Effects of seedling and clonal West Indian rootstocks irrigated with recycled water on ‘Hass’ avocado yield, fruit weight and alternate bearing

ABSTRACT

In modern avocado orchards, ‘Hass’ scions are grafted on selected rootstocks. Avocado rootstocks can be propagated via seeds (seedling rootstocks) or by vegetative propagation (clonal rootstocks). As choice of the proper rootstock influences productivity during the orchard's entire life span, it is critical to evaluate the effects of rootstocks on production parameters of the ‘Hass’ scion. We examined the effects of different rootstocks on fruit yield, fruit weight and alternate bearing of ‘Hass’ scions grafted on nine different West Indian rootstocks. The experiments were conducted from 2009–2019 in two commercial plots located in northwestern Israel irrigated with recycled water. Multiyear mean yield and alternate bearing index (ABI) in Afek orchard ranged from 11,032–15,214 kg ha⁻¹ and 0.26–0.5, respectively. Multiyear mean yield and ABI in Gesher Haziv orchard ranged from 12,746–15,047 kg ha⁻¹ and 0.34–0.57, respectively. Results revealed no significant differences in the multiyear production parameters of the examined rootstocks. However, a non-significant advantage was shown for the vegetative rootstocks, most of which had lower ABI values than the seedling rootstocks. Therefore, because vegetative rootstocks are more expensive to produce, it is practically recommended to use the seedling rootstocks. In the future, it is suggested to examine additional vegetative rootstocks.

KEYWORDS:

• Biennial
• low-quality irrigation
• multiyear yield
• Persea americana
• salt stress
• subtropical

Introduction

Avocado (Persea americana Mill.) is a subtropical fruit tree with high commercial value and increasing demand worldwide (Alcaraz et al. 2013; Migliore et al. 2017). The black-skinned Guatemalan–Mexican hybrid avocado cv. Hass has great commercial importance, as it currently dominates global production (Schaffer et al. 2013). For example, most of the total avocado supply in the United States consists of ‘Hass’, with about 97% of the total trade in 2018 (Ballen et al. 2021). Despite its popularity, however, this cultivar is characterised by variation in yearly orchard yields (Ziv et al. 2014). This Vphenomenon, referred to as alternate bearing, is considered a negative trait by both growers and distributers.

In all modern avocado orchards, ‘Hass’ scions are grafted on various selected rootstocks, providing a dual plant system to increase orchard productivity (Ben-Ya’acov and Michelson 1995; Ahsan et al. 2019). Typically, the scions are characterised by high fruit yield and other qualities of importance to growers and consumers, whereas the rootstocks are commonly selected for their ability to donate their disease and stress tolerance, as well as vigour qualities. For instance, avocado rootstocks have been selected and developed based on their tolerance to high soil and irrigation-water salinity, adverse climate conditions, and soilborne fungal pathogens such as Phytophthora cinnamomi and Verticillium dahliae (Krezdorn 1973; Menge et al. 1992; Ben-Ya’acov and Michelson 1995; Mickelbart and Arpaia 2002; Castro et al. 2009; Haberman et al. 2020; Kourgialas and Dokou 2021).

Avocado rootstocks can be propagated by one of two main methods: via seeds (seedling rootstocks) or by vegetative propagation (clonal rootstocks). The production of seedling rootstocks is very easy to operate and relatively inexpensive. The main drawback of these rootstocks is their nonuniformity, due to high genetic variation (Hiti-Bandaralage et al. 2017). Clonal rootstocks have been commercially available since the mid-1970s; they are produced by rooting cuttings taken from mature trees. A major advantage of the clonal propagation method is its ability to produce rootstocks of known uniformity. However, whereas the production of seedling rootstocks is considered relatively uncomplicated, creating clonal rootstocks is more complex and laborious (Ben-Ya’acov and Michelson 1995). Moreover, avocado trees grafted on clonal rootstocks are usually more expensive for growers than those grafted on seedling rootstocks. Because choice of a proper rootstock influences productivity throughout the orchard's entire life span, it is highly important to investigate, evaluate and compare the effects of different rootstocks, both seedling and clonal, on production parameters of the ‘Hass’ scion. Therefore, the aim of this study was to examine the effects of seedling and clonal rootstocks on ‘Hass’ fruit yield, fruit weight and alternate bearing. We hypothesised that even though yield and fruit weight parameters may be similar for both seedling and clonal rootstocks, alternate bearing, measured by the alternate bearing index (ABI), will be lower with the latter, due to their homogeneous nature (Ben-Ya’acov and Michelson 1995).

