Growth Characteristics of Five Plum Varieties on Six Different Rootstocks Grown in Containers at Different Irrigation Levels

Abstract

In contrast to other European countries, the nurseries in Hungary mainly use Myrobalan seedlings as rootstock for plum trees; no experience with other rootstocks is available yet. The aim of our research was to gather experience and recommend rootstocks to growers that are adapted to the climatic conditions and can be used in high-density intensive orchards. In the experiment, six rootstocks (Myrobalan seedlings, ‘St. Julien GF655/2’, ‘St. Julien A’, ‘Fereley’, ‘Wangenheim’ and ‘WaVit’) and five varieties (‘Topper’, ‘Toptaste’, ‘Topfive’, ‘Čačanska lepotica’ and ‘Jojo’) were tested. The experimental plantation was established in Kecskemét, in the lowland region of Hungary, which is characterised by low humus content and dry climatic conditions. The trees received two different irrigation doses from 2012 onwards, one for each half of the trees (six trees) of a combination. The first irrigation dose was 2 L and the second was 4 L per hour. Trees with the highest trunk cross-sectional area (TCSA) and canopy volume were ‘Topper’, ‘Toptaste’ and ‘Topfive’ on the ‘Fereley’ rootstock, but this rootstock showed lower survival, with only 33–50% of trees surviving the 2× dose irrigation. The average TCSA values of the rootstocks showed that ‘Wangenheim’ and ‘WaVit’ rootstocks resulted in too poor growth. In contrast, the Myrobalan seedling (Sdlg) and ‘St. Julien GF655/2’ provided strong growth to the trees based on the TCSA and canopy volume data. The results of the analysis of shoot growth showed that irrigation did not significantly affect shoot length, except for ‘Jojo’. The plum trees developed fruit trees typical of the cultivars after fruiting, with little influence of the rootstocks. ‘Topper’ and ‘Čačanska lepotica’ were dominated by bunched and bunched shoots, while ‘Topfive’ was dominated by short shoots and ‘Toptaste’ developed the highest proportion of long shoots.

1. Introduction

In Hungary, plum trees are predominantly grown on Myrobalan rootstocks [1,2], as the trees are grown on open canopies in mechanically harvested orchards. However, growers producing for the fresh market are increasingly planting their trees on less vigorous rootstocks, allowing the introduction of intensive nursery systems and cultivation techniques [3,4].

Abroad, the use of rootstocks is much more varied, taking into account their origin and vigour. However, the parentage and pedigree are rather complex [5,6]. Plum rootstocks are usually classified as (I) Myrobalan and its hybrids, (II) Prunus insititia (Saint Julien), (III) Prunus domestica and (IV) hybrids of different Prunus species [7]. Their adaptability to different soils and climates is quite variable, which underlines the importance of evaluating rootstocks. Although the authors more or less agree on the classification of rootstocks according to vigour [8,9,10,11,12], some literature data suggest that the scion and groundstock combination may modify the effect of rootstock on scion growth [9,13,14,15,16,17,18]. Furthermore, the type of cultivation and pruning applied may affect growth; in particular, the cultivation of high-density orchards may reduce growth differences between rootstocks [2,3,14,15]. Similarly, soil, climate and ecological conditions [15] are very important characteristics to select rootstocks.

The seedling Myrobalan (Prunus cerasifera Ehrh. var. cerasifera Scheid. cv. myrobalana, 2n:16, 2n:24 in hybrids) is a common planted rootstock in eastern and south-eastern European countries [7,8,9,13,19,20,21], but in western European countries, vegetatively propagated varieties or hybrids (‘Myrobalan B’, ‘Marianna GF 8-1’, ‘Myrobalan 29C’) are more popular. Myrobalan sdlg and the clonal rootstocks ‘Myrobalan’ are vigorous; the main disadvantage of the seedlings is the intensive budding. Recent breeding projects [17,22,23,24,25] have targeted clonally propagated hybrids that are less vigorous and resistant to soil-borne fungal diseases and Nematodes. The East Asian plum salicina (Prunus salicina Lindl. 2n:16), closely related to Myrobalan and used in several breeding projects, and the hybrid ‘Fereley’ (P. salicina x P. spinosa) are reported to be vigorous to moderately vigorous plum rootstocks.

The St. Julien plum (Prunus insititia Jusl.; 2n:40,48) is one of the most important plum rootstocks in western European countries, from vigorous (‘St. Julien A’) to moderately or weakly vigorous rootstocks (‘St. Julien GF 655/2′, ‘Pixy’) being the most important plum rootstocks in western European countries [4,8,9,20,23,26,27]. It is probably a natural hybrid derived from Prunus cerasifera and Prunus spinosa species in many parts of the region [5,6]. ‘St. Julien GF 655/2’ has been shown to be moderately vigorous and productive compared with Myrobalan Sdlg and ‘Marianna GF 8-1’ [1,7,21,28].

