variation (Table 4). The high coefficient of variation in fruit traits was also reported by Erfani et al. (2014). They evaluated pear genotypes and wild species in Europe and Asian. In their research, the coefficient of variation of fruit diameter was very close to the coefficient of variation of this trait in the present study (46.74%). Similarly, Petruccelli et al. (2013) reported that the coefficient of variation of flower and fruit traits were more than the coefficient of variation of other traits. In the research done by Petruccelli et al. (2013) the coefficient of variation of pear fruit volume was 28.5%, which was less than the measured values in the present study (39.57%). High diversity in the characteristics of wild-type fruit in temperate regions has been reported by many researchers (Ahmed et al., 2011; Tahzibihagh et al., 2012; Zaffar et al., 2004). More genetic diversity helps the plant to overcome environmental changes and also provides more chances to select new cultivars (Liu, 2006).

The coefficient of variation is not related to the trait measurement unit and has an effective role in the comparison of the studied traits. The value of the coefficient of variation can indicate the ability of a trait for distinguishing among genotypes. Morphological traits with a higher coefficient of variations are more reliable than other indicators for identifying cultivars and genotypes (Petruccelli et al., 2013). Fruit characteristics such as size, diameter, and skin color were also used to distinguish among sweet and sour cherry cultivars (Perez-Sanchez et al., 2010).

It was reported that leaf traits had the highest diversity among the grape studied cultivars (Mousazadeh et al., 2014). Doulati Baneh et al. (2013) and Tahzibihagh et al. (2012) also observed a high variation in the leaf characteristics of grape and pear cultivars. In the current study, the diversity coefficient of leaf width was high compared to other traits (Table 4).

The range of leaf length and width was between 65–42 and 41–17 mm, respectively. Paganova (2009) reported the leaf length and width of wild pear species in Slovakia were from 37.21 to 64 and 32.65 to 38.33, respectively. The range of leaf length and width in the examined pears by Paganova (2009) was lower than the reported values in the present study. Genetic and environmental factors are responsible for leaf size variation (Ahmed et al., 2013). The petiole length was varied between 25 and 53 mm (Table 4). Mozafari (2009) reported the petiole length of the studied pear genotypes in Kurdistan was between 45 and 74.2 mm, which had a wider range than petiole length of genotypes in Isfahan province.

In the present study, the range of flower diameter and peduncle length was from 20 to 41 and 20.66 to 37 mm, respectively. In a study, flower diameter of ‘Sebri’, ‘Shahmiveh’, ‘Domkaj’, ‘Sardroudi’, ‘Shahak’, ‘Natanzi’, ‘Dargazi’, ‘Tashkandi’ and ‘SeifeTabriz’ cultivars were ranked by means of 44, 39.5, 39, 35, 34.6, 34, 33.2, 32, 25.5 mm, respectively (Tahzibihagh et al., 2012), which were almost corresponding to the flower diameter in the evaluated pears in this study. The range of peduncle length in this study was more than the identified pears by Mozafari (2009) in Kurdistan (31.4–36.4 mm).

In our study, the fruit length and diameter were varied from 18–100 and 13–75 mm, respectively. According to Bell et al. (1996), the pear fruit diameter varies between one and more than 120 mm in different cultivars. The range of fruit length and diameter in this study was more than the values reported by Paganova (2009) and Dehghani et al. (2013). Paganova (2009) reported the range of fruit length and diameter in Slovak pears from 23.08 to 28.37 and 25.89 to 29.1, respectively, and Dehghani et al. (2013) reported these values in Asian pears from 49.75 to 77.51 and 43.25 to 67.97, respectively.

Fruit weight and volume were varied between 25–325 g and 42–260 ml, respectively, that higher values of these two traits were related to SV1 genotype and the lower values were belonged to SV2 genotype. Similarly, in a study by Dehghani et al. (2013), KS8 and KS6 cultivars had the highest and lowest of volumes and weight of fruit, respectively. The fruit weight in a conducted research by Tahzibihagh et al. (2012) was between 45.26 (in cv. ‘Domkaj’) and 383.5 g (in cv. ‘Mohammadali’). The values of this trait in other research were between 66.21 g (in KS6 genotype) and 206.2 (in KS8 genotype) (Dehghani et al., 2013). In both of these studies, the minimum of fruit weight was higher than the minimum of fruit weight in the present study, while in the other research, the minimum and maximum fruit weight were recorded between 21.34 and 214 g in the wild and European species, respectively (Erfani et al., 2014) that were lower than values in the current study.

The SV2 genotype showed the most fruit firmness and KB2 and KB4 genotypes showed the least of this trait with means of 7.66 and 3 kg/cm², respectively. In the research that was carried out on native cultivars, ‘Domkaj’ and ‘Shahak’ cultivars had the lowest firmness and ‘Sardroudi’ and ‘Sebri’ cultivars had the highest fruit firmness (Tahzibihagh et al., 2012).

TSS varied between 7-13%. A significant difference in the TSS of pear cultivars was also proposed by Erfani et al. (2014). In research carried out by Tahzibihagh et al. (2012), TSS was between 14.04 and 17.4% that had a wider range than the values of the present study. It has been reported that the diversity of pear cultivars is based on differences in fruit composition, such as TSS. The fruit tissue traits have lower heritabilities and are affected by environmental factors (Chen et al., 2007). In our study, there was no significant difference among genotypes for TA trait, which was inconsistent with Tahzibihagh et al. (2012) that reported TA was between 7.4 and 2.84%.