It is important to note that in many agricultural areas worldwide, freshwater for agricultural irrigation is very limited. With the predicted drying trends associated with climate change, availability is expected to decrease even further, while soil salinity is expected to rise (Acosta-Rangel et al. 2019; Amanullah et al. 2020; Corwin 2021). Specifically, most of the avocado orchards in Israel, as well as in other countries, are irrigated with recycled water that has high concentrations of soluble salts (Branson and Gustafson 1972). Therefore, in this study, we only investigated West Indian rootstocks, which are considered salt-tolerant compared to Mexican rootstocks (Mickelbart and Arpaia 2002; Oster et al. 2007; Lazare et al. 2021).

Materials and methods

Plant material and experimental sites

‘Hass’ scions were grafted on different rootstocks (Table 1) and planted in two locations in northwestern Israel in 2009. Both orchards were irrigated using recycled water with high chloride values (Table 2). Therefore, only the salt-tolerant West Indian rootstocks were chosen for this study. ‘Hass’ scions from the same tree were grafted in 2008 by Haskelberg Nursery Ltd. (Kfar-Vitkin, Israel) and Rahan Meristem Ltd. (Rosh Hanikra, Israel) on nine different West Indian rootstocks: four seedling (‘Ashdot 17’, ‘Degania 117’, ‘Degania 113’ and ‘Fairchild’) and five clonal (VC28, VC66, VC51, VC26 and VC27) rootstocks (Table 1).

Table 1. Rootstocks.

	Race	Propagation method	Origin	Nursery
Rootstock
Ashdot 17	West Indian	Seed	Israel	Haskelberg Nursery
Degania 117		Seed	Israel	Haskelberg Nursery
Degania 113		Seed	Israel	Rahan Meristem
Fairchild		Seed	Florida, USA	Haskelberg Nursery
VC28		Vegetative	Israel	Haskelberg Nursery
VC66		Vegetative	Israel	Haskelberg Nursery
VC51		Vegetative	Israel	Haskelberg Nursery
VC26		Vegetative	Israel	Haskelberg Nursery
VC27		Vegetative	Israel	Haskelberg Nursery

Table 2. Experimental orchards.

	Gesher Haziv	Afek
Orchard location
Coordinates	33°02'33''N 35°07'14''E	32°50'16''N 35°08'57''E
Altitude (m a.s.l.)	20	30
Soil
Lime (%)	20–30	1–5
Clay (%)	50	60
Irrigation
Irrigation rate (m^3 year ^−1 ha ^−1)	8,500	9,000
Chloride (mg l ^−1)	200	230–260
Nitrate (ppm)	<10	<10
Potassium (ppm)	25	30–35
EC (dS m^−1)	1.1-1.4	1.3-1.5
Fertilisation
Nitrogen (units ha ^−1)	300	300
Potassium oxide (units ha ^−1)	350	350
Phosphorus (units ha ^−1)	30	30
Orchard structure
Planting spacing between rows (m)	6	5
Planting spacing between trees (m)	4	2.5
Tree density (trees ha ^−1)	420	800
Tree row orientation	north-south	east-west

The experiments were conducted from 2009–2019 in two commercial ‘Hass’ avocado plots located in Kibbutz Afek and Kibbutz Gesher Haziv in northwestern Israel (Table 2, Fig. S1). Data were collected over six consecutive growing seasons, from the season of 2013–2014 to the season of 2018–2019. Orchard location, soil, irrigation, fertilisation and structure are detailed in Table 2. Trees were irrigated using drip-irrigation at an amount of 8,500-9,000 m³ year ⁻¹ ha ⁻¹. Irrigation intervals were between 2–3 days and liquid fertiliser was supplied with each irrigation. In both orchards, yearly fertilisation included nitrogen (300 units ha ⁻¹), potassium oxide (350 units ha ⁻¹) and phosphorus (30 units ha ⁻¹). The experimental plots were completely randomised, with four repeats for each treatment and at least 11 (Afek) or 15 (Gesher Haziv) trees in each repeat. ‘Ettinger’ trees were planted every third tree in every third row to serve as pollinizers.