The European plum (Prunus domestica L.; 2n:40,48) is generally described as a semi-dwarf or dwarf rootstock [7,20,28,29]. The old German rootstock ‘Wangenheim’, widely planted in Germany and Poland and known as the dwarf rootstock, is propagated as seedling and by hardwood grafting [9,29,30,31,32]. More recently, its selected clone is propagated in nurseries under the name ‘WaVit’.

Taking into account the interest of growers in planting new varieties on several imported rootstocks, which have not been previously tested in Hungarian lowlands on light sandy soils, we decided to conduct rootstock trials with the new varieties (‘Topper’, ‘Topfive’, ‘Toptaste’, ‘Jojo’ and ‘Čačanska lepotica’) and rootstocks (Myrobalan sdlg. ‘Myrobalan’, ‘St. Julien A’ as control, ‘St. Julien GF 655/2’, ‘Fereley’, ‘Wangenheim’ and ‘WaVit’). Based on the literature results, ‘Myrobalan’ and ‘St. Julien A’ are described as vigorous, ‘St. Julien GF 655/2’ and ‘Fereley’ as medium vigorous, and ‘Wangenheim’ and ‘WaVit’ as dwarf plum rootstocks [1,33,34,35,36,37].

The rootstock cultivars ‘Topper’, ‘Toptaste’ and ‘Topfive’ from Geisenheim, Germany, are the results of a breeding programme [28] to improve PPV resistance, disease, pest and climatic resistance and to improve the ability to withstand the relevant resistance. ‘Jojo’ also originates from Hohenheim, Germany, and is a cross of ‘Ortenauer’ and ‘Stanley’. ‘Čačanska lepotica’ is a Serbian breeding project, a disease-resistant variety that is widely planted in Hungary.

In this paper, we presented the results of the studies on vigour and shoot growth characteristics of the cultivars planted on different rootstocks, including their suitability for high-density intensive orchards and their adaptation to the Hungarian climate. This paper presented the results on tree growth and fruit tree development in the first 7 years.

2. Materials and Methods

The experimental orchard was established in April 2010 in Kecskemét, Great Plain of Hungary (Figure 1). Kecskemét is located on the sand ridge between the Danube and Tisza; the soil is sandy formed from the Danube alluvium, characterized by a low humus content. The climate is a warm-dry continental, and the average annual temperature is 11 °C. The average temperature in January is −1 °C and in July it is 21 °C, while the average temperature in the growing season is 18 °C. The number of heat days is 16–20 days. The mean annual sunshine hours are 2040. The average annual rainfall is 550 mm, but in 2012, the rainfall did not exceed 400 mm [29]. The meteorological data refer to the average of a 50-year data set.

One-year-old feathered trees were planted, purchased from Germany. The variety and rootstock combinations are presented in

Table 1. Variety and rootstock combination planted in experimental orchard.

	‘Myrobalan’	‘St. Julien GF655/2′	‘St. Julien A’	‘Fereley’	‘Wangenheim’	‘WaVit’
‘Topper’	✓	✓		✓
‘Toptaste’	✓	✓	✓	✓	✓
’Topfive’		✓	✓	✓		✓
‘C. lepotica’	✓		✓
‘Jojo’	✓		✓

. Not all the varieties were planted on each rootstock. Not all rootstock–variety combinations were available at nurseries. Because they were not available, they were not included in the experiment.

Unpruned trees were installed in a spacing of 2.5 × 1.5 m, trained to slender spindle; laterals were bent close to the horizontal (Figure 2), and the vertical growth of the trees was stopped by heading at a height of 3.2 m.

The criteria for analysing the rootstock and variety combinations were the trunk cross-sectional areas (TCSAs), from which we can infer the growth rate. In addition, the canopy volume and height of the trees were also measured. These data show the growth characteristics. These growth characteristics are necessary for intensive cultivation technology.

In the experimental orchard of each scion and rootstock combination, 12 trees were planted in a random block design. The trees received two different irrigation dosages from 2012, each on half of the trees (6 trees) of one combination. The trees were planted in 170-L containers (0.6 m diameter and 0.6 depth) sunk into the soil to avoid affecting the water supply from the neighbouring differently irrigated trees. The containers were drilled at the bottom to drain out the surplus irrigation water; in this way, no stagnant water at the bottom of the container was left. The trees received two different water-supply standards by using drip irrigation. The different irrigation regime started in 2012; every second tree received 2 L and the others 4 L per hour doses Two tensiometers were installed in the field, inserted into the 20 cm depth of the soil. The soil moisture content was measured 2–3 times a week with SMM1 equipment in randomly selected containers. The irrigation was controlled on a soil moisture basis, and watering started when the tensiometer and soil moisture metre showed a moisture content of 26–50% in the container of the sample trees.