Tree vigor in the most genotypes was intermediate and vigorous. Examined pears by Tahzibihagh et al. (2012) were also intermediate and vigorous. Foliage density was intermediate and high in most genotypes. Foliage density was also high in ‘Tashkandi’ and ‘Ghosi’ cultivars and was low in ‘Sefasleh’, ‘Seife-Tabriz’ and ‘Sebri’ cultivars (Tahzibihagh et al., 2012). The fruit size was varied between very small and very large. In another study, ‘Tashkandi’ and ‘Mohammad Ali’ cultivars had the largest fruit size and ‘Domkaj’ and ‘Seife Tabriz’ cultivars had the least rates of this trait (Tahzibihagh et al., 2012).

Doulati-Baneh et al. (2013) were reported some morphological differences among the wild grapes in West Azerbaijan. In the present study, three species of wild pears were significantly different in leaf length as well as length, weight, and volume of fruit. In general, although inter and intraspecies diversity is not reliable only based on morphological traits, some of the traits in this research, including fruit weight and size, as well as fruit and leaf dimensions, can be used for identification of pear genotypes.

A high correlation can be used to estimate other traits. The tree growth vigor had a significant positive correlation with foliage density. Similarly, it was found that the tree growth vigor was highly correlated with the foliage density in European pear seedlings (Rahmati et al., 2015). The correlation was not observed among leaf length and width with other traits, while the results of other research showed that leaf length and width had a positive and significant correlation with fruit weight and size traits (Erfani et al., 2014). It was also reported that late-ripening cultivars had more TSS than early ripening cultivars and had a more fruit firmness (Mozafari, 2009) but this relationship was not seen in the present study.

Fruit weight, length, diameter and volume had a positive and significant correlation with each other. Dehghani et al. (2013) also expressed a strong relationship among the fruit diameter, length, volume and fresh weight in Asian pear cultivars. A positive correlation has also been reported between fruit size and weight by Petruccelli et al. (2013). In general, if there is a positive correlation between two traits in a plant, a breeding program is easier for that plant (Mousazadeh et al., 2014).

Cluster analysis classified genotypes into four groups at five squared Euclidean distance. In the first group, KB2, KB4, and SV3 genotypes were placed along with with the SV4 (Khoj) species. High peduncle length, fairly large flower diameter, low petiole length, relatively low fruit weight, high fruit firmness, relatively high TSS, intermediate and vigorous growth, low alternative bearing, low to the high suckering tendency, medium leaf density, and small or medium fruit size were the common characteristics of these genotypes. SV4 (Khoj) species that was located in this group also belongs to Pyrus communis.

In a research that was carried out to determine the genetic diversity of pear cultivars using microsatellite markers, ‘Shahmiveh’, ‘Domkaj’ and ‘Sebri’ cultivars were placed in Khoj group that showed these cultivars were likely derived from Khoj and had no relationship with Asian pears (Safarpour Shourbakhloo et al., 2009). In this regard, Sabeti (1995) notes that the origin of most Iranian native pears is Khoj.

In the second group, KB3, KB5, and SV1 genotypes were placed. The genotypes of this group had a high flower diameter, fairly long petiole length, intermediate to vigorous growth, low alternative bearing and suckering tendency, high flower density, large and very large fruit size, intermediate flowering time and good edible quality. The highest fruit weight, length, diameter and volume, and most TSS along with the appropriate fruit firmness were the main characteristics of this group.

In the third group, only the KB1 wild pear (Amroud) was placed. This genotype had the longest leaf and peduncle length and the smallest leaf width. Other characteristics of this genotype were low weight, diameter, and volume of fruit, small fruit size, without alternative bearing and suckering tendency, early flowering time and weak edible quality.

The SV2 wild pear (Kalanjir or Anchuchak) was placed alone in the fourth group. This genotype was distinctly different from the other genotypes. The lowest peduncle and petiole length, flower diameter, fruit weight, length, diameter, volume and TSS and the most fruit firmness were the characteristics that belonged to this group. Other characteristics of this genotype were weak vigor of the tree, narrow leaf shape, very small fruit, middle ripening fruit time and very weak edible quality.

Grouping of genotypes showed that there was no correlation between grouping and the geographical location of genotypes. This result was also reported by Erfani et al. (2014). The existence of genotypes belonging to different regions in common groups can be attributed to their common source or change morphological characteristics of seedlings after transplantation and cultivation in new conditions (Khakdaman et al., 2007). Similarly, in another study, ‘Dargazy’ and ‘Tashkandi’ cultivars with Khorasan origin and ‘Sebri’ and ‘Natanzi’ cultivars with Isfahan origin were classified in the same group (Tahzibihagh et al., 2012). In some studies, the genotypes of each region are located in similar groups, for example, pear genotypes of two populations of Karaj and Mashhad were placed in separate groups (Rahmati et al., 2015). This relationship between the geographical locations with grouping was found by Santos et al. (2011). It seems that the origin of seedlings has a specific effect on groupings.

Diversity allows breeders to be able to select the best genotypes that have been adapted over the years in a region. Identification of vegetative characteristics, as well as qualitative and quantitative characteristics of the fruit, provides the possibility of selection and improvement of genotypes. Identification of genotypes also helps to protect them. The results of this research showed that in the studied areas in Isfahan province, there are many cultivars as a single tree, and the trees of these cultivars have been able to survive under drought conditions and the other destructive effects. According to the grouping, different species of pear were classified into different groups. The highest edible quality and the desirable characteristics of the fruit were observed in the European species included Brazouk and Hendi local pears from Kashan and late-ripening pears from Semirom. These genotypes will be compared to local check and can be cultivated in orchards. Wild pears were placed in distinctive groups. These species have been exposed to severe climate changes, various diseases and stresses over the years and have been able to adapt to these conditions. These genotypes likely have genes related to different resistance so will be protected for use in breeding programs.