Inheritance of Apple (Malus × domestica (L.) Borkh) Resistance against Apple Scab (Venturia inaequalis (Cooke) Wint.) in Hybrid Breeding Material Obtained by Gene Pyramiding

Abstract

Apple scab caused by the pathogenic fungus Venturia inaequalis causes significant damage to apples. The creation of apple-scab resistant varieties is considered an alternative to pesticide-based management. To evaluate R gene inheritance, 862 apple hybrid samples were analyzed and divided into populations depending on the resistance genes (Rvi6 and Rvi5) present in the parents’ genotypes and their combinations. The field evaluation was carried out in 2016–2018 using the international VINQUEST scab evaluation methodology, but the genes were identified using specific scab resistance-gene molecular markers. Parent plants were genotyped using 22 SSR markers. The study determined that the field resistance of apple scab is influenced not only by the resistance genes present in the genotype but also by the genetic background of the apple cultivar, the tree’s general health status and resistance to other diseases. It was found that the apple scab resistance genes Rvi6 (Vf) and Rvi5 (Vm) are inherited according to Mendelian laws, and when both genes are pyramided in offspring, they are inherited as separate genes. In general, the inheritance of both genes in the offspring is not a determining factor for apple field resistance, as field resistance to scab is influenced by a combination of several factors.

1. Introduction

Domestic apple (Malus × domestica Borkh.) belongs to the Malus genus of the subfamily Maloideae of the Rosaceae family [1,2]. It is assumed that the Malus genus consists of between 30 [3] to 55 [1] species and subspecies. However, due to difficulties in taxonomic distinctions, different sources cite different numbers [4]. Apples are cultivated widely and globally, M. × domestica being the most popular species for commercial growing. It is the fourth most popular fruit crop globally, after citrus, grapes and bananas, being the most important commercial fruit tree crop in the temperate climate region [5]. There are approximately 10,000 apple cultivars. However, only a small number are widely cultivated globally—by replacing local and traditional cultivars with popular commercial ones, genetic diversity is decreasing, thus affecting the quality of apple breeding in the long run [6].

One of the biggest threats to apple cultivation is apple scab, a disease distributed worldwide [7] but more common in temperate regions characterized by cool and humid springs and summers, and its causal agent is the hemibiotrophic fungus Venturia inaequalis Cooke (Wint.) [5]. Fruit and leaf deformities and premature fall occur due to apple scab, which affects the tree’s overall health and decreases its resistance to unfavorable environmental factors [8]. Intensive use of pesticides is needed during the season, spraying up to 20 times per season to suppress the spread of apple scab [9]. Such intensive farming creates a risk for the environment and consumers and facilitates the evolution of more resistant pathogen races. A solution to this problem is breeding resistant apple cultivars, whose resistance is durable in large commercial plantations and under favourable environmental conditions for pathogen development [10]. Most modern scab-resistant apple cultivars are based on the Rvi6 (Vf) resistance gene, which was first identified in Malus × floribunda 821 and has remained the most studied apple scab resistance gene [11]. Another gene important in apple breeding is the Rvi5 (Vm) gene, whose donors are Malus × atrosanguinea 804 and Malus micromalus 245-38 [12]. As mentioned in the literature, resistance to apple scab is dependent on a cultivar’s interaction with a specific fungi population; thus, plant resistance can differ not only in different regions but also in orchards within one region [13]. Therefore, one of the apple scab controlling strategies is gene pyramiding—especially the combination of R genes and quantitative trait loci (QTL) [10,14], ensuring new cultivars with at least two or more resistance genes ensure long-lasting resistance to apple scab [12,15].

Marker-assisted selection (MAS) is a method for streamlining the breeding process, with molecular markers being a surefire way to follow the inheritance of valuable traits [16]. Molecular markers in apple breeding are employed to determine the presence of resistance genes in the breeding population, select parental lines with desired resistance gene combinations, and perform gene pyramiding to assess gene combinations in the population that cannot be evaluated by phenotyping [14]. To date, 20 different Rvi resistance genes have been mapped in apples [11] and each of them has at least one matching molecular marker developed for research purposes. Many papers focus on using several markers to identify the Rvi6 (Vf) resistance gene [17,18,19], with dominant and codominant markers employed to distinguish homozygous and heterozygous genotypes [20]. MAS is a method usually validated for Mendelian traits, of which apple scab is one, however, the success rate of predicting it can be from 35 to 95% [21]. Despite a wide margin of success, markers and methods for improving MAS have been developed in a large-scale project in the European Union [22], as well as in a decade-long project to detect the best methods for introgressing apple scab resistance genes to ensure high resistance [23]. Molecular markers have been used as part of the breeding program in Latvia to detect the Rvi6 gene [24] and have become a routine method in apple hybrid material screening. The current Latvian apple breeding program is targeted at: cultivars adapted to the cool climate of Baltic countries and Scandinavia; complex disease resistance; fruit quality suitable for commercial production; improved storage and biochemical content. Crosses with donors of scab resistance genes Rvi6 and Rvi5 began in 1989. Cultivars ‘Roberts’, ‘Dace’, ‘Gita’, ‘Monta’, ‘Inese’, ‘Edite’, ‘Ligita’ and others possess the Rvi6 gene because of controlled breeding [25]. Genetic analysis of Rvi6 alleles has shown that modern cultivars inherited Rvi6 in a dominant state [26].