Yield, fruit weight and ABI measurements

Four years after planting (season of 2013–2014), fruit were manually picked during the commercial harvest period—during the winter (January–February) of each year (approx. 9–10 months after fruit set)—for 6 consecutive years (until the season of 2018–2019). For each repeat, total ‘Hass’ fruit weight was divided by tree number to obtain the average fruit yield per tree (kg tree⁻¹). Yearly mean yield of each rootstock was calculated by averaging the yield of each of the four repeats. Multiyear mean yield of each rootstock was calculated by averaging the yields of each of the 6 years of the experiment.

In addition, fruits from eight individual trees (two trees from each repeat) of each rootstock were numbered and weighed. Mean fruit weight of each rootstock was calculated by averaging the quotient of the yield of each tree and its fruit number.

ABI calculation was based on the method of Hoblyn et al. (Hoblyn et al. 1937) and expressed as an absolute number: $\mathrm{A}\mathrm{B}\mathrm{I}={\displaystyle \frac{\left|\mathrm{y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d}\mathrm{o}\mathrm{f}\mathrm{y}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{x}-\mathrm{y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d}\mathrm{o}\mathrm{f}\mathrm{y}\mathrm{e}\mathrm{a}\mathrm{r}\left(\mathrm{x}-1\right)\right|}{\mathrm{y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d}\mathrm{o}\mathrm{f}\mathrm{y}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{x}+\mathrm{y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d}\mathrm{o}\mathrm{f}\mathrm{y}\mathrm{e}\mathrm{a}\mathrm{r}\left(\mathrm{x}-1\right)}}$A bearing cycle was defined as two consecutive seasons. Multiyear ABI was calculated by averaging all of the ABI values of each rootstock.

Statistical analysis

All results from the same experimental site were subjected to one-way ANOVA followed by Tukey–HSD test in JMP version 11.0.0 (SAS Institute, Cary, NC, USA).

Results

Effect of different rootstocks on production parameters – Afek orchard

During the first and second measurement seasons, yields ranged from 12.6–22.3 and 15–27.4 kg tree⁻¹, respectively, with no significant differences between the different rootstocks (Figure 1A). In 2015–2016, yields were 3.6–14.9 kg tree⁻¹. Yields of three seedling rootstocks (Ashdot 17, Degania 117 and Degania 113) were significantly lower than those of three clonal rootstocks (VC66, VC51 and VC27). In 2016–2017, yields were 7.1–25.3 kg tree⁻¹. In this season, the yields of the three same seedling rootstocks—Ashdot 17, Degania 117 and Degania 113—were significantly higher than those of two clonal rootstocks (VC66 and VC51). In 2017–2018, yields were 10.4–34 kg tree⁻¹ and VC27 had the greatest yield, which was significantly higher than those of all of the seedling rootstocks. In 2018–2019, yields were 9.3–23 kg tree⁻¹. The yields of two seedling rootstocks (Degania 117 and Degania 113) were significantly higher than those of three clonal rootstocks (VC28, VC51 and VC27).

