Potential of Different Actinidia Genotypes as Resistant Rootstocks for Preventing Kiwifruit Vine Decline Syndrome

: Kiwifruit Vine Decline Syndrome (KVDS) is currently affecting Italian kiwifruit cultivation, causing dramatic yield and economic losses. The syndrome’s aetiology is due to soil-borne pathogens and waterlogging, leading to the decay of roots and then the canopy. Current knowledge about the disease is limited, and the techniques to control the syndrome are ineffective. The use of tolerant rootstocks is one of the most promising tools. Six genotypes of Actinidia were tested for two years at four infected experimental sites in Friuli Venezia Giulia (NE Italy). Plant evaluation and analysis were carried out on the root system and the vegetative parts. At all experimental sites, three genotypes, all belonging to the A. macrosperma group, grew normally. In contrast, plants of A. polygama died earlier and those of A. chinensis var. deliciosa ‘Hayward’ declined during the ﬁrst year. A. arguta ‘Miss Green’ survived the ﬁrst year but started to decline during the second year. After two years of study, we were able to identify three putative resistant genotypes: A. macrosperma accession numbers 176 and 183, and ‘Bounty71’, which will be a useful resource as rootstocks or as parents for breeding owing to their potential genetic resistance traits.


Introduction
Kiwifruit is a recently domesticated plant belonging to the family Actinidiaceae and the genus Actinidia, which contains ±54 species [1]. However, two taxa, A. chinensis var. deliciosa and A. chinensis var. chinensis, dominate the kiwifruit industry [2]. Breeders have shown an increasing interest in kiwifruit, and long-term breeding programs have recently resulted in the development of several cultivars [3]. Consumption of kiwifruit has been increasing steadily, with the fruit recognised as a superfood, owing to its high vitamin C content. Kiwifruit is a highly profitable crop with a long shelf life suited to global trade [4]. In Italy, the area planted with kiwifruit is currently 25,000 ha [5].
Recently, a new and complex disease called Kiwifruit Vine Decline Syndrome (KVDS), also known as "Moria del kiwi", has appeared in different kiwifruit growing regions, being reported initially in Italy and then worldwide [6,7]. This has led to significant yield and economic losses, around EUR 300,000,000 in 2020 [8]. KVDS was initially observed in poorly drained soils where waterlogging conditions occur most easily [9]. The abiotic factor of waterlogging is thought to favour the development of this syndrome [10], although KVDS can occur in many soil types. Several soil-borne pathogens have been implicated as aetiological agents (biotic factors) [10]. Phytophthora spp., Pythium spp., Phytopythium spp., Fusarium spp., Cylindrocarpon spp. and Desarmillaria tabescens were frequently isolated from affected plants [9,[11][12][13]. Pathogenicity on kiwifruit cuttings has been demonstrated for some isolates: Phytophthora cryptogea, Phytophthora citrophthora, Phytopythium vexans and

Plant Material and Experimental Design
The trial was conducted in the period 2020-2021 in the Udine and Pordenone provinces in the Friuli-Venezia Giulia Region (FVG) (Italy, NE), at four different sites including three orchards of A. chinensis var. deliciosa Hayward and one of A. chinensis var. chinensis Soreli. The locations of the sites are as follows: site-1 ( 45.4 E). The sites were chosen in accordance with previous research which recently found these sites are KVDS inducive due to the presence of the soil-borne pathogens already mentioned above [9]. The genotypes tested were A. macrosperma, accession numbers 176 (Ma176) (♂) and183 (Ma183) (♂), Bounty 71, which also belongs to the A. macrosperma group, A. arguta Miss Green (MG), and A. polygama (POL) (♂). A. chinensis var. deliciosa Hayward (HW) (♀) was used as a KVDS-susceptible control. The selected genotypes are represented in the University of Udine germplasm collection and were chosen according to putative tolerance to waterlogging and soil-borne pathogens [27,28] and upon preliminary results [29].
The cuttings were rooted in sterile Agriperlite during Spring 2019, then were grown in pots in a greenhouse for a year and planted in the field early in Spring 2020. Four plants at each experimental site were observed in our study; thus, there was a total of 16 replicates across the four experimental sites. The test plants were planted along the main plant rows, replacing dead plants, and equally distributed at each experimental site, by adopting a randomised block design. Local standard practices were followed for pest management and fertilisation and were comparable at the different sites. The plants were irrigated according to the weather conditions to restabilize the field capacity, using a micro sprinkler system at each experimental site. Each year, at the end of the growing season, the plants were carefully removed from the soil and investigated. Measurements were taken during autumn (November 2020 and November 2021) when the plants were dormant. Root biometric indices were recorded, and samples of thick roots were kept and stored, as they are the most metabolically active [30], to carry out further chemical analyses. The whole developed canopy was also harvested to be further analyzed.