Considering the above information, the study aimed to evaluate and characterize the heredity of the domestic apple (Malus × domestica) resistance to scab (V. inaequalis), assess the effect of resistance gene pyramiding, and evaluate its influence on apple resistance in the field.

2. Materials and Methods

2.1. Plant Material and Cultivation Practice

The study included 882 apple hybrids of 6 breeding populations (

Table 1. Characterization of apple hybrid populations.

Population Number	Population Code	Combination of Resistance Genes	Parent Plants	Number of Hybrids
1	MA-2	Rvi6 × susceptible cultivar	‘Arbat’ × ‘Signe Tillisch’	274
2	CO-3A	Rvi6 × susceptible cultivar	‘Arbat’ × ‘Zane’	119
3	VM-5	Rvi6 × Rvi5	‘Dayton’ × ‘Pervinka’	135
4	VF-3	Rvi6 × Rvi6	‘Kandil Orlovskij’ × ‘Florina’	117
5	VF-4	Rvi6 × Rvi6	‘Kurnakovskoje’ × ‘Rewena’	132
6	VM-2	Rvi6 × Rvi5	‘Scarlett O’Hara’ × D-1-92-32	85

) and 11 parent cultivar plants, selected at the Institute of Horticulture (LatHort) experimental fields in Dobele, Latvia (56°37′0″ N, 23°16′0″ E). The hybrid material has been developed by different crosses of parent plants varying in resistance to the apple scab (Table S1). The crosses were performed in 2009−2010 with the aim of pyramiding scab resistance genes and other traits valuable in breeding. Parent cultivars included donors of genes Rvi6 (‘Arbat’, ‘Dayton’, ‘Florina’, ‘Kandil Orlovskij’, ‘Kurnakovskoje’, ‘Rewena’, ‘Scarlett O’Hara’) and Rvi5 (‘Pervinka’, D-1-92-32), winter hardiness (‘Arbat’, ‘Kandil Orlovskij’, ‘Kurnakovskoje’, ‘Pervinka’), compact growth (‘Arbat’, ‘Pervinka’, D-1-92-32, ‘Zane’) and fruit quality (‘Dayton’, ‘Florina’, ‘Rewena’, ‘Scarlett O’Hara’, ‘Signe Tillisch’, ‘Zane’). Seedlings were evaluated in the 1st growth year for tree habit and general health, including naturally occurring mildew infection. The seedlings selected after first evaluation were planted in field and used for gene marker analysis (23 to 43% from a cross combination). The hybrid populations for this study were chosen based on the presence of apple scab resistance genes Rvi5 and Rvi6 in donor genotypes. To assess the heritability of apple scab resistance and resistance gene pyramiding effectiveness, the hybrids were divided into three groups, with two hybrid populations in each: (1) one parent has the resistance gene Rvi6, but the other parent can be considered susceptible to apple scab (populations 1 and 2); (2) each parent had a different resistance gene (Rvi5 or Rvi6, respectively, populations 3 and 6) and (3) both parents had the Rvi6 gene (populations 4 and 5).

The seedlings were planted on their own roots, at distances of 1 × 5 m. Tree strips were maintained free of weeds, and grass was grown in alleyways, mowing it 2 times in season. Fertilizing was conducted in spring according to soil analysis data. The same insecticide and acaricide treatment were used as in commercial orchards, uniformly applied to all the seedlings (1–2 treatments in a growing season, depending on pest occurrence). No fungicide sprayings were applied. The response of cultivars to diseases was assessed in natural field conditions of infection.

The meteorological data were collected by a “Lufft” meteorological station located at the LatHort site (Figure S1).

2.2. Field Evaluation of Apple Scab Severity and Tree Health Status

Field evaluation was performed to gather data on the overall tree health status and severity of apple scab on leaves. The evaluation was performed visually in the summers of 2016, 2017 and 2018, at the end of the primary apple scab development season. The overall health status of the tree was evaluated on a scale from 0 to 9: 0—tree is dead/evaluation is not possible, 1—tree has an excellent health, 2—transitory state, 3—good health, 4—transitory state, 5—average health, 6—transitory state, 7—weak tree, 8—transitory state and 9—the tree is dying. The dead trees were not evaluated in subsequent years to exclude their impact on the general improving average health of evaluated trees. A scale developed by VINQUEST (www.vinquest.ch/monitoring/collection.htm, accessed on 11 July 2022) was used in the evaluation of apple scab severity on leaves (

Table 2. Apple scab severity scoring scale.

