Application of Hybridization Breeding Technique for Fire Blight Resistance on Cydonia Oblonga: A Base Study on Susceptibility, Heterosis, and Heterobeltiosis Parameters

ABSTRACT

Erwinia amylovora: Burrill. is the most important disease affecting quince cultivation and breeding of resistant varieties is seen as the key solution. As a result of hybridizations, 223 hybrids were obtained from 9 different combinations and susceptibility indexes (SI) of these hybrids were determined by artificial inoculations. Among the tested seven parental cultivars, Ege 25, Altın 35, Zeybek 35, and Ayva B-35 were found fire blight resistant while others were detected as moderately susceptible. SI of hybrids were changed between 14.77 and 52.94% regardless of hybrid combinations and 77.60% of obtained hybrids took place in resistant (10.1–30%) group. Reciprocal combinations showed that using resistant cultivar as a male parent was more effective than using maternal parent for transferring resistance. Heterosis (Ht) rates between combinations were found significant (p < .0004), while heterobeltiosis (Hbt) were not (p < .0623). Ht and Hbt ranged between −2.82–13.54 and −4.20–6.72 in populations, respectively. The maximum and significant heterotic effect was observed in Quince A × Ayva B-35 and its’ reciprocal combination. In moderately susceptible × resistance combination, 83.33% of F₁ were determined to be resistant. On the other hand, hybrids from resistance × resistance combination were distributed in three groups as resistant, moderately susceptible and susceptible. According to these results, it could be said that (1) fire blight resistance genes in quince are recessive, (2) use of resistant varieties as pollinators is more successful in transferring resistance to hybrids, (3) open pollination treatment is an essential and easy way in terms of obtaining resistant hybrids if the orchard has resistant pollinator, and (4) heterobeltiosis for fire blight resistance is important.

KEYWORDS:

• Cydonia oblonga M.
• crossbreeding
• open-pollination
• self-pollination
• resistance breeding
• inheritance

Introduction

The genus Cydonia is monospecific, and has single species quince “Cydonia oblonga” (Bell and Leitao, 2011). It ranks 3^rd place following apple and pear in respect of economic importance in pome fruits, and its production has increased 41.3% between 2007 and 2017 worldwide (FAO, 2019). To optimize quince production, great importance should be attached to fighting against the most devastating disease, the fire blight.

Fire blight disease caused by Erwinia amylovora (Burril.) is the most perilous disease, seen in Rosaceae family; including apple, pear, quince. Since struggle methods are inadequate and most cultivated varieties and rootstocks are susceptible to fire blight, cultivation is threatened and high priority is given to the improvement of resistant genetic material. Resistance breeding for fire blight disease on apple and pear, that has higher economic importance, has continued for over 100 years all around the World (Durel et al., 2004; Gardner et al., 1980; Hedrick et al., 1921; Hunter, 1993; Layne and Quamme, 1975; Ryugo, 1982; Van der Zwet et al., 1974) including Turkey (Evrenosoğlu et al., 2010; Mertoğlu and Evrenosoğlu, 2017; Öztürk et al., 2011). Additionally, resistance levels of genetic resources of pear, apple, other Malus and Pyrus species have been previously reviewed in several researches, in order to determine for their potential to be used as a parent in breeding studies (Cameron et al., 1968; Aldwinckle, 1974; Van der Zwet et al., 1974; Gardner et al., 1980; Westwood, 1982; Le Lezec et al., 1987; Lespinasse and Paulin, 1990; Momol et al., 1992; Demir and Gündoğdu, 1993; Aysan et al., 1994; Keck et al., 1996; Saygılı et al., 1999; Norelli et al., 2001; Bell et al., 2005; Fischer et al., 2004; Korba et al., 2008; Ozrenk et al., 2011; Bell and Leitao, 2011; Przybyla et al., 2012; Tóth et al., 2013; Sobiczewski et al., 2015; Kostick et al., 2019; Mertoglu and Evrenosoglu, 2019).

Heredity of resistance to fire blight differs in pome fruit species such as Pyrus communis shows polygenic inheritance (Bell et al., 2005; Durel et al., 2004; Evrenosoğlu et al., 2019; Van der Zwet et al., 1974) while monogenic resistance is observed in P. pyrifolia, P. ussuriensis and some Malus species (Bokszczanin et al., 2012; Gardner et al., 1980; Thompson et al., 1962). Resistant individuals obtained from crossing susceptible × susceptible parents confirm the hypothesis that the fire blight resistance genes in pears are recessive (Evrenosoğlu et al., 2019; Rosati et al., 2002). However, there is no study relating to the inheritance of fire blight resistance in quince.