In the experimental orchard of each scion and rootstock combination, 12 trees were planted in a random block design. The trees received two different irrigation dosages from 2012, each on half of the trees (6 trees) of one combination. The trees were planted in 170-L containers (0.6 m diameter and 0.6 depth) sunk into the soil to avoid affecting the water supply from the neighbouring differently irrigated trees. The containers were drilled at the bottom to drain out the surplus irrigation water; in this way, no stagnant water at the bottom of the container was left. The trees received two different water-supply standards by using drip irrigation. The different irrigation regime started in 2012; every second tree received 2 L and the others 4 L per hour doses Two tensiometers were installed in the field, inserted into the 20 cm depth of the soil. The soil moisture content was measured 2–3 times a week with SMM1 equipment in randomly selected containers. The irrigation was controlled on a soil moisture basis, and watering started when the tensiometer and soil moisture metre showed a moisture content of 26–50% in the container of the sample trees.

The measurements were carried out from 2010 to 2016 (

Table 2. Measured and counted parameters of trees during investigation.

	2010	2011	2012	2013	2014	2015	2016
The trees survival rate			✓				✓
Trunk cross-sectional area	✓	✓	✓	✓	✓	✓	✓
Canopy volume							✓
The height of the trees				✓
Shoots				✓
Fruiting laterals				✓

). The trunk diameter was measured with a calliper at a height of 40 cm, from which we calculated the trunk cross-sectional area (TCSA). We measured the diameter of the crown with a tape measure and the height of the tree and trunk with a measuring stick, and we calculated the volume of the crown from these values. The length of the shoots was measured with a tape measure. The length of the shorter fruiting laterals was measured with a calliper and that of the longer ones with a measuring tape.

Soil analysis was performed in the year prior to planting (2009) to determine the nutrient content of the soil (

Table 3. Soil Properties (2009).

	Sandy Soil
pH-KCl [-]	8.01
Soil cohesion index (KA) *	28
All water soluble salts [m/m%]	<0.02
CaCO3 [m/m%]	2.5
Humus [m/m%]	0.67
NO3−-N+NO2−-N [mg/kg]	<1
P2O5 [mg/kg]	344
K2O [mg/kg]	61

* [30].

). Based on this, organic manure was applied to the soil before planting to increase the organic matter content of the soil. Then, in later years, K overweight complex fertilizer was added to the containers.

In the interests of sustainable fruit production, we applied irrigation in order to have a positive effect on the growth of the trees [31]. As the quality of irrigation water can affect the tree growth [32,33], we investigated the quality of water. The garden irrigation water supply contains the following ingredients and concentrations: Ca²⁺ 171 mg/L; Mg²⁺ 40 mg/L; Na⁺ 12.3 mg/L; K⁺ 0.43 mg/L; NH⁴⁺ 0.38 mg/L; NO³⁻ 0.88 mg/L; PO₄ ³⁻ 0.175mg/L; and Fe 5630 µg/L. Nutrients dissolved in the irrigation water can also have a positive effect on growth and yield [33]. The mineral content of the irrigation water is related to depth. The water in the deeper wells has lower Ca and Mg content, but Ca is dominant over Mg [32], which may affect the vegetative and generative development of the trees.

Statistical Methods

In the data analysis, we applied two-way analysis of variance (ANOVA), where both factors were considered. One-factor analysis of variance or two-sample t-test was applied depending on whether we had to compare the properties of at least three or two groups. At hypothesis testing procedures, the null hypothesis at a given level of significance was considered, that is, the equality of the mean value of the examined property in each case. Two conditions were examined: the normality test was performed by the Kolmogorov–Smirnov and/or Wilk–Shapiro test, while the standard deviation was analysed by the Levene test. On data meeting the requirements, analysis of variance or t-test was performed. If the conditions of the procedures were violated, the non-parametric equivalent of the procedures was used, i.e., the Welch and Brown-Forsythe test for ANOVA and the Mann–Whitney test for two-sample t-tests. Finally, if the null hypothesis was rejected as a result of the procedure used, i.e., we found a significant difference in the mean value of the given property in each group, then we created homogeneous classes using Duncan’s multiple range test (and denoted them by b, c). [34]. The significance level in the analysis was 0.05. All data were statistically tested by using SPSS software (IBM SPSS Statistics 27.0, Armonk, New York, NY, USA).

3. Results

All the trees of the rootstock and scion combination developed well; however, in terms of the survival rate, trunk cross-sectional area (TCSA), shoot growth and fruiting wood development, we found significant differences between the trees on different rootstocks. The results of the statistical analysis did not prove any significant effect of the water dosage except for the TCSA increment and total shoot length of ‘Čačanska lepotica’ and ‘Jojo’ trees.

3.1. Survival Rate of Trees

The trees of the five varieties survived the first 7 years at a high rate, 83% to 100%, on all rootstocks except for ‘Fereley’ receiving a double amount of irrigation (

Table 4. Tree number (columns 2012 and 2016) and survival rate at different water dosage (W1 and W2).