Score	Definition of the Symptoms	Proportion of Affected Organs (%)
0/n	No observation (missing plant)	-
1	No visible symptom	0%
2	One or very few lesions detectable on scrutiny of the tree	0 to 1%
3	Immediately apparent lesions in general clustered in few parts of the tree	1 to 5%
4	intermediate	±15%
5	Numerous lesions widespread over a large part of the tree	±25%
6	intermediate	±35%
7	Severe infection with half of the leaves badly infected by multiple lesions	±50%
8	intermediate	±75%
9	Tree completely affected with (nearly) all the leaves badly infected by multiple lesions	>90%

), using a 9-point system.

2.3. Genotyping Using Scab Resistance Gene and Genetic Relatedness Molecular Markers

Samples of apple leaves were collected in June—when new leaves had developed but had not yet experienced apple scab symptoms and stored in −80 °C until DNA extraction. DNA isolation was performed using the Genomic DNA Purification Kit (Thermo Scientific™, Vilnius, Lithuania) according to the manufacturer’s instructions. The DNA was then quantified, its quality was assessed by using NanoDrop™ 1000 Spectrophotometer (ThermoFisher Scientific™, Waltham, MA, USA) and samples standardised.

The presence of Rvi5 and Rvi6 genes in apple hybrid populations was determined using molecular markers specific for these genes according to published PCR protocols [27,28] using DreamTaq Green PCR Master Mix (Thermo Scientific™, Vilnius, Lithuania) and a Mastercycler epgradient thermal cycler (Eppendorf, Hamburg, Germany). Identification and length determination of amplification fragments were performed using 1% agarose gel in 1 × TAE buffer, fragment length standard GeneRuler™ 100 bp Plus DNA Ladder (Thermo Scientific™, Vilnius, Lithuania), visualised by ethidium bromide staining. The amplification fragment bands were scored manually.

The identification of apple scab resistance genes in parental genotypes was performed using molecular markers described by A. Patocchi and colleagues [29]. About 50 ng of genomic DNA was used for PCR amplification in a 20 µL reaction using DreamTaq Green PCR Master Mix and a Mastercycler epgradient thermal cycler. Genetic relatedness of parental genotypes was tested using 22 SSR molecular markers (Table S2) according to published marker information [30,31,32,33] using DreamTaq Green PCR Master Mix and a Mastercycler epgradient thermal cycler. In both cases the testing for the presence of PCR products was performed on a 1% agarose gel in 1× TAE buffer, visualised by ethidium bromide staining. The same PCR products were then analysed on an ABI PRISM^® 3100 Genetic Analyzer (Applied Biosystems, Waltham, MA, USA) and genotyped using GeneMapper^® Software v4.0 (Applied Biosystems, Waltham, MA, USA).

2.4. Data Analysis

Data analysis was carried out with R version 4.1.3 [34]. A one-way ANOVA test, LSD Post Hoc and a Duncan test was performed [34] to compare the mean values of apple tree health, apple scab severity, as well as mean values of meteorological data (precipitation, average temperature) throughout a three-year study period. Spearman correlation analysis was performed to understand the relationship between apple scab severity and apple tree health, precipitation, leaf moisture, air humidity as well as average temperature.

Analysis of the applied markers was performed with the software ‘SSRs’ and GenAlEx 6.5 [35] to determine marker heterozygosity, polymorphism information content and discrimination power. Marker data was converted to a binary format where the presence of a fragment was written down as 1, but its absence as 0. The same process was conducted for the data of markers detecting the genes Rvi6 and Rvi5—if Rvi5 was present in the sample it was written down as 1, and 0 in the opposite case. In the case of Rvi6 allele state was recorded, thus a homozygotic recessive state (vfvf) was recorded as 1, heterozygotic (Vfvf) state as 2 and homozygotic dominant (VfVf) state was recorded as 3. Principal component analysis (PCA) was performed with the R package v. 2.6-2 (R Foundation for Statistical Computing, Vienna, Austria) on the marker data to determine the relationship of the parent genotypes [34].

3. Results

3.1. Resistance of Hybrid Seedling Material against Apple Scab in the Field

Two parameters were assessed to evaluate the hybrid seedling resistance to apple scab—overall health of tree and apple scab severity on leaves. Initially it was planned to evaluate apple scab severity on fruit, however, many of the hybrid trees were not mature and did not yield fruit or had unreliable yields; therefore, it was not possible to gather data throughout the entire study period and the data was discarded. Apple tree health score was low throughout the study years, with a moderate improvement each year, mean scores were 1.9, 1.7 and 1.4 each year, respectively. Apple scab severity scores were similar in 2016 and 2017: 4.08 and 4.12, respectively, whereas 2018 saw a drop in apple scab severity with a score of 2.90.