Although quince is known to be one of the most susceptible fruit species to the disease (Bobev et al., 2009; Postman, 2008), it has been reported that there are tolerant genotypes, cultivars and rootstocks (Abdollahi et al., 2008; Bobev and Deckers, 1999; Mehrabipour et al., 2012; Ozrenk et al., 2012; Papachatzis et al., 2011; Şahin, 2017; Şahin et al., 2019, 2020a). In general, susceptibility indexes of genotypes were determined under natural infection conditions or artificial inoculations, but resistant or very low susceptible genotypes (SI 0–10%) have not yet been found. As a result of artificial inoculations, the mean susceptibility ratios of quince genotypes, cultivars and rootstocks were found between 13.1–90%, 28.58–50%, 29.11–37.16%, and 28.58–51.69%, respectively (Abdollahi et al., 2008; Mehrabipour et al., 2012; Ozrenk et al., 2012; Şahin et al., 2020a).

Only two researches focused on the selection breeding of quince to fire blight resistance in the world (Turkey and Bulgaria). In these researches, resistance was evaluated under natural epidemic conditions. Firstly, fifteen genotypes were determined as resistant via selection breeding in Turkey (Şahin, 2017). On the other hand, the disease severity was examined based on varieties and it was seen that this ratio was lower in Ege 22 variety than Eşme, Limon and Ekmek (Şahin et al., 2019). Secondly, Bobev et al. (2011), observed 274 hybrids for resistance and 19 of them were selected as resistant in Bulgaria. Determination of susceptibility levels under natural conditions requires a long period, thus testing them via artificial inoculation is very important issue for the first stage endurance control.

Heterosis refers to the superior performance of F₁-hybrid compared to the average of parents and heterobeltiosis refers to hybrids which are better than the best parent for investigated properties. These terms, widely use in field crops and vegetable breeding, are also used in fruit trees and especially important in determining transfer of quantitative characteristics and disease resistance (Cardoso et al., 2014; Marin et al., 2006; Vivas et al., 2014). Studies on this subject are concentrated on tropical fruits like papaya (Vale et al., 2016). In pome fruit species, heterosis and heterobeltiosis are only determined on loquat hybrids (Liu et al., 2019).

The aims of the current study were (1) to obtain hybrids from controlled hybridization, open and self-pollination with using moderately susceptible and resistance parents, (2) to determine their susceptibility levels to disease with artificial inoculations in two different periods, and (3) to identify inheritance of disease on quince for the first time in the world. Besides, the number of individuals showing heterosis, heterobeltiosis, and their rates were determined in order to give ideas for further hybridization studies.

Material and Methods

Plant Material and Hybridizations

Ege 2, Ege 22, Ege 25, Altın 35, Zeybek 35 varieties, Quince A, Ayva B-35 rootstocks, and genotype Kdz 12 were used as parent. Parents except genotype Kdz 12, selected as tolerant under natural epidemic conditions, were tested against the disease. Hybridization combinations and the number of tested plants were given in Table 1. In total, 223 F₁ quince hybrids were obtained from crossbreeding, open and self-pollination studies carried out in Aegean Agricultural Research Institute (AARI) quince field gene bank in 2014 as shown Figure 1. Crossbreeding and open-pollinations were made to obtain resistant rootstocks (except Ege 2 × Ege 25). In the scope of these studies, reciprocal crosses were performed only between Quince A and Ayva B-35 rootstocks. With the aim of the cultivar breeding just self-pollinations were performed.

Table 1. Combinations, codes, and number of plants tested

Combinations (♀×♂)	Combination code	Number of plants tested
Quince A× Ayva B-35	AxB-35	5
Ayva B-35× Quince A	B-35xA	106
Zeybek 35× Ayva B-35	ZxB-35	14
Ege 2× Ege 25	Ege2x25	6
Altın 35× Altın 35 (self pol.)	Altın35SP	10
Ege 22× Ege 22 (self pol.)	Ege22SP	10
Ege 25× Ege 25 (self pol.)	Ege25SP	26
Kdz 12 – Open pollination	K12OP	12
Ayva B-35 – Open pollination	B-35OP	34
TOTAL		223

Figure 1. Process of crossbreeding, open, and self-pollination studies (a: whole tree isolation and pollination, b: flowers isolation, c: emasculation of flowers, d: controlled hybridization of emasculated flowers, e: first stages of hybrid fruits, f: ripe fruits, g: seeds kept in moist perlite, h: seedling from viols, i: transferred seedlings to 4 liter pots)

Flower isolation and emasculation were made in controlled hybridizations. Pollination was carried out with a brush and the pollens collected from the anthers dehisced under artificial light the day before. Whole tree isolation was performed for self-pollination and flowers pollinated with own pollens via brush by entering into the isolation cabinet every two days. Isolation was not applied to open-pollination. Seeds were obtained from mature fruit, dried in shade and kept at +4°C. Before planting, seeds were kept in moist perlite for 1 month at +4°C and planted in viols in March 2015 and seedlings were transferred to 4-liter pots in 2016 (Figure 1).