Figure 1. Effect of rootstocks on avocado yearly yield (A), fruit weight (B), and alternate bearing index (ABI) (C) at Afek experimental site during six consecutive growing seasons. ‘Hass’ scions were grafted on eight different rootstocks and planted in the Afek orchard in 2009 at a density of 800 trees ha⁻¹. Data were collected during six consecutive growing seasons, from 2013–2014–2018–2019. A, Yield values are means ± SE of four repeats (n = 4), at least 11 trees in each repeat. B, Fruit weight values are means ± SE of eight different trees for each rootstock (n = 8). C, ABI values were calculated for each pair of consecutive growing seasons: seasons 1–2 (2013–2014 and 2014–2015), 2–3 (2014–2015 and 2015–2016), 3–4 (2015–2016 and 2016–2017), 4–5 (2016–2017 and 2017–2018) and 5–6 (2017–2018 and 2018–2019). Values are means ± SE of four repeats (n = 4) in each year, at least 11 trees in each repeat. Columns marked with different letters differ significantly by Tukey–HSD, p < 0.05.

During the first, second and third measurement seasons, mean fruit weight ranged from 216–265, 185–229 and 221–287 g, respectively, with no significant differences between the rootstocks (Figure 1B). In 2016–2017, mean fruit weight was 171–238 g. The mean fruit weight of VC27 was significantly higher than those of Ashdot 17, Degania 117, Degania 113 and VC28. In 2017–2018, mean fruit weight was 181–253 g, and the mean fruit weight of Degania 113 was significantly higher than those of both VC28 and VC51. In 2018–2019, mean fruit weight ranged between 202 and 267 g, with no significant differences between the rootstocks.

ABI values for seasons 1 and 2 ranged from 0.05–0.54, with VC66 and VC51 having significantly lower values than Degania 113 (Figure 1C). ABI values for seasons 2–3 ranged between 0.12 and 0.6, but with no significant differences between rootstocks. ABI values for seasons 3 and 4 were 0.09–0.72. During these consecutive seasons, the ABI values of two clonal rootstocks (VC28 and VC27) were significantly lower than those of three seedling rootstocks (Ashdot 17, Degania 117 and Degania 113). ABI values for seasons 4 and 5 ranged from 0.1–0.54. Fairchild had the lowest ABI value, significantly lower than those of VC66 and VC51. VC28 also had a low ABI value of 0.12, significantly lower than VC51. The ABI values for seasons 5 and 6 were 0.13–0.56, with no significant differences between the rootstocks.

The multiyear mean yield in the Afek orchard ranged between 13.8 and 19.01 kg tree⁻¹ (11,032 and 15,214 kg ha⁻¹), with no significant differences between the different rootstocks (Figure 2A). The multiyear mean ABI values in this orchard ranged from 0.26–0.5, also with no significant differences between the rootstocks (Figure 2B). The multiyear average ABI value for all rootstocks was 0.35: multiyear mean values of three seedling rootstocks (Ashdot 17, Degania 117 and Degania 113) were above this value, and those of all of the clonal rootstocks and the seedling rootstock Fairchild were below it.

Figure 2. Effect of rootstocks on avocado multiyear (over 6 years) mean yields (A) and alternate bearing index (ABI) (B) at Afek experimental site. A, Yield values are means ± SE of six (n = 6) consecutive yearly yields. B, ABI values are means ± SE of five (n = 5) consecutive biennial ABI values. Dashed line indicates the multiyear average ABI value of all rootstocks.

Effect of different rootstocks on production parameters – Gesher Haziv orchard

In general, fruit yield for each tree in the Gesher Haziv orchard was higher than that in the Afek orchard, because tree density was much lower (420 trees ha⁻¹ compared to 800 trees ha⁻¹, respectively). During the growing seasons of 2013–2014, 2014–2015, 2015–2016, 2016–2017 and 2017–2018, yields in Gesher Haziv orchard ranged from 23.6–36.2, 40.1–55.3, 10.3–20.2, 38.9–54.6 and 10.3–19.2 kg tree⁻¹, respectively, with no significant differences between the different rootstocks (Figure 3A). In 2018–2019, yields were 32.1–59.8 kg tree⁻¹, and the yield of Ashdot 17 was significantly higher than those of three clonal rootstocks (VC28, VC66 and VC51).