The lowest recorded apple tree health score throughout the study was 1, meaning the trees were in excellent health, which was observed each year of the study. The number of trees with excellent health varied from 408 in 2016, to 491 in 2017 and 621 in 2018. The highest tree health score was recorded in 2016 for the tree VM-5-100, which had a score of 9, meaning the tree was close to dying. The following years the highest score recorded was 8: three genotypes in 2017 (VM-5-143, VM-5-151, VM-5-154) and two in 2018 (VF-4-268, VM-5-13). The highest score of apple scab severity recorded was also 9 and this observation was made each year, which meant a severe disease incident in the case of scab. Three genotypes—VM-5-130, VM-5-169, VM-5-3—scored a 9 in apple scab severity in 2016. The following year saw an increase in severely diseased apple trees, as 49 genotypes had a score of 9, however the average score of the entire sample group was not strongly impacted. Only one genotype, VM-5-169, had a score of 9 in 2018. The lowest score of apple scab recorded was 1, which was the lowest score each year in our study. The number of genotypes fluctuated between years: 62 in 2016, 125 in 2017 and 87 in 2018. When comparing data between the years (Figure 1), significant differences (p < 0.05) between the years were found both in the case of tree health and apple scab severity.

Meteorological data was recorded for the April–June period as it is the time of primary apple scab infection. The average temperature of this period fluctuated between years: 12.9 °C in 2016, 10.6 °C in 2017, reaching the highest temperature, 14.1 °C, in 2018. Statistically significant (p < 0.05) differences were found between the average temperature of 2017 and the other two years. When comparing months, statistically significant differences were found in 2016 and 2017 between all three months, whereas in 2018 May and June did not have statistically significant differences, the temperature was similar: 16.5 °C in May and 17.1 °C in June. Despite 2018 being the warmest year on average, the warmest month was June 2016, with an average temperature of 17.3 °C.

Precipitation gradually decreased throughout the years, with 2016 having the highest sum—158.5 mm over a three-month period, whereas 2017 and 2018 had lower precipitation—97.7 mm and 80.4 mm, respectively. Every year May was the month with the lowest precipitation, with April coming in second in 2016 and 2017, while in 2018 June was the month with the second-lowest precipitation sum. June 2016 had the highest precipitation out of all years—96.1 mm, contributing more than half of that year’s rainfall in the study period.

The evaluation of the entire population showed statistically significant (p < 0.001) correlations observed in all cases (Table S3)—negative with average temperature (r = −0.05), positive in the case of apple scab severity and tree health (r = 0.36), precipitation (r = 0.03), and air and leaf humidity (r = 0.04 and r = 0.03). The same relationships were observed for correlations within groups of populations; however, it is important to note that in most cases the correlations were almost negligible except a weak correlation between tree health and apple scab severity.

Tree health was good in all populations, with an average score of 1.6 across all six hybrid seedling populations. Individually, the mean health score of each population was as follows—1.6, 1.4, 2.3, 1.5, 1.8, 1.2 (populations 1 through 6). When comparing highest and lowest scores, every single population had seedlings whose tree health could be described as excellent (1 point), however, the portion of seedlings in excellent health varied by population. Out of all seedlings, 37% had a score of 1 in population 1, 45% in population 2, 15% in population 3, 30% in population 4, 31% in population 5 and 66% in population 6. Not a single population had a health score higher than 7 points; only MA-2-327 from population 1 had a score of 7. Overall, the highest score in a population was observed in a low number of seedlings. In populations 2 and 4 the highest scores were 6: seedling CO-3A-205 (population 2), seedling VF-3-100 (population 4). Seedling VM-5-165 of population 3 had a score of 6.3, same as seedling VF-4-268 from population 5. Population 6 had two seedlings score 3.3—VM-2-17 and VM-2-37—which was the highest score in the population.

Apple scab severity had an average score of 3.7 across all populations. Populations 1 and 2 had a mean score of 3.3, population 3 had the highest score (5.0), population 4 had a score of 3.6, population 5 had a score of 4.2 and population 6 had the lowest mean score—2.9 points. Four seedlings in population 3 (VM-5-102, VM-5-125, VM-5-126, VM-5-169) had the highest score out of all populations—9 points. The lowest score recorded was 1 and each population had at least 1 seedling with zero visible apple scab symptoms. Population 1 had 10 seedlings with a score of 1, the rest of the populations had under 5 seedlings each.

The overall health of trees and apple scab severity mean scores were significantly (p < 0.05) different between all populations (Figure 2). Population 6 had the healthiest trees with an average score of 1.2, population 3 had the worst tree health with a score of 2.3. While population 6 also had the lowest scores of apple scab severity, the highest severity of apple scab was observed in population 3. It is important to note that populations 6 and 3 had seedlings whose both parents had at least one resistance gene, whereas seedlings of population 1 had a susceptible parent.