Fire Blight Susceptibility Evaluations of Parents and Hybrids

The susceptibility indexes of hybrid plants and parental cultivars were defined by artificial inoculations. Inoculation was carried out twice on the same material in a year. At the end of the first inoculation, plants were cut off over five buds and allowed to regrow. Mixture of three virulent Erwinia amylovora strains; Ea-163, Ea-211, and Ea-223, known to have high virulence with 100% disease severity (Şahin et al., 2020a) were used in artificial inoculations. Strains were developed in different petri dishes and inoculum prepared with an equal mixture (10⁸ cfu ml⁻¹) of 24 hours old culture cultivated on nutrient sucrose agar. The greenhouse-resistance assessments were performed in a fully automatic climate controlled greenhouse with 24–27°C temperatures, 70–75% relative humidity. When shoots were above 40 cm long with an active growing period, artificial inoculations were conducted just below the terminal leaves with the syringe No:1 (Şahin et al., 2020a). After inoculation, plants were covered with plastic bags for 72 hours, in order to provide 95–100% relative humidity conditions (Layne and Quamme, 1975; Şahin et al., 2020a).

Shoot blight symptoms were measured at 28^th day after inoculation and the necrotic parts of inoculated shoots in proportion to whole shoot length were found. The average of two inoculations was taken and SI of parental cultivars and hybrid individuals were calculated (Thompson et al., 1962). Mean, maximum, and minimum SI values of each combination and SI classes of parental cultivars obtained by using Gardner scale: 0–10% highly resistant (HR), 10.1–30% resistant (R), 30.1–50% moderately susceptible (MS), 50.1–90% susceptible (S), and 90.1–100% highly susceptible (HS) (Gardner et al., 1980). Hybrids between 0 and 20% in terms of SI value were selected. In the next stage of the study these hybrids will be planted in orchard for fruit observations. Remaining plant materials, pots and soil were destroyed under quarantine conditions.

Heterosis and Heterobeltiosis Rates

Heterosis (Ht) and heterobeltiosis (Hbt) rates (%) were determined only in crossbreeding combinations with resistance levels (RL) obtained from SI values with given Formula 1. Especially, in the calculation of these parameters RL was used instead of SI, because of numerically high SI level did not mean that individuals have superior performance. Estimation of Ht and Hbt were calculated based on Formula 2 and 3 given below (Fehr, 1993; Hallauer and Miranda, 1981) on two period average in populations. In addition, the percentage of individuals showing Ht and Hbt was determined in each combination.

Formula 1. Resistance Level $\left(RL\right)=100-SI$

Formula 2. $Heterosis\left(Ht\right)=\frac{F1-MP}{MP}\ast 100$

Formula 3. $Heterobeltiosis\left(Hbt\right)=\frac{F1-BP}{BP}\ast 100$

F₁ = mean RL of hybridization combination.

MP = mid parent RL value of the particular F₁ cross [(Parent 1 + Parent 2)/2].

BP = better parent RL value in the particular F₁ cross (Parent 1 or Parent 2).

Statistical Analyses

For determination of parental cultivars’ SI, experiment was arranged in a completely randomized design with three replicates and each replicate contains four seedlings per cultivar. Statistical analyses were conducted using one-way ANOVA followed by LSMeans Student’s t test (p < .05) to determine the differences of parents, hybridization combinations, parental effects in reciprocal group, Ht and Hbt parameters.