Pedological, Climatic and KVDS Characterisation of the Experimental Sites
Samples of soil were randomly collected at each site at 20 cm below ground and analysed according to the official Italian protocol (as indicated by Italian Law Decree n. 79/1992 and Law Decree n. 185/1999) [31]. Rainfall and temperature data were taken at a meteorological station located near the experimental sites. The presence of KVDS at each experimental site was initially assessed as incidence (% of symptomatic plants) and severity (% of attack per plant), using a previously used approach based on visual inspection of the aerial parts of the vines [32].

Survival of Genotypes at the Study Site
Each year, the percentage of plants of each genotype surviving was determined by a visual inspection at the end of growing season (November).

KVDS Root Symptom Survey
For each plant, the percentage of roots with necrosis or a rat-tail appearance was assessed. The root systems were photographed with a digital camera. The photos were processed with ImageJ (National Institutes of Health, Bethesda, MD, USA) to individuate the symptoms presence.

Root Biometric Indices: Volume, Width and Depth
The profile wall method of Böhm [33] was adopted to determine root indices. The width and depth (cm) of the root system were measured using a 35 cm high and 53 cm wide grid system positioned against the root system, keeping the vine trunk centred as a reference. The grid was divided internally into 1 cm × 1 cm squares. The volume (mL) of each below-ground part was recorded by dipping the whole sample into a graduated beaker partially filled with tap water and measuring the pre-and -post immersion difference in water volume.

Protein Content
The total protein content was measured in samples of thick roots because degradation of protein is known to be influenced by soil-borne pathogens [34]. The samples were dried in an oven at 65 • C for 1 day. The method of Fazeli et al. (2007) [35] was used: 1 g of root material was homogenised in 5 mL of 1M Tris-HCl buffer (pH 6.8). The homogenate was centrifuged in a refrigerated centrifuge at 10,000× g for 20 min, and the supernatant obtained was used for protein content determination. All the steps were carried out at 4 • C. The protein content of the extracts was determined according to Bradford [36] using bovine serum albumin (BSA) as a standard. A spectrophotometer (Dynatech MR5000, Dynex Technologies, Chantilly, VA, USA) was used to perform absorbance readings, at 560 nm. The values are expressed as mg g −1 Dry Weight (DW). Discontinuous sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE) was carried out in order to separate proteins and then determine their patterns.

Measurement of Total Non-Structural Carbohydrates (TNCs)
An assay of TNCs (sucrose, glucose, and fructose) on the thick roots was performed owing to the importance of TNCs in subsequent years for accumulation of starch reserves [37]. Briefly, TNCs were assessed using anthrone reagent (Merck, Darmstadt, Germany) and absorbance readings at 620 nm were performed using a Shimadzu UV Mini-1240 spectrophotometer (Kyoto, Japan) [38]. The data are expressed in mg g −1 DW.

Determination of Mineral Elements in Roots
Thick root (±50 g) samples were firstly washed with distilled water and then analyzed as described by Ari et al. [39] with the slight modifications of   [40]. Briefly, after having obtained the dry material, 1 g DW was digested in nitric acid HNO 3 (65%) (Sigma Aldrich, St. Louis, MO, USA), then diluted in a solution of HNO 3 (1%). All the analytical solutions were prepared using ultrapure water (resistivity ≥ 18.2 MΩ cm) obtained from a Milli-Q ® System (Millipore, Bedford, MA, USA). 15 mL of the resultant extract was filtered. The concentration of mineral elements was determined using an Inductively Coupled Plasma-Atomic Emission Spectrometry (ICP-AES) (5800 Agilent. Perkin Elmer Optima 3000) instrument, equipped with a crossflow nebulizer and an auto sampler. The detected elements in solution are expressed as mg g −1 DW.