An analysis of yearly changes within populations revealed that tree health differed significantly (p < 0.05) between all three years in populations 1, 3 through 6, whereas in population 2 tree health scores did not significantly differ between 2017 and 2018; in both years the score was 1.4. In the case of apple scab severity, scores differed significantly within all populations in a three-year period, reflecting the overall changes in apple scab severity scores.

The hybrid seedlings were also grouped depending on the resistance of their parents and the health of the trees and severity of apple scab was assessed (Figure 3); in all cases, there were significant differences (p < 0.05) between the different groups and the values of resistance parameters. Average scores of tree health were good: 1.5 in the Rvi6 × susceptible cultivar group, 1.9 in the Rvi6 × Rvi5 and 1.7 in the Rvi6 × Rvi6 group.

The best score recorded in all three groups was 1, in the Rvi6 × susceptible cultivar group, 39% of seedlings had a tree health score of 1, in the Rvi6 × Rvi5 group the percentage was 35 and 31% in the third group (Rvi6 × Rvi6), indicating that a similar number of seedlings in each group had excellent tree health. The worst tree health score recorded was 7 in the Rvi6 × susceptible cultivar group for seedling MA-2-327. Interestingly, the healthiest trees were observed in the group where one parent was susceptible, it also had the lowest severity of apple scab.

The highest apple scab severity score was observed in the group of seedlings whose parents contained the Rvi6 and Rvi5 genes, respectively; the mean score was 4.2. This group also had seedlings with the highest score (9 points) overall: seedlings VM-5-102, VM-5-125, VM-5-126, VM-5-169 each had a score of 9. The group which consisted of hybrids whose parents each had the Rvi6 gene scored 3.9 in apple scab severity. The lowest mean score, 3.3 points, was recorded in the Rvi6 × susceptible cultivar group. This group also had the highest number of seedlings which scored apple scab severity at 1 point—13 seedlings. The other two groups (Rvi6 × Rvi5 and Rvi6 × Rvi6) had four and five seedlings with a score of 1, respectively.

3.2. Apple Scab Resistance Specific Gene Identification in Hybrid Seedlings

The genes Rvi6 and Rvi5 were introduced into the seedling population via marker-assisted breeding (MAB) and genotyping was performed to determine the heritability of both genes and to evaluate the relationship between resistance genes in the population and disease severity in the field; 860 genotypes were genotyped overall. Rvi6 was identified as homozygotic recessive (vfvf) in 40% of genotypes, homozygotic dominant (VFVF) in 23% of genotypes and heterozygotic (VFvf) in 37% of genotypes. On the other hand, 63% of all seedlings had the gene Rvi5 (Vm) (138 genotypes).

Allele frequencies were calculated in each of the seedling populations to better assess the heritability of the Rvi6 and Rvi5 genes (Figure 4). It was observed that in populations 1 and 3, Rvi6 was mostly inherited in a homozygotic recessive state; the frequency was 52% in population 1 and 64% in population 3. The opposite was observed in population 5—Rvi6 was mostly homozygotic dominant with a frequency of 55%. In populations 3 and 6 a low frequency was observed for Rvi6 in a homozygotic dominant state (1% in population 3, 5% in population 6), but Rvi5 had a high allele frequency: 54% and 76%, respectively. As heterozygotic, Rvi6 had the highest frequency in populations 4 and 2–56% and 41%.

When evaluating allele frequency by grouping the seedlings according to parent resistance gene occurrence (Figure 5) it was observed that in cases where a parent with the Rvi6 gene was crossed with a susceptible cultivar, the progeny’s Rvi6 allele frequency was highest in a homozygotic recessive state (44% of genotypes), followed by heterozygotic (33%) and then homozygotic dominant (23%).

On the other hand, when comparing Rvi6 × Rvi6 and Rvi5 × Rvi6 crosses, progeny with higher resistance were produced in cases where both parents had the Rvi6 gene—21% vfvf, 39% VFvf, 40% VFVF. In the case of crossing Rvi5 and Rvi6 parents, Rvi5 (Vm) had a high allele frequency in the progeny at 63%. The effectivity of gene pyramiding was observed in the case of Rvi6 × Rvi6 crosses; however, gene pyramiding was not observed in the Rvi6 × Rvi5 seedling population.

When comparing data of allele frequencies with field observations, equal patterns were not observed between populations. Populations 1 and 2 had similar levels of apple scab severity but their allele frequency data differed. Population 3 had the highest apple scab severity with the lowest homozygotic dominant Rvi6 genes, however, population six, which had similarly low homozygotic dominant Rvi6 genes, had the lowest apple scab severity. Comparing populations based on parent plants showed that the seedling population where one parent was susceptible had the lowest apple scab severity. Populations with pyramided genes (populations 3 and 6) showed significantly different field resistance evaluation results. Moreover, population 3 has the highest scab severity among all studied populations, while population 6 has the lowest one with the best resistance. The highest resistance in the field was observed in the Rvi6 × Rvi6 population, even though the Rvi6 × Rvi5 population had two resistance genes.