Results and Discussion

The susceptibility indexes of 7 parents and 223 hybrids obtained from 9 combinations were investigated over two periods with artificial inoculations. The main effect of genotype was found significant (p ≤ 0.0001). Ege 25, Altın 35, Zeybek 35 varieties and Ayva B-35 rootstock were determined as resistant according to Gardner scale with 29.63%, 26.82%, 30.53%, and 28.51% SI, respectively. On the other hand, Ege 2, Ege 22 varieties and Quince A rootstock were stated moderately susceptible with 31.97%, 40.35% and 41.63% SI, respectively (Table 2). Şahin et al. (2020a), found these rootstocks (Quince A; 37.16%, Ayva B-35; 29.11%) and varieties (Ege 2; 31.49%, Ege 22; 38.71%, Ege 25; 30.49%, Altın 35; 28.58%, Zeybek 35; 30.91%) as moderately susceptible and resistant. Moreover, Ekmek variety from Turkey and Isfahan variety from Iran had 45.44–46.9% and 50% SI values, respectively (Abdollahi et al., 2008; Ozrenk et al., 2012; Şahin et al., 2020a). In the field surveys, the severity of the disease was found to be lower in Ege 22 cultivar than Eşme, Limon and Ekmek varieties (Şahin et al., 2019). These differences could be caused by age and growth force of trees, environmental and growing conditions, cultural processes, and inoculum concentration as well as genetic factors (Harshman et al., 2017; Paulin and Lespinasse, 1990; Şahin et al., 2020a, 2020b; Sobiczewski et al., 2015; Steiner, 2000).

Table 2. SI of parents and classification according to the Gardner scale

Parent*	Susceptibility indexes (%)**	Classes
Ege 2	31.97 b***	Moderately susceptible (MS)
Ege 22	40.35 a	Moderately susceptible (MS)
Ege 25	29.63 bc	Resistant (R)
Altın 35	26.82 c	Resistant (R)
Zeybek 35	30.53 bc	Resistant (R)
Ayva B-35	28.51 bc	Resistant (R)
Quince A	41.63 a	Moderately susceptible (MS)

*Genotype Kdz 12 was collected via seeds and only open pollinated seedlings tested.

**Decrease in numbers indicates that resistance increases.

***Mean separation within rows by LSD test (p < 0.05).0

The number of hybrids obtained from each combination, and distribution of these individuals into resistance classes are given in Table 3. Susceptibility indexes of hybrids were predominantly found in the range of 10.1–30% class. When the results are interpreted according to the Gardner scale, none of the hybrids was found in high resistant and highly susceptible group, while 77.6% of hybrids took place in resistant (10.1–30%), 21.5% moderately susceptible (30.1–50%), and only 0.9% susceptible group (50.1–90%) (Table 3). Neither parents nor hybrids have been found immune to fire blight. Similar to our findings, none of pear, apple and quince cultivars, selections, and hybrids was immune to fire blight (Evrenosoğlu and Mertoğlu, 2018; Kellerhals et al., 2010; Ozrenk et al., 2012; Şahin et al., 2020a; Van der Zwet et al., 1974).

Table 3. Number of individuals in classes according to percentage of lesion length and maximum minimum and mean of SI values

	Resistance classes					Statistical data*
Combination code	0– 10%	10.1–30%	30.1–40%	50.1–60%	90.1–100%	Mean (%SI)	Min. (%SI)	Max. (%SI)
A × B-35	0	4	1	0	0	26.28 abc	19.22	34.23
B-35 × A	0	78	28	0	0	27.47 ab	14.77	45.83
Z × B-35	0	8	5	1	0	31.51 a	18.64	52.39
Ege2 × x 25	0	5	1	0	0	24.90 bc	18.55	32.83
Altın35SP	0	7	3	0	0	28.50 ab	17.25	47.65
Ege22SP	0	7	3	0	0	27.07 abc	19.29	37.87
Ege25SP	0	21	4	1	0	25.69 bc	17.94	52.94
K12OP	0	11	1	0	0	24.36 bc	16.11	32.85
B-35OP	0	32	3	0	0	23.62 c	15.06	42.06

* Decrease in numbers indicates that resistance increases and mean separation within rows by LSD test.

