1. Introduction
Kiwifruit,
Actinidia chinensis (Planch., 1847) originated in China and it has been domesticated in New Zealand from where it has become a global crop [
1]. Kiwifruit market is dominated by two major varieties, the green-flesh (
A. chinensis var.
deliciosa) and yellow-flesh cultivars (
A. chinensis var.
chinensis) [
2,
3]. Despite kiwifruit cultivation is very successful, and it is widely spread with an increase of total fresh fruit from 577 to 887 Mt/year in the last 20 years [
4,
5], this crop is facing several phytoiatric problems. Among them,
Pseudomonas syringae pv.
actinidiae [
6], kiwifruit vine decline syndrome, KVDS [
7] and
Halyomorpha halys [
8] pose the highest risks.
In Italy, more than 45% of the kiwifruit production is concentrated in Lazio (specifically in the Agropontino, Latina province), followed by Emilia-Romagna (in particular Ravenna province), which accounts for 15% of national production. Among the other Italian regions, the most important are Veneto and Piedmont (with approximately 10% each of the Italian kiwifruit production), Calabria and Campania (together with another 13% of the total national production) [
9]. The Italian kiwifruit production covers a crucial economic importance for both the national market and international export, mainly in Europe [
10]. Given the value generated by this crop for the Italian growers and for the whole kiwifruit industry, it is necessary to guarantee high quality standards while minimizing the inputs of plant protection products [
11]. In 2020, China produced about the 50% of the global kiwifruit production, whereas New Zealand and Italy accounted for 14% and 12%, respectively [
12]. Prior 2020, Italy ranked second in total kiwifruit production, ahead of New Zealand and this shift in global ranking could be related to the adverse climatic conditions in Italy in the recent years and to the rise yield losses due to several factors, including KVDS and
H. halys.
The brown marmorated sting bug,
Halyomorpha halys (Stål) (Hemiptera: Pentatomidae) is a very concerning invasive insect species that caused important economic losses in most countries outside its Asian Native range [
13,
14,
15,
16]. Major damage is reported mainly on tree fruit crops, including kiwifruit [
8,
17,
18,
19]. This pest was introduced in Europe in the early 2000s, likely with international trades. In Italy the first record dates back to 2007 in Liguria [
20], but about ten years ago the first region where economic losses were reported was Emilia-Romagna [
21,
22]. In kiwifruit crop, the feeding punctures of
H. halys cause internal white corking damage to the fruit flesh, which is visible only peeling the fruit skin [
23,
24] and in some cultivars the fruit injuries can lead to an early drop [
25].
The
H. halys management requires the adoption of several complementary control methods to be applied with an integrated approach [
16]. The first pillar is to create an ecological equilibrium, with the implementation of the biological control, mainly carried out by egg parasitoids [
26,
27,
28] that were recorded also within kiwifruit orchards [
8,
29]. The use of exclusion netting systems also proved to be effective in a number of crops towards several insect pests, including
H. halys [
30,
31,
32], and it can reduce fruit damage also in kiwifruit orchards [
33]. Then, the practical implementation of chemical ecology knowledge, exploiting for instance synthetic
H. halys aggregation pheromones [
34], had led to the creation of monitoring networks at regional scale with a shared alert system, the issuance of weekly bulletins and the development of a forecasting model in order to properly time the insecticide applications against this pest [
35].