3.3. Genetic Relationship of Parent Plants

PCA analysis was performed based on the genotyping of parent plants with 22 SSR markers (Figure 6) to evaluate the relationship of the plants used for breeding hybrid seedlings. The biplot shows four groups, starting with the cultivar ‘Signe Tillisch’, which is furthest away from the rest; DI-1-92-32, ‘Pervinka’ and ‘Rewena’ group together in the second group, whereas ‘Kandil Orlovskij’ and ‘Zane’ are part of the third group. The fourth group contains most cultivars: ‘Arbat’, ‘Dayton’, ‘Florina’, ‘Kurnakovskoje’ and ‘Scarlett O’Hara’.

A second PCA analysis was performed in order to determine genetic relationships between the parent cultivars based on fragments that identify apple scab resistance genes (Figure 7). It was observed that ‘Kandil Orlovskij’, ‘Dayton’ and ‘Scarlett O’Hara’ had identical resistance genes (Rvi6, Rvi11, Rvi15), evidenced by them grouping together in the biplot (Table S1). The remaining cultivars in the cluster shared some genes: ‘Signe Tillisch’ and ‘Florina’ both had Rvi11, whereas ‘Signe Tillisch’ also shared the gene Rvi15 with ‘Kandil Orlovskij’ and the rest of the identical cultivars, whereas ‘Florina’ shared Rvi6 with the group. The two Rvi5 donors, ‘Pervinka’ and D-1-92-32 share the same parent, SR 0523, which too was a Rvi5 donor, which is why the two cultivars clustered together and were opposite from the rest of the cultivars, which did not possess the Rvi5 gene. ‘Rewena’ did not cluster together with any other cultivar, it also had the highest number of resistance genes: Rvi2, Rvi4, Rvi6, Rvi11, Rvi15. ‘Kurnakovskoje’, ‘Zane’ and ‘Arbat’ formed a unique cluster, as ‘Zane’ is a susceptible cultivar and only has one resistance gene (Rvi15), whereas ‘Kurnakovskoje’ shares Rvi2 and Rvi4 with ‘Rewena’, while also sharing Rvi6 with ‘Arbat’.

4. Discussion

4.1. Relationship between Field Data and Marker Data

The most modern cultivars resistant to apple scab are derivatives of crosses with M. × floribunda 821, however, there have been documented cases of V. inaequalis overcoming apple scab resistance [36,37]. This created a necessity for breeding programs that investigated different approaches to establishing long-term apple scab resistance. More than 20 apple scab resistance genes have been identified [11], a database for marker assisted breeding has been developed [22], projects that included pyramiding multiple apple scab resistance genes [38] and quantitative trait loci associated with apple scab [39] have been carried out. Despite there being a robust amount of data and materials for marker assisted breeding and molecular analysis, there is limited data that evaluates how introducing new combinations of R genes into an apple population affects resistance in the field, which this study aims to remedy.

Data of field observations showed that the best tree health and lowest disease severity for apple scab were observed in 2018 (Figure 1); however, it is important to note that tree health changed very minimally and was good, evaluated at 2 points out of a 9-point scale. Correlation analysis showed a weak positive correlation between apple scab severity and tree health, meaning that the lower the health of the tree, the higher the severity of apple scab, which is a positive indicator of tree health being a valid predictor of a cultivar’s ability to resist apple scab infection.

Apple scab, despite being an issue worldwide [7], proliferates best in the temperate climate, which is cool and humid [13]. The main environmental factors that promote primary apple scab infection are air temperature, precipitation and humidity [7]. There were statistically significant differences between average temperature and precipitation between all three years of study (Figure S1), however, changes in weather conditions did not neatly align with changes in disease severity. The year with the highest apple scab severity was 2017, however, precipitation was second lowest that year and average air temperature was the lowest out of the three years. It is also important to note that correlation analysis between apple scab severity and environmental factors all showed a very weak correlation. This is most likely due to the factor that the margin of optimal weather conditions for the development of V. inequalis is wide enough that environmental factors show a low influence on apple scab severity.

To assess the genetic aspects of apple scab resistance, phenotypical data (tree health, apple scab severity) was compared between hybrid seedling populations (Figure 2). Apple scab phenotype can vary even in cultivars that have known resistance genes, the variance is affected by modifier genes of the main resistance genes, for example, quantitative trait loci (QTL) [40], as well as by the aggressiveness of the pathogen race [41]. When comparing populations, there were statistically significant differences between all populations based on phenotypic data (tree health and apple scab severity).