Regardless of hybrid combinations, SI values ranged between 14.77 and 52.94% (Tables 3 and 4). When the average SI values of the combinations were compared, it was determined that the differences were found significant (p < .0076). Population held from open pollination of Ayva B-35 had the best resistant level with 23.62% SI, while Zeybek 35 × Ayva B-35 hybrid combination was determined as the most susceptible group with 31.51% SI (Table 3). Evrenosoğlu and Mertoğlu (2018) determined SI values of hybrid pear combinations in the range of 19.69–92.31%. Although, quince hybrids are seen more resistant than pear hybrids that tested in Turkey in terms of fire blight resistance some differences in addition to genetic structure such as virulence of strain, environmental conditions, inoculation methodology affect the disease resistance. Forty hybrids were selected with 10.1–20% SI and will be transferred into advanced observation parcel for further investigations (Table 4). Nineteen of these individuals were obtained from different hybrid combinations (Quince A × Ayva B-35, Ayva B-35 × Quince A, Zeybek 35 × Ayva B-35, Ege 2 × Ege 25), whereas 15 from open-pollination (Kdz 12, Ayva B-35) and 6 from self-pollination (Altın 35, Ege 22, Ege 25). Bobev et al. (2011) selected 19 individuals from 274 hybrids for showing high levels of resistance under natural fire blight infection conditions. Ten of these hybrids were obtained from open-pollination (Triumph, Asenitza, Hemus varieties) and nine of them from different hybrid combinations (Hemus × Tzarigradska, Asenitza × Triumph, Asenitza × Du Portugal, Hemus × Tzarigradska, Asenitza × Du Portugal, Asenitza × Triumph, Tzarigradska × Du Portugal, Asenitza × Triumph, Tzarigradska × Triumph).