However, the chemical control of
H. halys using insecticides remains an indispensable and crucial component for the successful management of this pest. The achievement of a satisfactory chemical control of
H. halys is very challenging, because of the characteristics of this pest (
e.g., polyphagy, mobility across different crops, recovery ability when exposed to sublethal doses of insecticides) and to the features of the available effective active substances (
e.g., short-time residual activity, lack of selectivity towards non-targets) [
36,
37,
38,
39,
40,
41]. Organic insecticides have shown some promising results under laboratory conditions [
42], but insecticides authorized on kiwifruit crop in organic farming (
e.g., pyrethrins) do not match the performances of synthetic insecticides (
e.g., etofenprox, deltamethrin) [
43]. In addition, the European legislation is moving toward the reduction of synthetic plant protection products, with the ban of several active substances and numerous limitations in the use of the authorized ones [
44,
45]. Therefore, the optimization of all the available management practices for
H. halys, considering the above-mentioned factors, is crucial.
Especially the application of insecticides needs to be carefully rationalized and optimized, aiming at coupling the maximum efficacy with the best cost-effective convenience. Besides that, from a sustainability standpoint, also the environmental and social impacts need to be properly evaluated in order to guarantee the lowest possible impact of these practices on the agroecosystems and the minimum possible residues on fruit at harvest. Furthermore, due to the peculiar training systems used on vines, plant protection products distribution in kiwifruit orchards is challenging. Indeed, kiwifruit vines are generally trained in pergola or T-bar trellis systems [
46] and in these conditions canopy size can reach 50,000 m
2/ha [
47]. Furthermore, the pergola system hampers air circulation, increases humidity inside the canopy and may reduce the efficient penetration of the plant protection products. In this context, the technique of the insecticide distribution within the crop canopy and the quantity reaching the target areas (
i.e., where the fruits are localized, to protect them from the
H. halys feeding activity) are crucial factors.
This study was aimed at comparing two insecticides spray techniques to control H. halys in two kiwifruit cultivars in the Emilia-Romagna region, Northern Italy. The specific objectives were to:
i) evaluate the mortality over time caused by pyrethrins/pyrethroids applied either with a conventional ray atomizer or a localized spray technique on H. halys adults;
ii) assess whether the different spray techniques had any effects on fruit damage caused by H. halys at harvest, considering the same grower insecticide spray program.
2. Materials and Methods
2.1. Spraying Machineries
In this study, the conventional ray atomizer (
Figure 1A) was compared to a localized spray technique where the insecticide product is delivered by means of trumpets that restrict the spray range in the area where fruits are located (
Figure 1D). The trumpets-modified atomizer does not differ in the body structure from the conventional ray atomizer, sharing the chassis and all the electronical and mechanical components. The difference between these spray machineries is in the area where the air flow exits and in the position of the nozzles; therefore, in the delivery trajectory of the plant protection products. In the localized spray, there is a metal deflector positioned in the area from which the air generated by the fan exits, conveying it through plastic pipes on which end nozzles are mounted. In the trumpets-modified atomizer the nozzles are mounted on a nozzle holder positioned on the side of the fan and connected to the frame via a metal structure, forming a so-called ‘trumpet’ (
Figure 1E).
The trumpets-modified atomizer is usually utilized by kiwifruit growers for artificial pollination during blooming, providing a localized air flow in the area of the canopy where most of the flowers occurs. In all the experiments, the same atomizer (model Vulcano Georgia 2000, Vulcano srl, Faenza, RA, Italy) was used, considering the conventional application technique when using the full ray operative (with 10 nozzles model ATR Albuz 80, of which 2 green, 6 grey and 2 red, for a nominal flow rate of 29.5 L min-1) and the localized system technique when the trumpets were installed on the atomizer (with 8 nozzles model ATR Albuz 80, all green, for a nominal flow rate of 27.4 L min-1). The same operative parameters (operational pressure 20 atm, speed 7.5-8 km h-1, and distributed water volume 800 L ha-1) were kept in each application and in all the experiments for both the insecticide spray techniques.
2.2. Study Sites and Biometric Plant Parameters
Experiments were carried out during 2021 and 2022 simultaneously in the yellow-flesh and the green-flesh kiwifruit cultivars ‘Jintao’ and ‘Hayward’, respectively. All the experiments were carried out in a farm located in Castel Bolognese (RA) (44°19’36.74”N, 11°47’38.17”E), where two neighboring commercial kiwifruit orchards (< 20 m apart) were managed by the same grower following the GAP (Good Agricultural Practices) and according to the IPM (Integrated Pest Management) guidelines. In the area,
H. halys infestations naturally occurs every year since 2015 [
8,
22].