Rootstock	’Topper’				’Toptaste’				’Topfive’
	2012	2016	W1%	W2%	2012	2016	W1%	W2%	2012	2016	W1%	W2%
’Myrobalan’	6 + 6	6 + 6	100	100	6 + 6	6 + 5	100	83	-	-	-	-
’GF 655/2′	6 + 6	6 + 6	100	100	6 + 6	6 + 6	100	100	6 + 6	6 + 5	100	83
’St. Julian A’	-	-		-	6 + 6	6 + 5	100	83	6 + 6	6 + 6	100	100
’Fereley’	6 + 6	5 + 3	83	50	6 + 6	6 + 2	100	33	6 + 6	6 + 2	100	33
’Wangenheim’	-	-		-	6 + 6	6 + 5	100	83	-	-	-	-
’WaVit’	-	-		-	-	-	-	-	5 + 6	5 + 4	83	67
	’Čačanska lepotica’				’Jojo’
	2012	2016	W1%	W2%	2012	2016	W1%	W2%
’Myrobalan’	6 + 6	6 + 6	100	100	6 + 6	6 + 6	100	100
’St. Julien A’	6 + 6	5 + 6	83	100	6 + 6	6 + 6	100	100

). The trees of the varieties of ’Topper’, ’Toptaste’ and ’Topfive’ on the ’Fereley’ rootstock showed considerably low survival rates, 33–50%, when they received a double amount of irrigation. Tree death happened to the ‘Toptaste’ and ‘Topfive’ trees after 2012 when the different irrigation regime was running. The ‘Topfive’ trees on the ‘WaVit’ rootstock also showed a lower survival rate, 67%.

3.2. Tree Size in 2016 Presented in TCSA, CV and Tree Height

The results of the statistical analysis proved the significant effect of rootstocks on the trunk cross-sectional areas (

Table 5. Means of trunk cross-sectional area (TCSA, cm²) and canopy volume (CV, m³) in 2016.

Rootstock	‘Topper’		‘Toptaste’		‘Topfive’
	TCSA	CV	TCSA	CV	TCSA	CV
‘Myrobalan’	18.34 b	1.22 a	10.59 ab	0.74 ab	- -	- -
‘GF 655/2′	17.51 b	1.15 a	11.71 b	0.8 b	12.51 b	0.67 a
‘St. Julian A’	- -	- -	11.81 b	0.92 bc	13.41 b	0.51 a
‘Fereley’	11.16 a	0.93 a	15.47 c	1.06 c	14.01 b	0.87 b
‘Wangenheim’	- -	- -	8.64 a	0.56 a	- -	- -
‘WaVit’	- -	- -	- -	- -	7.12 a	0.63 a
	‘Čačanska lepotica’		‘Jojo’		Mean of rootstock
	TCSA	CV	TCSA	CV	TCSA	CV
‘Myrobalan’	17.36 a	1.57 a	19.27 a	1.60 a	16.51 b	1.30 d
‘GF655/2′	- -	- -	- -	- -	13.95 b	0.88 bc
‘St. Julien A’	19.32 a	1.63 a	20.16 a	1.44 a	16.20 b	1.12 cd
‘Fereley’	- -	- -	- -	- -	13.55 b	0.95 c
‘Wangenheim’	- -	- -	- -	- -	8.64 a	0.56 a
‘WaVit’	- -	- -	- -	- -	7.12 a	0.63 ab

Note: Means are separated by Duncan’s multiple range test; different letters following the data stand for significant differences at p = 0.05.

) of the ’Topper’, ’Toptaste’ and ’Topfive’ trees on different rootstocks in 2016, and the canopy volume was also significantly different for ‘Toptaste’ and ‘Topfive’, while the ‘Čačanska lepotica’ and ‘Jojo’ trees did not differ significantly in either the trunk cross-sectional area or canopy volume. The trees of ’Topper’, ’Toptaste’ and ’Topfive’ differently performed considering their TCSA. The largest TCSA and CV reached the ‘Toptaste’ and ‘Topfive’ trees on the ‘Fereley’ rootstock, while the ‘Topper’ trunk was the thickest on the ‘Myrobalan’ seedling. The trees of ‘Topper’ did not show significant differences in the CV. The TCSA and CV of ‘Toptaste’ on ‘Wangenheim’ and ‘Topfive’ on ‘WaVit’ were significantly smaller than those on ‘Fereley’.

The mean TCSA results (Table 5) also showed that the cultivar ‘Čačanska lepotica’ on ’Myrobalan sdlg’ gave similar tree sizes to the ’Topper’ and ’Myrobalan’ combination without significant differences. The ’Myrobalan’ and ’St. Julien A’-rooted ‘Jojo’ trees had an overall strong growth, while ‘Toptaste’ resulted nearly half the size of the trees on the same rootstocks. The development of the TCSA and canopy volume (CV) in ‘Topfive’ was almost similar. Examining the tree height as a variety (

Table 6. Height of trees in 2013 on different rootstocks (m).