Table 4. SI values of quince hybrids for each combination

♀x♂	Hybrid number	Mean SI (%)	♀x♂	Hybrid number	Mean SI (%)	Hybrid number	Mean SI (%)	♀x♂	Hybrid number	Mean SI (%)
Ayva B-35 – OP*	B-35OP/1	17,31	Ayva B-35 x Quince A	B-35xA/1	27,62	B-35xA/68	22,27	Zeybek 35 x Ayva B-35	ZxB-35/7	38,80
	B-35OP/2	27,58		B-35xA/2	21,13	B-35xA/69	14,77		ZxB-35/8	38,20
	B-35OP/3	27,08		B-35xA/3	27,38	B-35xA/70	23,60		ZxB-35/9	29,69
	B-35OP/4	42,06		B-35xA/4	28,47	B-35xA/71	18,46		ZxB-35/10	33,81
	B-35OP/5	23,53		B-35xA/5	28,10	B-35xA/72	27,48		ZxB-35/11	28,23
	B-35OP/6	22,35		B-35xA/6	34,80	B-35xA/73	22,29		ZxB-35/12	19,06
	B-35OP/7	17,46		B-35xA/7	30,89	B-35xA/74	19,12		ZxB-35/13	28,93
	B-35OP/8	28,80		B-35xA/8	30,61	B-35xA/75	36,06		ZxB-35/14	28,59
	B-35OP/9	24,48		B-35xA/9	30,94	B-35xA/76	30,48
	B-35OP/10	18,95		B-35xA/10	23,14	B-35xA/77	27,83	Altın 35 x Altın 35	Altın35SP/1	18,89
	B-35OP/11	15,06		B-35xA/11	23,34	B-35xA/78	22,26		Altın35SP/2	22,28
	B-35OP/12	27,41		B-35xA/12	29,40	B-35xA/79	20,10		Altın35SP/3	47,65
	B-35OP/13	18,96		B-35xA/13	29,94	B-35xA/80	26,27		Altın35SP/4	24,07
	B-35OP/14	21,90		B-35xA/14	32,74	B-35xA/81	20,81		Altın35SP/5	36,84
	B-35OP/15	22,15		B-35xA/15	32,98	B-35xA/82	22,27		Altın35SP/6	17,25
	B-35OP/16	17,76		B-35xA/16	27,79	B-35xA/83	37,71		Altın35SP/7	28,95
	B-35OP/17	19,53		B-35xA/17	25,35	B-35xA/84	32,14		Altın35SP/8	25,91
	B-35OP/18	24,18		B-35xA/18	21,90	B-35xA/85	23,72		Altın35SP/9	34,81
	B-35OP/19	18,98		B-35xA/19	21,79	B-35xA/86	45,83		Altın35SP/10	28,37
	B-35OP/20	22,01		B-35xA/20	25,44	B-35xA/87	22,50
	B-35OP/21	27,32		B-35xA/21	26,18	B-35xA/88	25,64	Ege 22 x Ege 22	Ege22SP/1	21,96
	B-35OP/22	29,92		B-35xA/22	34,08	B-35xA/89	19,37		Ege22SP/2	23,06
	B-35OP/23	19,94		B-35xA/23	26,53	B-35xA/90	23,40		Ege22SP/3	19,86
	B-35OP/24	27,91		B-35xA/24	31,84	B-35xA/91	22,48		Ege22SP/4	23,24
	B-35OP/25	26,90		B-35xA/25	24,20	B-35xA/92	29,17		Ege22SP/5	19,29
	B-35OP/26	26,46		B-35xA/26	20,26	B-35xA/93	21,73		Ege22SP/6	37,87
	B-35OP/27	23,34		B-35xA/27	29,25	B-35xA/94	40,84		Ege22SP/7	30,06
	B-35OP/28	19,98		B-35xA/28	26,67	B-35xA/95	41,63		Ege22SP/8	31,99
	B-35OP/29	23,37		B-35xA/29	25,91	B-35xA/96	24,15		Ege22SP/9	35,95
	B-35OP/30	23,26		B-35xA/30	19,70	B-35xA/97	34,88		Ege22SP/10	27,41
	B-35OP/31	19,53		B-35xA/31	26,04	B-35xA/98	23,98
	B-35OP/32	28,76		B-35xA/32	32,17	B-35xA/99	30,71	Ege 25 x Ege 25	Ege25SP/1	34,78
	B-35OP/33	31,17		B-35xA/33	36,85	B-35xA/100	38,40		Ege25SP/2	52,94
	B-35OP/34	17,84		B-35xA/34	23,61	B-35xA/101	32,03		Ege25SP/3	17,94
				B-35xA/35	23,76	B-35xA/102	24,51		Ege25SP/4	25,92
Kdz 12 – OP	K12OP/1	26,59		B-35xA/36	37,45	B-35xA/103	33,02		Ege25SP/5	22,84
	K12OP/2	26,16		B-35xA/37	22,61	B-35xA/104	16,65		Ege25SP/6	29,86
	K12OP/3	30,39		B-35xA/38	22,87	B-35xA/105	23,22		Ege25SP/7	26,89
	K12OP/4	24,04		B-35xA/39	31,62	B-35xA/106	28,96		Ege25SP/8	30,44
	K12OP/5	21,74		B-35xA/40	20,40				Ege25SP/9	26,64
	K12OP/6	26,69		B-35xA/41	19,07				Ege25SP/10	24,86
	K12OP/7	24,99		B-35xA/42	21,30				Ege25SP/11	21,86
	K12OP/8	16,11		B-35xA/43	26,49				Ege25SP/12	20,93
	K12OP/9	16,31		B-35xA/44	28,73				Ege25SP/13	26,45
	K12OP/10	16,89		B-35xA/45	32,87				Ege25SP/14	30,05
	K12OP/11	32,85		B-35xA/46	41,47				Ege25SP/15	21,80
	K12OP/12	29,50		B-35xA/47	34,36				Ege25SP/16	34,32
				B-35xA/48	29,23				Ege25SP/17	28,47
Ege 2 x Ege 25	Ege2x25/1	32,83		B-35xA/49	37,66				Ege25SP/18	21,17
	Ege2x25/2	18,55		B-35xA/50	22,33				Ege25SP/19	21,10
	Ege2x25/3	22,44		B-35xA/51	32,05				Ege25SP/20	32,85
	Ege2x25/4	25,83		B-35xA/52	37,22				Ege25SP/21	23,29
	Ege2x25/5	25,01		B-35xA/53	30,62				Ege25SP/22	29,59
	Ege2x25/6	24,77		B-35xA/54	20,98				Ege25SP/23	23,52
				B-35xA/55	33,84				Ege25SP/24	32,81
Quince A x Ayva B-35	AxB-35/1	25,05		B-35xA/56	19,45				Ege25SP/25	21,52
	AxB-35/2	19,22		B-35xA/57	28,97				Ege25SP/26	23,47
	AxB-35/3	26,87		B-35xA/58	30,22
	AxB-35/4	26,05		B-35xA/59	34,77
	AxB-35/5	34,23		B-35xA/60	19,75
				B-35xA/61	34,53
Zeybek 35 x Ayva B-35	ZxB-35/1	27,88		B-35xA/62	23,51
	ZxB-35/2	41,00		B-35xA/63	28,68
	ZxB-35/3	18,64		B-35xA/64	29,70
	ZxB-35/4	24,00	B-35xA/65	15,58
	ZxB-35/5	52,39	B-35xA/66	24,16
	ZxB-35/6	31,97	B-35xA/67	28,22

*OP: open pollination.