The orchard of the cultivar ‘Jintao’ (0.5 ha) was planted in 2007 with a planting space of 4.5 m between rows and 1.5 m between plants (density: 1481 plants ha-1). The orchard of the cultivar ‘Hayward’ (1.2 ha) was planted in 2018 with a planting space of 4.6 m between rows and 2.0 m between plants (density: 1087 plants ha-1). Both cultivars were grafted on ‘Hayward’ rootstock and the orchards had the same density of male plants (1 every 5 female plants). In both orchards, the pergola training system was used and vines were irrigated by a drip system. The ‘Jintao’ orchard had an anti-hail net cover combined with lateral nettings (i.e., monoblock netting system) to protect the crop from both hail and insect pests such as H. halys, while no nets were installed in the ‘Hayward’ orchard. The monoblock netting system is usually deployed in this orchard cultivar ‘Jintao’ after blooming, at the end of May.
Despite the same training system was used, the differences between the cultivars ‘Jintao’ and ‘Hayward’ in terms of vigor and pruning technique affect the canopy structure and density, and could impact the outcomes of insecticide applications performed with different spray techniques (
Figure 2). In particular, the green-flesh cultivars are usually pruned leaving a higher number of productive shoots and longer shoots compared to the yellow-flesh cultivars. Therefore, the main biometric plant parameters of the two cultivars under study were recorded at the end of September 2021 (crop BBCH 79-81). Briefly, a random sample of 10 representative plants per each cultivar was selected within each orchard to count the number of productive shoots per plant. The length of all the shoots per plant was measured and the total number of leaves in each shoot was counted; then, in one randomly selected shoot per each of the 10 plants, all the leaves were collected and scanned to measure the leaf area by means of the image processing software ImageJ [
48].
2.3. Halyomorpha halys Mortality Experiments
Halyomorpha halys individuals used for these experiments were field collected in unsprayed agroecological contexts and were previously kept in controlled conditions in the entomological laboratory of ASTRA Innovazione e Sviluppo (Faenza, RA, Italy), where they were fed with fresh organic vegetables (carrots, green beans and tomatoes), sunflower and soybean seeds, and a cotton ball soaked with water. Halyomorpha halys individuals were held at a 12:12 light:dark photoperiod, a temperature of 25 ± 1 °C with a 60 ± 10% relative humidity prior to the field exposure in the mortality experiments.
To assess the insect mortality in function of the spray techniques, namely conventional ray and trumpets-modified atomizer, an artificial infestation procedure was performed during 2021-2022 seasons. The experiments were conducted following the methods described in Preti et al. [
43] and according to the EPPO guidelines [
49]. Three
H. halys adults (sex ratio male:female = 1:2) were placed in mesh cages (20 cm × 30 cm, mesh size < 0.5 mm), which were wrapped around a single kiwifruit.
Each experiment was carried out following a block design with three replicates. Ten mesh cages were installed approximately in the center of each plot (30 H. halys individuals per replicate and 90 individuals per treatment). Cages were deployed on different plants and randomly selecting fruits with a variable degree of canopy coverage, to simulate the pest occurrence within the orchard. Each experimental plot comprised a minimum of 4 neighboring rows and 10 consecutive plants within each row, for a minimum area of approximately 350 m2. All plots were located in the central part of the orchard at a minimum distance of 50 m from the borders in order to avoid perimeter-driven effects. The insect mortality was assessed at 1, 3, 7 and 10 days after the insecticide sprays, recording the number of alive and dead H. halys individuals in each mesh cage. The natural mortality was assessed exposing groups of three H. halys adults in the field (within mesh cages) in plots unsprayed with insecticides (three control plots per experiment).