Canopy Development
The total canopy biomass (stem plus leaves) produced each year was recorded during winter. The canopy biomass was collected from 10 cm above the root collar. This was dried in an oven at 65 • C for three days in order to record the total amount in g DW.

Statistical Analysis
One-way analysis of variance was performed by using "R" free software (version 4.0.3 2020-10-10) [41]. Statistical analysis to determine significant differences between treatment means was carried out using the Student-Newman-Keuls test (p ≤ 0.05), as already used for these kind of parameters and types of studies [42]. Changes between study years in the measured parameters were calculated as mean values as this can provide a good indication of agronomic performance. The differential was calculated as follows: ∆2020-2021: #value2021-#value2020.

Characterisation of the Study Sites
The incidence of KVDS at each site at the beginning of the experiment (2020) ranged from 30% (site-4) to 90% (site-2), whilst KVDS severity ranged from 30% (site-4) to 45% (site-2) ( Table 1). KVDS incidence was similar at sites -1, -2 and -3, and was only relatively minor at site-4. Nevertheless, the KVDS severity was similar at all sites. The average values of the physical characteristics of the soils are listed in Table 1. Following the international nomenclature of the United States Department of Agriculture, the soils are classified as silt-loam, are similar at all the sites and are typical of this part of Italy, apart from the C content. The study areas have a typical Mediterranean climate (Table 2), with warm summers and cool winters. The annual rainfall was approximately 1150 mm in 2020, with a mean temperature of 14.10 • C. In 2021, the annual rainfall was about 1200 mm, whilst the average temperature was 13.50 • C. The two years were climatically similar. This was confirmed by the global radiation: 4,926,780 KJ m −2 in 2020 compared with 5,018,202 KJ m −2 in 2021 (data not shown).

Genotype Survival in KVDS-Inducing Soils
The percentage of plants alive at the end of each vegetative season in the two study years is reported in Table 3. In 2020, all the plants of Ma176, Ma183, Bounty71 and MG survived, whereas only 16% of HW plants survived. None of the POL plants survived: they declined and died during the summer, 3-4 months after planting, so no further data were collected for these genotypes. Additionally, when the POL plants were removed, they had no roots, although at the time of planting in the field, all the genotypes showed similar levels of root growth in the pots (data not shown). Similar to POL in the first year, a serious decline in HW occurred during the second year of the study (the root system was heavily compromised and the plants were severely KVDS affected); hence, no measurements were taken for HW at this time. At the end of the two years, some MG plants also declined, with about 30% of the plants dying compared with none the first year.

KVDS Root Symptom Survey
The root symptom survey data are listed in Table 4. Ma176 and Ma183, 'Bounty71' and MG did not show any symptoms in 2020, whilst for HW, about 90% of the root system was compromised. In 2021, Ma176, Ma183 and Bounty71 continued to show no KVDS symptoms, but all the HW plants died, and in MG, an average of 25% of roots were symptomatic. Table 4. KVDS incidence in 2020 and 2021. Within columns, values that are assigned different letters are significantly different at α ≤ 0.05, 0.01, 0.001. Data are the mean of 16 replicates (4 for each site). Statistical analysis (ANOVA) was carried out using the Student-Newman-Keuls test (p ≤ 0.05). "-" = not measured.

Genotype KVDS Incidence (%)-2020 KVDS Incidence (%)-2021
Ma176  Table 5 reports all the root measurements from the two years of study and Figure S1 illustrates examples of the roots. In 2020, HW had the lowest mean root volume (56 mL), and Ma176 and MG had the highest volumes (407. 33 Between 2020 and 2021, the root volume of Ma176 increased slightly (13.67 mL) whilst Ma183 and Bounty71 showed the greatest increases (108.78 and 249.91 mL, respectively). Interestingly, the volume of MG decreased significantly (−266 mL), which might be correlated with KVDS attack. Thus, all A. macrosperma showed increased root volumes. Regarding root depth, only Bounty 71 increased (+4.86 cm), whereas the other genotypes seemed to decrease in depth. There was a parallel outcome for width, which increased in Bounty71 (+6.76 cm) and decreased in the remaining genotypes. The most marked growth occurred in the A. macrosperma group, and disregarding HW, the greatest decline occurred in MG during the second year of the study.