Figure 3. Effect of rootstocks on avocado yearly yield (A), fruit weight (B), and alternate bearing index (ABI) (C) at Gesher Haziv experimental site during six consecutive growing seasons. ‘Hass’ scions were grafted on eight different rootstocks and planted in the Gesher Haziv orchard in 2009 at a density of 420 trees ha⁻¹. Data were collected during six consecutive growing seasons, from 2013–2014–2018–2019. A, Yield values are means ± SE of four repeats (n = 4), at least 15 trees in each repeat. B, Fruit weight values are means ± SE of eight different trees for each rootstock (n = 8). C, ABI values were calculated for each pair of consecutive growing seasons: seasons 1–2 (2013–2014 and 2014–2015), 2–3 (2014–2015 and 2015–2016), 3–4 (2015–2016 and 2016–2017), 4–5 (2016–2017 and 2017–2018) and 5–6 (2017–2018 and 2018–2019). Values are means ± SE of four repeats (n = 4) in each year, at least 15 trees in each repeat. Columns marked with different letters differ significantly by Tukey-HSD, p < 0.05.

During the growing seasons of 2013–2014, 2014–2015, 2015–2016, 2016–2017 and 2017–2018, mean fruit weight ranged from 207–241, 176–190, 232–286, 176–215 and 222–271 g, with no significant differences between the rootstocks (Figure 3B). Technical difficulties prevented the collection of fruit weight data in the 2018–2019 season.

The ABI values for seasons 1–2, 2–3, 3–4, 4–5, and 5–6 ranged from 0.14–0.4, 0.4–0.68, 0.37–0.67, 0.47–0.6 and 0.39–0.56, respectively, with no significant differences between the rootstocks (Figure 3C).

The multiyear mean yield in the Gesher Haziv orchard ranged from 30.3–35.8 kg tree⁻¹ (12,746–15,047 kg ha⁻¹), with no significant differences between the different rootstocks (Figure 4A). The multiyear mean ABI value in this orchard ranged from 0.34–0.57, also with no significant differences between the different rootstocks (Figure 4B). The multiyear average ABI value of all rootstocks was 0.46: multiyear mean values of two seedling rootstocks (Degania 117 and Degania 113) were above this value; those of the other seedling rootstocks (Ashdot 17 and Fairchild) and one clonal rootstock (VC26) were similar to this value; those of the three other clonal rootstocks (VC28, VC66 and VC51) were lower than this value.

Figure 4. Effect of rootstocks on avocado multiyear (over 6 years) mean yields (A) and alternate bearing index (ABI) (B) at Gesher Haziv experimental site. A, Yield values are means ± SE of six (n = 6) consecutive yearly yields. B, ABI values are means ± SE of five (n = 5) consecutive biennial ABI values. Dashed line indicates the multiyear average ABI value of all rootstocks.

Discussion

Fruit yield is a major economically important trait in crop trees, as well as other crops. In the Afek orchard, significant differences in fruit yield between the different rootstocks were observed in four of the six seasons examined (Figure 1A). However, in the Gesher Haziv orchard, significant differences were only observed in one season (Figure 3A). Nevertheless, none of the examined rootstocks achieved significantly higher multiyear yields than the others, in either orchard (Figures 2A and 4A). This suggests that under the conditions of these experiments, there was no clear advantage of a specific rootstock for this important parameter. Similar results were found in a study conducted in Cyprus, in which there were no significant differences in the cumulative yields of avocado cultivars ‘Ettinger’, `Fuertè and ‘Hass’ grafted on `Lulà and West Indian rootstocks (Gregoriou and Economides 1991). However, in other studies, rootstocks did have an effect on the yield of ‘Hass’ scions. For example, in a study conducted in southern California, ‘Hass’ trees grafted on ‘Duke 7’ and ‘Borchard’ had significantly greater yields than those grafted on G755A, G755B and G755C—all vegetative rootstocks (Mickelbart et al. 2007). In two other studies conducted in Israel, there were significant differences in the yield of ‘Hass’ trees grafted on various rootstocks and grown under biotic stress in a plot inoculated with a high level of Verticillium dahliae (Haberman et al. 2020) or under high salt stress (Lazare et al. 2021). Such differences in yield have also been found in other fruit crops. For example, results from another study conducted in Cyprus showed that there were significant differences between the yields of ‘Shamouti’ orange grafted on 14 different rootstocks (Georgiou and Gregoriou 1999). The difference between the above-listed results in avocado and those in our study may be explained by differences in growing conditions and the rootstocks examined. Moreover, effects of rootstock on yield are strongly associated with the rootstock's tolerance to biotic diseases or abiotic stress (Bowman and Joubert 2020). Thus, it is possible that under other growth-constraining conditions, such as exposure to soil pathogens or adverse climatic events such as frost, differences in production parameters between the examined rootstocks would be more obvious.