Differences in phenotypic data could not be reliably explained by environmental factors, however, the current data shows a relationship between the genetic makeup of the hybrid seedlings and the phenotypic data. Populations 6 and 3 had the best and worst average tree health scores, respectively, and it is known that the origin of a cultivar affects its ability to adapt to the climate its grown in [25]. Seedlings from both populations had one parent plant which originated from the USA [42] which would mean the plants are more sensitive to the climate of Latvia. However, the seedlings of population 6 were the offspring of D-1-92-32 (SR 0523 × /’Lobo’ × ‘Iedzenu’/), which is a hybrid developed in Latvia and is well suited for the local climate; this therefore explains the different tree health outcomes. The importance of the ability of a tree to adapt to a changing environment could influence discrepancies in other genetic aspects; population 3 inherited the most apple scab resistance genes but had the highest average apple scab severity, whereas seedlings that had a susceptible parent showed the lowest average apple scab severity (Figure 2 and Figure 3). This could mean that despite resistance genes being an important aspect of apple breeding, other factors should not be overlooked, such as the pedigree of the plant material.

Interpretation of the results was made more difficult by the fact that there are not many studies that combine field data and molecular analysis. In most cases, phenotyping happens in greenhouses in a controlled environment, infecting plants with a pathogen suspension and then genotyping the affected plants [43], which can affect the performance of both the plant and the pathogen. These studies are valuable; however, they leave out information about the interaction between the plant and its natural environment.

4.2. Analysis of Parent Plants

SSRs markers were used to genotype 12 parent genotypes (Table S2), and even though two markers, EB14743a and CN493139, had relatively lower genetic diversity scores, the rest of the markers showed high values in heterozygosity, gene diversity and discrimination power, meaning that the chosen markers were suitable for analysis of the plant material. Cluster analysis revealed that the grouping of some parent plants could be related to their geographic origins, for example, ‘Scarlett O’Hara’ and ‘Dayton’ originate from the USA [42]. Both cultivars showed genetic similarities according to 22 SSR markers (Figure 6) and to the Rvi gene specific markers (Figure 7). However, it does not completely explain the huge disparities in field evaluation data between the offspring of either cultivar in populations 3 and 6. It could be argued that genotypes in each population possess modifier genes which lead to different results in phenotypical evaluation, despite their common ancestor SR 0523 (parent of both ‘Pervinka’ and D-1-92-32).

Cultivars meant for Scandinavia and the Baltics must be adapted to a specific climate; therefore, a lot of widely grown cultivars are not suitable for commercial growing. However, a selection of specifically adapted cultivars have been developed over the years [25]. PCA analysis based on 22 SSR markers revealed that parent genotypes split into three groups (Figure 6): cultivars ‘Pervinka’, ‘Rewena’, ‘Kandil Orlovskij’, ‘Zane’ and DI-1-92-32 grouped together in one side of the plot, and they originated from Latvia, Russia and Germany. Opposite of group one was a group of genotypes from the USA (‘Scarlett O’Hara, ‘Dayton’), France (‘Florina’), and Russia (‘Kurnakovskoje’, ‘Arbat’). One cultivar stands out— ‘Signe Tillisch’—which comes from Denmark and is situated separate from the two groups. As Russian cultivars are spread out evenly in the plot, geographic origin does not seem to be the defining factor for grouping the different parent genotypes. Many of the parent genotypes have M. × floribunda 821 as an ancestor, which too could have affected their genetic similarity and shows similar patterns of apple breeding. ‘Signe Tillisch’, on the other hand, is an old cultivar with unknown pedigree, thus its difference from the rest of the cultivars cannot be fully explained by pedigree data; however, this could be an indication how historical and modern breeding practices do result in genetically different cultivars.

4.3. Effects of Gene Pyramiding

To understand the inheritance patterns of the Rvi6 and Rvi5 gene, genotyping was performed on the hybrid material. There is no doubt that Rvi6 is inherited as a dominant monogenic trait. It has been proven by several studies in which a heterozygotic parent is crossed with a susceptible parent, which results in a 1:1 (resistant/susceptible) ratio and this effect is observed in following generations [40]. This was observed in our study as well, as in a cross between a susceptible cultivar and a cultivar carrying a heterozygotic Rvi6 gene, the gene was inherited at a 1:1 ratio, whereas by crossing two Rvi6 carriers the offspring inherited the genes at a 3:1 ratio. Thus, the basics of the apple scab resistance gene function is understood, but resistance genes can be affected by modifier genes. These genes cannot be identified, as they manifest only in combination with the main resistance gene, similarly to QTL. These genes cause variation in phenotype and can manifest as susceptibility, regardless of the presence of the R gene [40]. This can explain the discrepancies of seedlings of susceptible plants showing better apple scab resistance in the field. Despite that Rvi6 alleles in a Rvi6 × susceptible cross were inherited at a 1:1 ratio and at a 3:1 ratio in a Rvi6 × Rvi6 cross, lower average disease severity scores were found in the populations that had a susceptible parent (Figure 3). This could suggest that other R genes, QTL or a modifier gene is affecting expression of resistance of apple trees. Such an effect in the cross Rvi6 × Rvi6 could also have been caused by inbreeding depression for this gene, although the parents of these populations are genetically quite distinct (Figure 6). The cultivar ‘Arbat’ carries the gene Rvi13 which could not be found in other Rvi6 donors, leading to believe that it affects the functions of the Rvi6 gene. Rvi5 has a high heritability of 63% of all samples inherited the gene, however, the alleles of the gene were not investigated deeper. It is also important to note that the weakest of the hybrid material was removed after the first year after crossing and only potentially resistant seedlings were planted in the experimental field. Such a selection could have affected the data of Rvi6 and Rvi5 heritability but had no effect on the relationship between marker data and field evaluations.