Protein Content in the Root System
The total protein content in the root system of the tested genotypes in the two-year study is reported in Table 6. In 2020, Ma176 showed the highest value (5.60 mg g −1 DW) and HW the lowest (2.84 mg g −1 DW), with this difference being statistically significant. The other genotypes had higher values than HW (ranging from 4.80 to 5.10 mg g −1 DW), but no strong statistical significance was found apart from the difference from HW. Indeed, in 2021, Ma176 once more recorded the highest value (7.97 mg g −1 DW), but again there was no statistical significance among the other genotypes. Bounty71 showed 6.55 mg g −1 DW, and MG was similar (6.48 mg g −1 DW), whereas for Ma183, the value was 7.01 mg g −1 DW. Lastly, concerning the changes between years, the more marked increase occurred in Ma176 (2.37 mg g −1 DW). The other genotypes showed a smaller increase (± 1.7 mg g −1 DW). There were no differences between genotypes in protein patterns ( Figure S2). Table 6. Genotypes protein content in roots in 2020 and 2021 and the changes between the two years. Within columns, values that are assigned different letters are significantly different at α ≤ 0.05, 0.01, 0.001. Data are the mean of 16 replicates (4 for each site). Statistical analysis (ANOVA) was carried out using the Student-Newman-Keuls test (p ≤ 0.05). "-" = not measured.

Accumulation of TNCs in the Root System
Concerning TNCs (Table 7), at the end of the 2020 season HW had the lowest value (13.33 mg g −1 DW) and was significantly different from the other genotypes. There were no differences among the other genotypes, where the values were around 22 mg g −1 DW. Again, in 2021, all the genotypes being tested showed similar values (around 30 mg g −1 DW). Unsurprisingly, the changes in TNCs in 2021 relative to 2020 were also similar among the genotypes tested (range: 7.82 to 9.98 mg g −1 DW). Table 7. Genotypes TNCs content in roots in 2020 and 2021 and the changes between the two years. Within columns, values that are assigned different letters are significantly different at α ≤ 0.05, 0.01, 0.001. Data are the mean of 16 replicates (4 for each site). Statistical analysis (ANOVA) was carried out using the Student-Newman-Keuls test (p ≤ 0.05). "-" = not measured.

Genotype
TNCs-2020 (mg g −1 DW)  Table 8 lists the mineral elements detected in the root system in 2020 and 2021. In 2020, HW had statistically lower values than the other genotypes for every element investigated. Apart from HW, no differences were found for nitrogen (N), carbon (C) or magnesium (Mg) between the tested genotypes. Nevertheless, Ma176 and Bounty71 had significantly higher phosphorus (P) contents (41.55 and 42.80 mg g −1 DW, respectively) than the other genotypes. At the same time, Ma176 also tended towards the highest calcium (Ca) value (7.56 mg g −1 DW), however, without statistical significance. In contrast, MG tended to have the lowest potassium (K) value (87.90 mg g −1 DW).

Mineral Elements Determined in the Root System
In 2021, no differences were found for the N, K and Mg contents. Significant differences were observed for P content with Ma183 having the highest value (39.79 mg g −1 DW). There were also significant differences in C content with Bounty71 and MG (666.42 and 662.96 mg g −1 DW, respectively) having lower values than Ma176 (738.11 mg g −1 DW) and Ma183 (719.24 mg g −1 DW). Lastly, Bounty71 had the lowest accumulation of Ca (5.13 mg g −1 DW).
Finally, regarding the changes between years (Table 9), every elemental content in each genotype was higher in 2021 than in 2020, with similar increases for each element across the genotypes. The most marked changes occurred in K and C accumulation.   Table 10 presents the DW of the canopy during 2020 and 2021. In 2020 there were significant differences among the varieties, with Ma176 along with Bounty71 possessing the greatest weights (23.53 and 22.22 g, respectively). Ma183 and MG had mid-range values (7.10 and 10.22 g, respectively), and HW had the lowest value (2.80 g). In 2021, once again, Ma176 and Bounty71 were similar and had the highest weight (27.08 and 25.90 g, respectively) and they were significantly different from Ma183 and MG, which had similar lower values (12.50 and 10.00 g, respectively). Between the study years, Ma176, Ma183 and Bounty71 showed similar small increases, whereas MG had a slight decrease in canopy DW, consistent with the other analyses, due to KVDS attack. Table 10. Dry weight of canopy for genotypes in 2020 and 2021 and the change between the two years. Within columns, values that are assigned different letters are significantly different at α ≤ 0.05, 0.01, 0.001. Data are the mean of 16 replicates (4 for each site). Statistical analysis (ANOVA) was carried out using the Student-Newman-Keuls test (p ≤ 0.05). "-" = not measured.