Fruit weight is also an important economic and commercial factor for avocado growers, as it may impact consumer preferences. In this study, significant differences in fruit weight were observed only in the Afek orchard in two of the six growing seasons (Figure 1B), and none were observed in the Gesher Haziv orchard (Figure 3B). These data suggest that under these growing conditions, the evaluated rootstocks did not have a substantial effect on fruit weight. Similar results were obtained in two different studies conducted in Cyprus (Gregoriou and Economides 1991; Gregoriou 1992). However, other studies have suggested that rootstock does play a role in determining fruit size in avocado and mango (Avilán et al. 1997; Mickelbart et al. 2007). Again, it is possible that under different growing conditions, which might cause significant differences in fruit yield between the rootstocks, there would be also significant differences in average fruit weight.

Variation in yearly orchard yields is considered a negative phenomenon in the avocado industry, as well as with other fruit crops (Monselise and Goldschmidt 1982). High alternate bearing may complicate orchard management and result in income and profit losses during low-yield (off) years, and glut during high-yield (on) years (Lovatt 2010). Therefore, it is important to screen for and assess the productivity of rootstocks with low ABI quality (Mickelbart et al. 2007). Although we did observe differences in the ABI values among rootstocks in some of the growing seasons (Figures 1C and 3C), the overall multiyear ABI values did not differ significantly, in either experimental plot (Figures 2B and 4B). Nevertheless, the multiyear ABI value for all clonal rootstocks in Afek and three of the clonal rootstocks in Gesher Haziv was below the multiyear average ABI value for all rootstocks. In contrast, only one seedling rootstock in Afek and none of the seedling rootstocks in Gesher Haziv had a lower multiyear ABI value than the average for all rootstocks (Figures 2B and 4B). This may give credence to our hypothesis that the homogeneous nature of the clonal rootstocks compared to the seedling rootstocks leads to more stable yields throughout the years. Note, also, that the multiyear ABI value of the clonal rootstocks at both experimental sites ranged between 0.26 and 0.47, generally lower than those of the 10 clonal rootstocks examined in southern California, which ranged between 0.33 and 0.6 (Mickelbart et al. 2007). Having said that, as there are many differences between the growing conditions in northwestern Israel and southern California, it is hard to compare the results of these studies. It would therefore be of interest to evaluate the clonal rootstocks examined in this study in other avocado-cultivation regions as well.

All in all, our results showed that there is no significant difference in the multiyear production parameters for any of the examined rootstocks. This might be due to the high tolerance of the selected rootstocks to the high salinity of the recycled irrigation water used in the orchards. Nevertheless, in accordance with our hypothesis, a non-significant advantage was observed for the vegetative rootstocks, as most of them had lower ABI values than those of the seedling rootstocks. Note, however, that it is practically recommended to use the seedling rootstocks when choosing from these rootstocks in new plantations because vegetative rootstocks are more expensive to produce. Further research should be conducted to evaluate the examined rootstocks, as well as additional rootstocks, under various topo-climatic conditions, especially in light of climate change which may disrupt world avocado production (Bar-Noy et al. 2019; Shapira et al. 2021; Zandalinas et al. 2021; Chernoivanov et al. 2022). Thus, as the frequency of extreme climatic events increases, rootstock selection may be the key to improving avocado performance under these conditions (Balfagón et al. 2021).

Acknowledgments

The authors wish to thank the Israeli Avocado Growers Board for financial support, and the Kibbutz Afek and Kibbutz Gesher Haziv avocado teams for the effort they invested in this study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by The Israeli Avocado Growers Board.