The mortality study considered a total of six replicated experiments (three per each kiwifruit cultivar). A single insecticide application was carried out per each experiment, using either the commercial product Asset Five® (a.s. pyrethrins 46.5 g L-1) applied at 0.96 L ha-1 or Decis Evo® (a.s. deltametrhin 25 g L-1) applied at 0.50 L ha -1. In the cultivar ‘Jintao’, experiment #1 was carried out with deltamethrin on August 30th 2021 (crop BBCH 77-78), experiment #2 was carried out with pyrethrins on September 21st 2021 (crop BBCH 79-81), while experiment #3 was carried out with deltamethrin on September 26th 2022 (crop BBCH 79-81). In the cultivar ‘Hayward’, experiment #4 was carried out with pyrethrins on September 21st 2021 (crop BBCH 78-79), experiment #5 was carried out with deltamethrin on October 15th 2021 (crop BBCH 79-81), while experiment #6 was carried out with deltamethrin on September 26th 2022 (crop BBCH 78-79). Insecticide application timings and active ingredient choice replicated what Italian kiwifruit growers usually do to control H. halys infestations, following the products labels.
2.4. Injury Assessment on Kiwifruit
The impact of the insecticide application technique on the crop damage was measured on kiwifruits at commercial harvest in 2021. The cultivars ‘Jintao’ and ‘Hayward’ were managed following the same insecticide spray program. The two spray techniques (standard ray atomizer and trumpets-modified atomizer) were compared with an unsprayed control. The same plots used to test the insecticide efficacy with the artificial infestation of H. halys in mesh cages were sampled to score fruit damage caused by H. halys individuals naturally occurring in the study orchards. For the cultivar ‘Jintao’, the two insecticide applications were carried out in August and September 2021 with deltamethrin and pyrethrins, respectively, whereas for the cultivar ‘Hayward’, pyrethrins and deltamethrin were sprayed in September and October 2021, respectively (see experiments #1-2 for cultivar ‘Jintao’ and experiments #4-5 for cultivar ‘Hayward’).
At commercial harvest (crop BBCH 85-87), on October 19th 2021 for ‘Jintao’ and on October 30th 2021 for ‘Hayward’, samples of 100 randomly selected fruits per replicate were collected to assess the fruit injury level (300 fruits per treatment). These fruits were not collected from the same plants where the mesh cages were installed for the mortality study, previously flagged to be excluded from this study. After a 2-month period of cold storage at 4 ± 1 °C, each fruit was completely peeled to detect any internal white corking due to the
H. halys feeding punctures and the number of injuries were recorded per each fruit (
Figure 3). The fruit injury incidence was calculated as percentage of damaged fruits out of the total number of observed fruits, while the fruit injury severity was calculated as mean number of feeding injuries (
i.e., internal white corking sites) per fruits considering all the observed fruits.
Finally, a monitoring trap (model Dead-Inn
® pyramid, AgBio Inc., Westminster, CO, USA) baited with
H. halys aggregation pheromones (model Pherocon
® BMSB lure, Trècè Inc., Adair, OK, USA) was placed at approximately 100 m from the two kiwifruit orchards to record the
H. halys population dynamics. The trap was checked from mid-May (crop BBCH 69-71) until the end of October 2021 (crop BBCH 85-87). The number of
H. halys catches, separating adults and nymphs, was weekly recorded removing and killing all the caught individuals (
Figure 4).
2.5. Statistical Analyses
Generalized linear mixed models (GLMM) with a binomial error distribution, a probit link function and an autoregressive covariance structure of the first type were used to test the effects of treatments (conventional atomizer, trumpet atomizer and untreated control) on the mortality of H. halys adults exposed within mesh cages. Days of exposure of the insects to the treatments were included as repeated measures, and the interaction “treatment x days of exposure” was tested as well. Mesh cages were nested within blocks and the Kenward–Roger method was used to estimate the degrees of freedom. Sequential Bonferroni methods was used to compare the levels of factors with significant effects setting the p level at 0.05. Separate models were run for each kiwifruit cultivar (‘Jintao’ and ‘Hayward’) and for each experiment because different insecticides were sprayed and the field trials were carried out in different periods of the seasons.