Genotype
Dry

Discussion
Some KVDS epidemiological traits are still under investigation, and at present there are no effective methods of control. Kiwifruit growers have empirically tried many strategies to maintain crop yields and their incomes [43], but these strategies are not sustainable nor effective. The lack of effective control strategies results in continually increasing economic losses [8]. The most promising approach, as with any soil-borne disease, is the use of tolerant or resistant species as rootstocks [24,44]. Within the Italian kiwifruit sector, this practice has not been widely used before. We have, therefore, screened genotypes of different Actinidia species intended for use as rootstocks or for breeding programs for their putative resistance or tolerance to KVDS [27,29,45,46].
In the current research, six genotypes of Actinidia spp. were planted at four infected experimental sites and tested for two years: A. macrosperma  (Tables 1 and 2). Consequently, the genotype behaviour recorded from each site was also similar. This allowed us to perform the experiment using four replicates per site, and this was confirmed by data normalization. In the statistical analysis, a one-way ANOVA was performed as a two-way ANOVA was not applicable because young plants grow quickly, so the year × treatment interaction would obviously be statistically significant (data not shown). Instead, changes between study years in the measured parameters were calculated as mean values because this provides a good indication of agronomic performance.
All the three genotypes of A. macrosperma, Ma176, Ma183 and Bounty71, survived in KVDS-inducing soil, suggesting that they might be useful tools for providing resistance to the syndrome (Tables 3 and 4). Furthermore, a genotype of known sensitivity, HW, proved to be susceptible and did not survive in the test soils (all plants were dead by the second year). For this reason, HW was not included in the second year of the study. Moreover, the POL genotypes died in the spring of the first year, indicating that this species is highly susceptible to KVDS. This result indirectly emphasizes the importance of biotic factors relative to environmental factors in KVDS aetiology [6]. This is supported by the observation that the parents of the A. polygama plants, as well as Hayward, used in the experiment are present in the germplasm collection of the University of Udine where KVDS is not reported (a few kilometers away from the experimental sites), and show excellent vegetative development and strong vigour, without any KVDS symptoms. The A. arguta Miss Green plants showed no symptoms of KVDS in the first year, but in the secondyear, root systems were affected (±25%), and consequently, many of the investigated root parameters were worse, with several plants dying. MG was affected at all experimental sites. This suggested that MG is not a resistant genotype but could be considered only tolerant. Consequently, where KVDS pressure was high, MG was not completely able to maintain a balanced root profile, nor was it fully capable of surviving.
A considerable range in kiwifruit root system agronomic parameters (biometric indices) was evident among the genotypes tested (Table 5). For A. polygama it was not possible to conduct any further analysis because all the plants died. Among the genotypes that survived the first year, HW had the worst performance. MG also performed poorly because by the second year it was affected by KVDS. In 2020, the best values were found in Ma176, and in the second year in Ma176 and Bounty71. Both Ma183 and Bounty71 showed strong increases in some indices (e.g., root volume) in the second year of the study, and Bounty71 in particular attained values that were similar to Ma176 during 2020. Bounty71 showed increases in all values. MG showed decreases in all the parameters due to KVDS. However, the small decreases in width and length of Ma176 and Ma183 were not significant.
The effect of soil-borne pathogens on protein degradation was clear ( Table 6) when considering the protein content in the resistant genotypes compared with the sensitive genotype (HW), as has been reported previously [47]. Unsurprisingly, HW showed the greatest reduction in protein content, even when accounting for the data from the second year, with the remaining genotypes not noticeably affected. Once again, Ma176 tended towards the highest protein content; however, no statistical differences emerged among the tested genotypes. As previously discussed, there were no differences in protein patterns observed ( Figure S2).