Ayva B-35 is a new quince rootstock (Şahin and Mısırlı, 2016) used in 3 hybrid combination (♀×♂) and open-pollination. Open-pollinated seedlings of Ayva B-35 had significant disease resistance ranging between 15.06 and 42.06% SI (Table 3), and 85.29% of them were found to have higher resistance than maternal parent (Tables 2 and 4). Increase of resistance may be due to Ayva B-35 that the source of resistance or it can be suggested that resistant genotypes which overlapping blooming period in quince genetic resources played a significant role in pollination. Besides, open-pollination (Kdz 12 and Ayva B-35) had the best distribution rate in resistance classes with 91.67–94.12%, respectively (Table 5). Similar to this study, Bobev et al. (2011) suggested that resistance cultivars play significant role in terms of resistance in open pollinated combinations, and via this way two hybrids obtained with high resistance from Azenitza cultivar which is very susceptible to fire blight.

Table 5. Distribution of hybrids through susceptibility classes according to Gardner scale (%)

Combination codes	Highly resistant (HS)	Resistant (R)	Moderately susceptible (MS)	Susceptible (S)	Highly susceptible (HS)
Quince A × Ayva B-35 (MS×R)	-	80.00	20.00	-	-
Ayva B-35 × Quince A (R× MS)	-	73.58	26.42	-	-
Zeybek 35 × Ayva B-35 (R × R)	-	57.14	35.71	7.14	-
Ege 2 × Ege 25 (MS×R)	-	83.33	16.67	-	-
Altın 35 Self-pollination (R)	-	70.00	30.00	-	-
Ege 22 Self-pollination (MS)	-	70.00	30.00	-	-
Ege 25 Self-pollination (R)	-	16.67	66.67	16.67	-
Kdz 12 Open-pollination -	-	91.67	8.33	-	-
Ayva B-35 Open-pollination (R)	-	94.12	5.88	-	-

Individuals of Zeybek 35 × Ayva B-35 combination had SI in the range of 18.64 and 52.39 SI, and 57.14% of them were found more resistant than maternal parent and 37.51% than male parent. Quince A × Ayva B-35 combination had 19.22–34.23% SI and all hybrids were found more resistant than maternal parent, while 80% than male parent. In Ayva B-35 × Quince A combination, SI varied from 14.77 to 45.83% and 57.54% of hybrids were found more resistant than maternal parent while 98.11% more resistant than male parent (Tables 2 and 4). In this combination, it was found that the resistance was higher and showed a wide variation. Ayva B-35, which is resistant than Quince A, had significant effect when used male parent in crossbreeding (Tables 3 and 5). On the contrary, Evrenosoğlu et al. (2019) suggested that using resistant cultivar as a female parent is more effective than using male parent for transferring resistance to pear hybrids.

Ege 2 and Ege 25 varieties are very important in terms of fruit quality. These two parents are well-known cultivars in Turkey (Şahin and Mısırlı, 2016) and they also had such an important resistance with 31.97 and 40.35% SI, respectively (Table 2). The hybrids of these varieties had 18.55–32.83% SI values and 83.33% of them were found more resistant than maternal parent and 16.67% than male parent (Tables 2 and 4).

As a result of self-pollination on Altın 35, Ege 22, and Ege 25 cultivars, SI values ranged between 17.25–47.65%, 19.29–37.87%, and 17.94–52.94%, respectively. Ege 22, which is in moderately susceptible group, had the best SI on self-pollinated hybrids and came into prominence for tolerance of disease (Tables 2 and 3). Ege 22 was also found more tolerant to disease compared to some other important varieties and have gained importance in İzmir province in recent years (Şahin et al., 2019). Altın 35 and Ege 25 were located in resistant group (Table 2) and their self-pollinated hybrids showed a wide variation in resistant, moderately susceptible and susceptible groups (Table 5).

Isfahan cultivar and 29 genotypes were examined for fire blight susceptibility via artificial inoculations and SI values were found between 13.1 and 90% in Iran (Abdollahi et al., 2008; Mehrabipour et al., 2012). In the same study, SI values of Isfahan cultivar was reported as 50%. Totally 62 quince genotypes were tested in Turkey up to date and SI values ranged between 17.8 and 51.69% (Ozrenk et al., 2012; Şahin et al., 2020a). To summarize of these sections with our study, 11 hybrids were found to be higher resistant than tested ones in Turkey.

Our research is the first study that conducted on heredity, Ht and Hbt rates (%) for fire blight resistance on quince. Distribution of hybrids through susceptibility classes according to Gardner scale was also seen in Table 4 and this can give preliminary information about the inheritance of resistance to fire blight in this species. As a result of hybridization, Quince A × Ayva B-35 (moderately susceptible × resistance) parents, 80% of obtained F₁ were determined to be resistant. Its reciprocal combination had slightly lower values. Reciprocal effects on fire blight resistance were reported in pears (Evrenosoğlu et al., 2020) and blackberries (Stewart et al., 2005), too. Additionally, considering the hybrids of Ege 2 × Ege 25 (moderately susceptible × resistance), 83.33% of them were determined to be resistant. Hybrids of resistance × resistance combinations were distributed in 3 groups as resistant, moderately susceptible and susceptible with distribution percentages of 57.14, 35.71, and 7.14, respectively (Table 5). According to these results, it could be said that the resistance to fire blight in the quince is transferred recessively.