Rootstocks	‘Topper’	‘Toptaste’	‘Topfive’	‘Čačanska lepotica’	‘Jojo’
‘Myrobalan’	2.68 b	2.63 a	- -	2.82 a	2.68 a
‘GF655/2′	2.76 b	2.72 a	2.71 a	- -	- -
‘St. Julien A’	- -	2.65 a	2.66 a	2.77 a	2.73 a
‘Fereley’	2.54 a	2.67 a	2.65 a	- -	- -
‘Wangenheim’	- -	2.64 a	- -	- -	- -
‘WaVit’	- -	- -	2.61 a	- -	- -
Mean of varieties	2.66	2.66	2.66	2.80	2.71

Note: Means are separated by Duncan’s multiple range test; different letters following the data stand for significant differences at p = 0.05.

), the results of the statistical analysis did not show a significant effect on rootstocks, except for the ‘Topper’ and ‘Fereley’ combination, which produced a smaller height in 2013 than the other combinations. The ‘Topper’ cultivar showed a significant difference according to the ‘Myrobalan’ and ‘St. Julien GF655/2′ rootstocks.

All in all, the different TCSA results showed that the mean trunk thickness was similar on ‘Myrobalan’, ‘St. Julien GF 655/2′, ’St. Julien A’ and ’Fereley’, while on the ‘Wangenheim’ and ’WaVit’ rootstocks, the TCSA was significantly smaller (Table 5).

Considering the fact that the growth rates of non-productive (2010–2012) and productive (2013–2016) years may differ from each other (

Table 7. Increment of TCSA (cm²) in two stages of tree development (2010–2012 and 2013–2016).

Rootstocks	‘Topper’		‘Toptaste’		‘Topfive’
	2010–2012	2013–2016	2010–2012	2013–2016	2010–2012	2013–2016
‘Myrobalan’	4.04 b	9.76 b	2.53 a	4.30 ab	- -	- -
‘GF655/2′	3.21 ab	9.88 b	2.89 a	4.57 b	2.61 ab	5.44 b
‘St. Julien A.’	- -	- -	2.57 a	3.89 ab	3.06 b	6.59 bc
‘Fereley’	2.24 a	5.03 a	3.03 a	9.02 c	2.97 b	7.89 c
‘Wangenheim’	- -	- -	2.51 a	2.09 a	- -	- -
‘WaVit’	- -	- -	- -	- -	2.24 a	2.75 a
Mean variety	3.16	8.22	2.71	4.77	2.2	5.67
	‘Čačanska lepotica’		‘Jojo’		Mean rootstocks
	2010–2012	2013–2016	2010–2012	2013–2016	2010–2012	2013–2016
‘Myrobalan’	3.17 a	10.22 a	3.84 a	12.35 a	3.40	9.16
‘GF655/2′	- -	- -	- -	- -	2.90	6.63
‘St. Julien A’	4.22 a	10.58 a	3.18 a	12.39 a	3.26	8.36
‘Fereley’	- -	- -	- -	- -	2.75	7.31
‘Wangenheim’	- -	- -	- -	- -	2.51	2.09
‘WaVit’	- -	- -	- -	- -	2.24	2.75
Mean variety	3.70	10.4	3.51	12.37

Note: Means are separated by Duncan’s multiple range test; different letters following the data stand for significant differences at p = 0.05.

), we examined the increases in the trunk cross-sectional area of the two periods. While the growth of ‘Čačanska lepotica’ and ‘Jojo’ was balanced, the ’Topper’ variety resulted a smaller trunk cross-sectional-area increase in the first stage compared with the ‘Myrobalan’ sdlg. rootstock and in both stages compared with the ‘Myrobalan’ sdlg. and ‘St. Julien GF 655/2′ rootstocks. In contrast, the increase in the trunk cross-sectional area of the ‘Toptaste’ and ‘Topfive’ trees in the productive phase (2013–2016) exceeded the TCSA growth measured in others on the ‘Fereley’ rootstocks by a significant difference. We measured the lowest TCSA growth in the ‘Wangenheim’ and ‘WaVit’ rootstocks at both stages.

3.3. Performance of Shoot Growth in 2013

The results of the analysis of the mean shoot length and total shoot growth per tree showed that irrigation did not significantly affect these parameters, except for the ‘Jojo’ variety. In the case of ’Jojo’, the single (W1) and double watering (W2) resulted in a significant difference in the total shoot growth of the trees. The average total shoot growth was 1478 cm with single irrigation and 2244 cm with double irrigation. Table 5 and Table 6 show the mean for the rootstocks. There was a significant difference in the mean shoot length measured on the sample branch (

Table 8. Mean shoot length of trees in 2013 (cm).

	’Topper’	’Toptaste’	’Topfive’	’Čačanska lepotica’	’Jojo’
’Myrobalan’	153.33 a	99.75 a	- -	125.92 a	167.92 a
’GF655/2′	142.83 a	84.67 a	100.18 a	- -	- -
’St. Julien A’	- -	99.58 a	119.42 a	164.17 a	172.25 a
’Fereley’	158.42 a	157.63 b	143.10 a	- -	- -
’Wangenheim’	- -	70.25 a	- -	- -	- -
’WaVit’	- -	- -	135.82 a	- -	- -

Note: Means are separated by Duncan’s multiple range test; different letters following the data stand for significant differences at p = 0.05.