Concerning the TNCs content (Table 7), HW was the worst performing, further indicating its KVDS susceptibility. This parameter was clearly affected by KVDS and has known susceptibility to several biotic factors in many crops [48]. Although the other genotypes tested showed higher values than HW, there were no statistical differences between any of the genotypes, or for the change in TNCs between years among the surviving genotypes. Nevertheless, we may assume from this assay that the genotypes that performed well will be able to accumulate more reserves in subsequent years, which is important for plant budburst, and flowering, etc. [49,50].
The mineral elements detected present in the root system (Tables 8 and 9) showed some genotype specific effects. It is of great importance to understand their accumulation in the roots since they are further translocated to the aerial parts of the plant (i.e., leaves and fruits). At first glance, HW had significantly lower values for each element investigated than the other genotypes. However, no great differences emerged between the resistant genotypes. Nonetheless, in 2020, Bounty71 recorded the lowest Ca accumulation value. Calcium is involved in determining fruit size, firmness and sugar translocation [51,52], so we can speculate that this rootstock might eventually result in fruit size and quality reductions. During 2021, the trend continued, with Bounty71 accumulating less Ca and C than most of the varieties (the C levels were similar to MG). Indeed, the fact that Bounty71 recorded the lowest C concentrations among the surviving genotypes may mean that this cultivar is not capable of further accumulation of reserve substances or of maintaining a well-balanced whole plant profile in the long term [53]. Regarding MG, despite it being affected by KVDS in the second year, it was surprising that the levels of mineral elements present in the root system tended to increase. We can assume that this was a concentration effect because the root volume of MG decreased in the second year of the study. Interestingly, the P content for MG in 2020 alone and for Ma183 in both 2020 and 2021 was significantly lower than that for the other genotypes. The P content should be carefully taken into consideration when selecting rootstock, because P is a component of essential molecules such as nucleic acids, phospholipids, and ATP (adenosine triphosphate), and is involved in controlling key enzyme reactions and the regulation of metabolic pathways [54]. Finally, Ma 176 seemed to have the highest elemental concentration of all the genotypes, as did Bounty71 for the majority of elements, but these values were not statistically significant. Similar results were just obtained in preliminary experiments [55,56].
Considering canopy development measured in terms of dry weight (Table 10), in both 2020 and 2021, Ma176 and Bounty71 had the highest values overall. In 2020, HW had a significantly smaller canopy than the other genotypes, and MG and Ma183 registered lower values than Ma176 and Bounty71 in both years. Perhaps owing to the effects of KVDS, canopy production in MG decreased in 2021. Finally, the changes in canopy dry weight between 2020 and 2021 were similar. We can report that the genotypes tested showed an interesting and different canopy development and appearance, which could be important for plant structure once grafting occurs, since the aptitude to develop canopy is an important factor for the whole aerial plant structure [57]. In fact, a different canopy structure not only affects the aerial plant structure, but the biodiversity of the cultivated land as well (e.g., light competition is one of the most important factors that limits both the growth of plant species and the biodiversity of the soil under kiwifruit canopy) [58].
As a final consideration, A. macrosperma and A. arguta showed better performance than HW and POL in every analysis carried out. Obviously, all the genotypes had different specific effects, but in the end, the majority demonstrated KVDS resistance.

Conclusions
In this paper we are reporting detailed evidence of resistance or tolerance to KVDS in Actinidia genotypes which can be used as rootstocks, once their affinity with the most important cultivars can be demonstrated, or as a source of resistance or tolerance in breeding programs. A. macrosperma genotypes are the most promising of those tested and, at present, they are the only way to overcome KVDS. Further studies will be required to understand how and why these genotypes are resistant or not, and experiments on these and other aspects are currently ongoing under controlled conditions (e.g., ability to modify the microbiota). In addition, we will further investigate the underlying genomics and the grafting long-term compatibility.