A factorial ANOVA model was used to test the effect of treatments (conventional atomizer, trumpet atomizer and untreated control) and the kiwifruit cultivar (‘Jintao’ and ‘Hayward’) on the mean number of H. halys feeding punctures per fruit. The interaction “treatment x cultivar“ was tested as well. Raw data were log transformed before running the model to match ANOVA assumptions and Tukey HSD test was carried out for multiple comparisons of the three levels of the treatment factor setting the p level at 0.05.
The biometric plant parameters are reported as descriptive statistic, considering the mean values ± standard deviation.
The software package IBM SPSS Statistics (ver. 26) was used for all the analyses and for graphical representations of data.
4. Discussion
The insect mortality recorded in this study testing
H. halys adults confined in mesh cages was in line with previous data collected with the same experimental design [
43] and lower than what observed with a natural infestation in open field [
50,
51]. The insects received a topical insecticide application and were forced to stay in a contaminated environment; they were, therefore, fully exposed to the residual activity of the plant protection products under study. This experimental approach may lead to an overestimation of the real insect mortality, but allows to evaluate the relative efficacy of the treatments under study with a comparable experimental set up [
49].
Considering the insect mortality recorded with the two spray techniques under study, in the cultivar ‘Hayward’ the trumpet atomizer resulted in a higher efficiency than conventional ray atomizer. However, in the cultivar ‘Jintao’ little or no difference emerged between the spray techniques. This result can be explained by the leaf area and canopy density in this cultivar. In fact, in comparison with ‘Hayward’, ‘Jintao’ had a lower number of shoots, which were shorter and with less leaves. In addition, the cultivar ‘Hayward’ had 1.3-fold wider leaves with an overall 2.9-fold higher total leaf area per plant compared to the cultivar ‘Jintao’. These differences in terms of canopy structure, shape and size influence the insecticide distribution using either the conventional ray atomizer or the trumpets-modified atomizer in the green-flesh cultivar ‘Hayward’. This effect was more evident using the synthetic pyrethroid, likely because of the higher activity against
H. halys compared to the natural pyrethrins [
43,
51].
According to the
H. halys monitoring trap catches (
Figure 4), the pest pressure started to increase from mid-September reaching a peak of adults in the first half of October, 12 and 23 days prior to the ‘Jintao’ and ‘Hayward’ kiwifruit harvest, respectively. The insecticide strategies provided a sufficient, but not fully satisfactory damage control, with an Abbott efficacy close to 60% for the cultivar ‘Jintao’ and close to 65% for the cultivar ‘Hayward’ in terms of damage incidence reduction compared to the untreated check.
Overall, the
H. halys damage incidence was higher for the cultivar ‘Hayward’, and the green-flesh fruits were significantly more injured (with a higher damage severity) compared to the yellow-flesh fruits. It has to be noted that the cultivar ‘Hayward’ was harvested 11 days after the cultivar ‘Jintao’ and therefore remained exposed to the feeding activity of
H. halys for a longer period. In addition, the ‘Hayward’ orchard was not covered with nets, resulting in a full exposure to the
H. halys infestation that in the ‘Jintao’ orchard was limited by the anti-hail net combined with lateral nettings. Likely, the monoblock netting system usually applied in the yellow-flesh kiwifruit orchard helped to physically control
H. halys [
30,
31,
32,
33]. Interestingly, these results are in line with Francati et al. [
8], who showed how the kiwifruit injury increase across the season and is higher in late summer. The fruit injury results from our study do not allow to discriminate a difference between the spray techniques, while both insecticide strategies significantly reduced the fruit injury severity compared to the untreated control. Likely, other factors involved in this field study, including the high pest pressure recorded within the experimental site and the spray timing of the insecticide applications, could explain this result.