) on ‘Toptaste’ and ‘Fereley’, while there was a significant difference in the total shoot length (

Table 9. Total shoot length of trees in 2013 (cm).

	’Topper’	’Toptaste’	’Topfive’	’Čačanska lepotica’	’Jojo’
’Myrobalan’	1919 a	1129 a	- -	1388 a	1794 a
’GF655/2′	1991 a	969 a	1254 a	- -	- -
’St. Julien A’	- -	1223 a	1449 a	2131 b	1928 a
’Fereley’	1734 a	1920 b	1911 a	- -	- -
’Wangenheim’	- -	875 a	- -	- -	- -
’WaVit’	- -	- -	1450 a	- -	- -

Note: Means are separated by Duncan’s multiple range test; different letters following the data stand for significant differences at p = 0.05.

) of the whole tree in the case of ‘Toptaste’ and ‘Fereley’, and ‘Čačanska lepotica’ and ‘St. Julien A’ combinations.

3.4. Performance of Fruiting Tree Development

The results of the statistical analysis showed significantly different performance of investigated varieties in the development of the fruiting wood type (

Table 10. Performance of rate of fruiting wood type on investigated varieties (piece %).

	Bouquets and Bouquet Spurs	Short Fruiting Laterals	Medium and Long Fruiting Laterals
‘Topper’	47.60 c	37.90 ab	14.50 a
‘Toptaste’	21.90 a	36.50 ab	41.60 c
‘Topfive’	17.30 a	52.80 c	29.90 b
‘Čačanska lepotica’	47.70 c	41.00 b	11.30 a
‘Jojo’	36.50 b	29.50 a	34.00 b

Note: Means are separated by Duncan’s multiple range test; different letters following the data stand for significant differences at p = 0.05.

). The varieties ’Topper’ and ’Čačanska lepotica’ predominantly developed bouquets (47.44% and 47.13%, respectively), the dominant fruiting wood type on ’Topfive’ were short shoots (52.9%), while the variety ’Toptaste’ developed long fruiting laterals at the largest rate (41.62%).

The rate of short fruiting laterals on ‘Topfive’ significantly differed from the other two fruiting lateral types. In the case of ‘Čačanska lepotica’, there was a significant difference between all types of the fruiting laterals, while the rate of bouquets and long fruiting branches of the ‘Jojo’ trees did not significantly differ; but the rate of short fruiting laterals was smaller. No significant difference was found between the rate of fruiting laterals of the cultivars on different rootstocks (

Table 11. Proportion of fruiting laterals in varieties on each rootstock (piece %).

Rootstock	‘Topper’			‘Toptaste’			‘Topfive’
	B	S	L	B	S	L	B	S	L
‘Myrobalan’	48.67	36.00	15.33	19.50	30.42	50.08	-	-	-
‘GF 655/2′	50.30	35.40	14.30	28.00	33.50	38.50	18.00	53.70	28.30
‘St. Julian A’	-	-	-	20.90	47.50	31.60	12.00	52.10	35.90
‘Fereley’	43.50	42.70	13.80	24.50	37.00	38.50	18.25	50.33	31.42
‘Wangenheim’	-	-	-	16.50	34.20	49.30	-	-	-
‘WaVit’	-	-	-	-	-	-	23.00	55.00	22.00
Mean of variety	47.49	38.03	14.48	21.88	36.52	41.60	17.81	52.78	29.41
	‘Čačanska lepotica’			‘Jojo’			Mean of rootstock
	B	S	L	B	S	L	B	S	L
‘Myrobalan’	49.80	40.00	10.20	33.00	33.00	34.00
‘GF655/2′				-	-	-
‘St. Julien A’	44.50	41.90	13.60	40.00	26.00	34.00
‘Fereley’	-	-	-	-	-	-
‘Wangenheim’	-	-	-	-	-	-
‘WaVit’	-	-	-	-	-	-
Mean of variety	47.15	40.95	11.90	36.50	29.50	34.00

Note: B: bouquet and bouquets spurs; S: short laterals; L: long fruiting wood.

).

4. Discussion

Based on the TCSA performance of the trees, rootstocks could be classified into two large groups: the vigorous ones are ‘Myrobalan’, ‘St. Julien GF 655/2′, ‘St. Julien A’ and ‘Fereley’, while the ‘Wangenheim’ and ‘WaVit’ rootstocks proved to be dwarfing. This corresponds to the literature data [10,14], except for ‘St. Julien GF 655/2′, which was reported to be moderately vigorous by [2,14]. It is probable that the vigour differences in the vigorous group of rootstocks are smaller than expected. The cause of this phenomenon might be, on the one hand, the smaller plant distance [2] and, on the other hand, the root restriction effect on the container-grown trees.