It was observed that, resistance could not be transferred optimally in resistant × susceptible or resistant × resistant hybrid combinations. On the contrary, when two moderately susceptible parents were crossed, a maximum resistance level could be obtained (Van der Zwet et al., 1974). Resistant individuals were also obtained from moderately resistant and/or susceptible combinations (Evrenosoğlu et al., 2014; Rosati et al., 2002). At this point, it should be considered that, genotype has an effect upon resistance to the disease at the rate of 1/3 in general and inheritance of fire blight resistance is low with minor effects (Bokszczanin et al., 2012). Indeed, it has been reported general combining and specific combining abilities of each cultivar is different in terms of fire blight resistance (Evrenosoğlu et al., 2019).

The results of the Ht and Hbt evaluations are presented in the Table 6. 88.67% of Ayva B-35 × Quince A hybrids showed Ht and 57.54% Hbt for fire blight resistance, reciprocal of this combination had 100% Ht and 80% Hbt. On the other hand, individuals of Ege 2 x Ege 25 had the same percentage in both parameters. Ht rates between combinations were found significant (p < .0004), while Hbt were determined statistically not significant (p < .0623). The lowest rates were found in Zeybek 35 × Ayva B-35 combination, also Ht and Hbt rates negative effects were obtained for fire blight resistance. Ht and Hbt rates were found between −2.82–13.54, −4.20–6.72 in combinations, respectively and identified positive effects for fire blight resistance except Zeybek 35 × Ayva B-35 combination. The maximum and significant heterotic effect were observed in Quince A × Ayva B-35 and its’ reciproc combination statistically, whereas heterobeltiotic effect was in Ege 2 × Ege 25 (statistically not significant).

Table 6. Ht and Hbt rates for fire blight resistance and % of individuals showing Ht and Hbt in combinations

Combinations	Heterosis (Ht)*	Heterobeltiosis (Hbt)	% of individuals showing**
			Ht	Hbt
Ayva B-35 × Quince A	11.70 a	1.45	88.67	57.54
Quince A × Ayva B-35	13.54 a	3.12	100	80.00
Zeybek 35 × Ayva B-35	−2.82 b	−4.20	50.00	35.71
Ege 2 × Ege 25	7.86 ab	6.72	83.33	83.33

Ht and Hbt were determined with resistance levels (RL) that obtained SI values with given Formula 1.

Mean separation within rows by LSD test (p < 0.05).

**Hybrids showing Ht and Hbt is being given as a percentage for each combination.

Studies on heterosis and heterobeltiosis in fruit species are limited and generally used for quantitative characteristics and disease resistance (Cardoso et al., 2014; Liu et al., 2019; Marin et al., 2006; Vivas et al., 2014). There is no study on fire blight resistance about genetic pattern in quince. The observed positive heterosis and heterobeltiosis for fire blight resistance could be a breeding advantage. By the way conducting such studies in this polygenic disease may contribute to achievement of more resistant hybrids for example, in more extensive hybridization studies for fire blight resistant rootstock breeding, it will be possible to obtain hybrids with very high resistance by planning more individuals from Quince A × Ayva B-35 combination that have high heterosis rate.

Conclusion

This study is the first study on controlled hybridization and heredity process of fire blight conducted on quince. It was concluded that heredity of fire blight is recessive and using resistant male parent in crossbreeding could be more effective than female parent on transferring resistance to hybrids. The presence of hybrids that are resistant to fire blight is important and they could be used as registered variety or genitor. The observed positive heterosis and heterobeltiosis for fire blight resistance give an information and could be a breeding advantage to get higher resistant hybrids. On the other hand, open-pollination is an easy and economical way for breeders in terms of obtaining resistance hybrids if the orchard has resistant pollinators.

Additional information

Funding

This research was carried out as a preliminary study within the scope of “Fire Blight Tolerant Quince Breeding Project (Project number: TAGEM/BBAD/16/A08/P03/04)” that was funded by the Republic of Turkey Ministry of Agriculture and Forestry General Directorate of Agricultural Research and Policies.