The effect of gene pyramiding was assessed by comparing field evaluation data and genotyping data; in the case of population 3 and 6, gene pyramiding was not effective, as population 3 had the highest apple scab severity, but population 6 had the lowest severity (Figure 2). There is literature on how, in some cases of plant-pathogen interactions, gene pyramiding may not have the desired resistance outcomes [44]. A better outcome can be reached by combining different resistance genes, as R genes are functionally similar [23]. This has been shown by pyramiding R genes with QTL, which results in long lasting resistance [44], whereas only selecting QTL genes results in worse resistance outcomes, as QTL genes have low heritability in comparison to R genes [41]. Gene pyramiding can be improved by studying the genetic relationships of R genes and QTL, several studies have noted their co-localisation [45,46]. Relationships between genes have been established by introducing wild plants into breeding pools; they tend to introduce genes that at first seem of low value but due to their interaction with other genes become important in ensuring resistance to the pathogen. Studies have shown that resistance genes that have strong connections and belong to one linkage group are inherited in their offspring as pyramided genes [14].

Studies of molecular mechanisms and genes are just as necessary; two resistance genes have been cloned thus far: Rvi6 and Rvi15 [47]. This has helped us to understand the molecular makeup of the genes as well as partially determine their functions [48], however, other resistance genes have yet to be studied. The parent plants of the hybrid seedlings used in the study were genotyped to identify resistance genes in the population and figure out the differences of the Rvi6 and Rvi5 genes. Several genes belong to the linkage group LG2 (Rvi4, Rvi2, Rvi15, Rvi11) and were inherited all at once as pyramided genes, which explains their presence together in different parent genotypes (Table S1). A study has shown that most QTL involved in resistance are mapped into linkage groups LG2, LG1 and LG17 [41]. It can be said that apple scab resistance genes are closely related to the LG2 linkage group, however, more studies are needed to confirm it. Several QTL have been observed to be of the LG1 linkage group [41], where Rvi6 has been mapped, as well as the linkage group LG17 where Rvi5 has been mapped which proves the relationship of R genes and QTL and their link. The link between both genes has also been confirmed by studies of Rvi5 where offspring express resistance differently and this phenotypical difference was caused by genes that were similar to Rvi6 [12].

To determine the most suitable genes for apple scab resistance, the VINQUEST project has collected data over a ten-year period [23] and concluded that the most promising genes for durable apple scab resistance breeding are Rvi5, Rvi11, Rvi12, Rvi14, and Rvi15. The researchers also note that genes Rvi2, Rvi4, Rvi6, Rvi7, Rvi9 and Rvi13 are useful for breeding, however, they are most effective in pyramids of at least 3 genes.

Our data showed an interesting relation to the results of the VINQUEST project; the population with the lowest scab resistance, population 3, had four of the five recommended genes, however, another population (population 6) which had the same number of recommended genes, had the highest average scab resistance (Figure 2). Both populations had a parent who was a Rvi6 donor (‘Dayton’ and ‘Scarlett O’Hara’) and a Rvi5 donor (‘Pervinka’ and D-1-92-32) (Table S1). Population 3 also had the lowest average tree health (Figure 2) and population 6 had a parent, D-1-92-32, which is a hybrid specifically bred for the climate of Latvia (Table S1), showing that even if the recommended genes are present, other factors such as tree health and adaptation to the local climate can affect resistance. The rest of the parent plants had either one or two of the recommended genes, Rvi11 and Rvi15, which shows that despite other factors influencing resistance to apple scab, having a certain diversity of R genes is still valuable, therefore, it is advisable to follow the recommendations of the VINQUEST project when planning a breeding programme.

5. Conclusions

• Overall, a combination of field evaluation and molecular studies offers a more measured look into how resistance plays out in a tree’s natural environment and is a viable path for future research.
• Resistance to apple scab is influenced by the overall health of the tree, genetic factors such as the pedigree of the cultivar, genetic composition of parent cultivars, incl. the presence of various resistance genes, whereas the environment alters the way resistance manifests.
• When Rvi6 and Rvi5 are pyramided, they are inherited as separate genes and inheriting both is not the only factor for apple scab resistance since in the field, several factors are at play and can significantly modify resistance outcomes.