In high-density orchards, the trunk cross-sectional area is the parameter of growth vigour, which is less influenced by pruning, and the planting distance [2,3,38]. Considering the TCSA, measured in 2016, the rootstocks significantly influenced the growth vigour of trees, but this effect was differently manifested by varieties (Table 5 and Figure 3).

The largest TCSA of the variety ‘Topper’ was measured on the ‘Myrobalan’ rootstock; similarly vigorous growth was measured on ‘St. Julien GF 655/2′ (95%) without any significant differences, while the TCSA on ‘Fereley’ was 61% compared with that of ‘Myrobalan’. In the canopy volume (CV) of ‘Topper’, there was no significant difference. On the contrary, ‘Toptaste’ produced the largest TCSA and CV on the ‘Fereley’ rootstock followed by ‘St. Julien A’ (76%), ‘St. Julien GF 655/2′ (76%) and ‘Myrobalan’ (68%) with a significantly smaller TCSA, while the trees were less vigorous on the ‘Wangenheim’ rootstock (56%). The CV followed this performance, too. The ‘Topfive’ trees again showed the largest TCSA on ‘Fereley’, too, without significant differences to the similarly vigorous St. Julien A (96%) and St. Julien GF 655/2 (89%), while the rootstock ‘WaVit’ produced a significantly smaller TCSA (51%). The CV similarly performed: a significantly larger CV on ’Fereley’ and a smaller one on the other rootstocks (77–58%). The varieties ‘Čačanska lepotica’ and ‘Jojo’ were vigorous on both investigated rootstocks without significant differences. Based on these results, the ‘Fereley’ rootstock can be considered very vigorous for ‘Toptaste’ and ‘Topfive’ but not for ‘Topper’, which produced moderately vigorous trees. Our further data on ‘Myrobalan’, ‘St. Julien A’ and ‘GF 655/2′ also confirm the opinion of the authors [9,13,36] who reported similar results. Corresponding to the literature data, the ‘Wangenheim’ and ‘WaVit’ rootstocks proved to be dwarfing for the ‘Toptaste’ and ‘Topfive’ varieties [23].

It is supposable that the trees of ‘Topper’, ‘Čačanska lepotica’ and ‘Jojo’ on the vigorous ‘Myrobalan’ rootstock are more vigorous than those of the ‘Toptaste’ variety. The TCSA of ‘Toptaste’ only reached 55% of that of the ‘Jojo’ trees. The TCSA of the trees on ‘St. Julien A’, ‘Toptaste’ and ‘Topfive’ similarly produced, only 59% and 67%, compared with the ‘Jojo’ trees. Based on these results, the varieties ‘Topper’, ’Čačanska lepotica’ and ‘Jojo’ can be considered as vigorous varieties, while ‘Toptaste’ and ‘Topfive’ are moderately vigorous. These data contradict the literature data on these varieties [39]. The overall mean of the TCSA of the varieties should not be compared with each other because ‘Jojo’ and ‘Čačanska lepotica’ were planted on vigorous rootstocks only; ‘Wangenheim’ and ‘WaVit’ were missing combinations. Besides these data, ‘Čačanska lepotica’, ‘Jojo’ and ‘Topper’ proved to be more vigorous than the ‘Toptaste’ and ‘Topfive’ varieties.

The increment of the TCSA of the trees on various rootstocks was different in the development stages. The ‘Topper’, ‘Toptaste’ and ‘Topfive’ trees on vigorous or moderately vigorous rootstocks showed a significantly larger TCSA increment than on the dwarfing ‘Wangenheim’ or ‘WaVit’ rootstocks (Table 6 and Figure 3). These characteristics of rootstocks should be considered when planning the orchard spacing [3,38,40,41]. The data presented in Figure 3 show the increasing TCSA increment in the years from 2013 to 2016 on vigorous and moderately vigorous rootstocks, while the TCSA increment on dwarfing rootstocks ‘Wangenheim’ and ‘WaVit’ is reducing.

In the case of ‘Toptaste’, the initial increase in the TCSA (2010–2012) was the same on rootstocks, with no significant difference between them (Table 7). After the trees had turned to bearing (2012–2016), there were already great differences between the rootstocks. The ‘Wangenheim’ rootstocks had a smaller (2.09 cm²) 4-year increase in the productive years compared with the non-productive 2-year increase (2.51 cm²). In contrast, on the ‘Fereley’ rootstock, the initial weak TCSA growth (3.03 cm²) began to very strongly thicken after the trees turned to bearing, at a rate of 9.02 cm². The trunk thickening was stronger on the ‘Myrobalan’ and ‘St. Julien’ rootstocks, too. ‘Myrobalan’, ‘St. Julien A’ and ‘St. Julien GF655/2’ can be classified in the same group, yet the trees with the ‘St. Julien GF655/2’ rootstock thickened a little more vigorously.

The final trunk cross-sectional area showed a significant difference among the rootstocks of the ‘Toptaste’ varieties (Figure 3). The smallest increase was shown by the combination of ‘Toptaste’ and ‘Wangenheim’, which differed, but not significantly, from the combination of ‘Toptaste’ and ‘Myrobalan’. There was a significant difference between the ’St. Julien’ and ‘Fereley’ rootstocks. The ‘Toptaste’ trees on the ‘Fereley’ rootstock significantly differed from all other rootstocks, as their trunk has grown thicker over the years. Our results underline the individual performance of rootstock and scion combinations.

Regarding the shoot growth, the mean shoot length per tree and total shoot length per tree followed the growth vigour differences based on the trunk thickness indices. The exceptions are the average shoot length of the ’Toptaste’ trees on ’Fereley’ and all the shoot lengths of ’Toptaste’ and ’Fereley’, and ’Čačanska lepotica’ and ’St. Julien’ in rootstock–variety combinations. The ‘Toptaste’ and ‘Fereley’ trees show robustness not only in the trunk thickness and crown volume but also in the shoot length (143.10 cm) and total shoot length (1920 cm) of the tree. All these data confirm our findings on the growth strength groups found for each combination.

Regarding the irrigation water doses, the results of the statistical analysis did not show any significant effect, except for the increase in the stem thickness measured in the years between 2013 and 2016 for ’Čačanka lepotica’ and ’Jojo’. In the case of ’Toptaste’ and ’Topfive’, which had a weaker growth than the other cultivars, it can be concluded that the amount of irrigation water did not have a significant effect on their growth. In contrast, due to irrigation, the stronger-growing ’Čačanska lepotica’ and ’Jojo’ cultivars provided an approximately 30–40% higher trunk cross-sectional area growth on the strong-growing rootstocks between 2013 and 2016 (

Table 12. Effect of irrigation on growths of TCSA between 2013 and 2016 (cm ²).

Irrigation	’Čačanska lepotica’	’Jojo’
1	8.71 a	10.41 a
2	11.94 b	14.33 b

Note: Means are separated by Duncan’s multiple range test; different letters following the data stand for significant differences at p = 0.05.

).

The evolution of the survival rate of the trees on the ’Fereley’ rootstock is remarkable between 2013 and 2016, when half of the trees irrigated with twice the water dose died (Table 4). This result clearly suggests, especially in the case of ’Toptaste’ and ’Topfive’, where this rootstock gave the strongest growth, that ’Fereley’ does not require the use of high doses of water anymore; moreover, it is particularly unfavourable to it.

The distribution of plum trees coming into production in 2013 was typical of their species, with little influence from the rootstocks (Table 10 and Table 11, Figure 4). As we did not find any literature data on the fruiting wood distribution of the varieties and the interaction with the rootstocks, we consider this as a new finding. The predominant fruiting wood type of the varieties ’Topper’ and ’Čačanska lepotica’ were bouquets and bouquet spurs, while the dominant fruiting wood type on ’Topfive’ were short shoots; but the variety ’Toptaste’ developed long fruiting laterals at the largest rate. Despite expectations, dwarfing rootstocks did not increase the rate of bouquets and bouquet spurs on the ‘Toptaste’ and ‘Topfive’ varieties.

5. Conclusions

The investigated varieties showed large differences in vigour; ‘Topfive’ and ‘Toptaste’ are less vigorous on all rootstocks than ‘Čačanska lepotica’, ‘Topper’ and ‘Jojo’. The rootstocks differently affected the vigour of the trees in the interaction with the vigour of the varieties. Based on our results, the ‘Fereley’ rootstock can be considered vigorous for ‘Toptaste’ and ‘Topfive’ but not for Topper, which produced moderately vigorous trees. Our data on ‘Myrobalan’, ‘St. Julien A’ and ‘GF 655/2′ also confirm the opinion of the authors who reported similar results. Corresponding to the literature data, the ‘Wangenheim’ and ‘WaVit’ rootstocks proved to be dwarfing for the ‘Toptaste’ and ‘Topfive’ varieties. The performance of the canopy volume and shoot length follows the growth vigour of scion and rootstock combinations. The growth of the trees on the dwarfing rootstocks slowed down after the third year, while the TCSA increment on the rootstocks of the vigorous group continued to increase. The fruiting wood development in 2013 showed large differences between the varieties, which had been hardly influenced by the rootstocks. The varieties ‘Topper’, ‘Čačanska lepotica’ and ‘Jojo’ produced more bouquets and bouquet spears, while a larger rate of short, medium and long fruiting shoots was produced by ‘Toptaste’ and ‘Topfive’. No considerable effect of the irrigation dosage was detected; however, ‘Čačanska lepotica’ and ‘Jojo’ on the vigorous ‘Myrobalan’ and ‘St. Julien A’ increased the TCSA increment after 2013. Furthermore, the trees on the ‘Fereley’ rootstock irrigated with a double dose of water died at a considerably higher rate (50–67%); the reason for this phenomenon needs further investigations.