New Tools for Mechanical Thinning of Apricot Fruitlets

Abstract

In this study, the thinner machine with yellow rod equipment was tested in relation to tree branch length and orientation in April 2019, in a narrow-canopied apricot orchard of Emilia Romagna Region, Italy. The trees were mechanically thinned with manual finishing, and comparative tests were carried out simultaneously with the ordinary hand thinning (control). Three groups of two plants were identified as replication for a total of six plants per row. Three rows were checked, considering field uniformity average. The branches were grouped into four classes according to their length: <30 cm, 30–60 cm, 60–90 cm and >90 cm. Branch inclination on the plant, radial or longitudinal with respect to the row, was evaluated. Fruit number before the thinning, after the first and the second machine intervention, after three days of the mechanical thinning and after the hand finishing was recorded. This experience showed satisfactory results in terms of thinning efficiency and reduced damage to both fruits and branches, as a function of the class length and insertion point in the main branch of the plant. Thinning efficiency was always kept above 37% of the left load after hand finishing, and on average between the treatments close to 44%. Fruit damages always remained below the economic thresholds to marketable production or to the plant.

1. Introduction

Apricot (Prunus armeniaca L.) is an important fruit crop grown worldwide, and is a highly appreciated fruit by consumers [1]. In apricot tree cultivation, thinning is a central cultural practice to regulate the fruit load, to improve fruit size, yield, and quality, and to decrease alternate bearing [2,3,4]. In commercial orchards, the excess fruits are manually removed around 40–60 days after full blooming (DAFB), to produce large size fruits for market [5,6]. Thinning responses depend on environmental and soil conditions, on timing, and on management practices, such as pruning. Pruning is the earlier method to adjust crop load [5]. In some species/cultivars, reproductive and vegetative performance is related to thinning severity and timing [7,8,9].

Thinning can be carried out in pre-bloom (e.g., floral buds), during bloom (e.g., flowers) or post-bloom (e.g., fruits/fruitlets), using different methods, such as hand, mechanical, and chemical thinning [10]. In apricots, hand thinning is still the standard method, although it requires a high level of labor force in a limited time, thus being very expensive [5,6]. Hand thinning is a costly and time-consuming management practice, and nowadays it is difficult to find qualified manpower to manually thin fruit crops and the farm labor cost is increasing. For hand thinning, labor load was estimated to be 31% of all cultural costs associated with cling peach production [11]. Chemical thinning is the most cost-effective method, but the advantages vary by crop [12,13], and phenological stage [14,15]. Chemical bloom thinners were more effective when more flowers were open, and more pistils were exposed to the chemical treatment [16,17]. Furthermore, flowers are more susceptible to surfactants than flowers at tight bloom or in balloon stages [18].

Christen et al., 2010 [19] showed that ammonium thiosulfate and 6-benzyladenine applications on apricot trees did not exhibit phytotoxicity, and high thinning efficiency was observed in many tested cultivars, except for ‘Kioto’. Spraying GA in the first week of June decreased flower populations in the following season and the need for hand fruit thinning was reduced in ‘Patterson’ apricot, improving quality, without phytotoxic effects [20]. Thinning with Metamitron was effective on ‘Sensação’ peach trees if realized up to 40 DAFB [21]. Weather conditions also influenced the result. A negative correlation was found between maximum temperatures and the effectiveness of thinning in the case of NES and Ned, and a positive correlation in the case of Rol-fruct [22].

Several mechanical thinning devices have been evaluated in stone fruits [4], from electro-mechanical limb shaker [23], to Horizontal String Blossom Thinner [24,25], and vertical rotor machine, to thin blossom and fruits [26,27]. There have been varying degrees of success, linked to the tree training system, tree architecture, and pruning intensity [28,29,30,31]. Narrow canopy training systems and novel apricot tree growth habits offer new opportunities to examine mechanical methods for apricot thinning. Various bloom mechanical thinning trials were set up on peach trees, testing a German mechanical device originally designed to thin apple flowers [24,28]. In Italy, first trials were realized in Piedmont, North-West Italy, due to the wide spread of peach orchards trained to narrow canopy systems, that allow optimal string penetration within the canopy [25,32]. However, the adoption of mechanical solutions cannot be satisfactory applied without wise pruning, in the ordinary management of the fruiting wall. In last years, CREA carried out evaluation tests of a vertical rotor machine with blue radial bars produced by a French company (La Canne Vale, model ECLAIRVALE), on various stone fruit [26,28]. For stone fruits, this experience immediately highlighted weak points on operating methods, related specifically to tree phenology, given that in some peach varieties the thinning technique has a good efficiency only in a narrow timeframe. First specific tests to evaluate the detachment force for stone fruit [31] and to test the thinning machine for peach and apricot [26,27,33,34] were carried out. This original version of the machine has been significantly implemented over the years at operational level especially in terms of rod types, material, and dimensional aspects. In the present study, the ECLAIRVALE^®1600 was tested in an apricot orchard, to evaluate its thinning efficiency with regard to tree branch length, to orientation of insertion points in the main branch, longitudinal (L) or perpendicular to the row (P), thus oriented towards the row center.

2. Materials and Methods

2.1. Experimental Site

The research was carried out in a narrow-canopied apricot orchard, cv Carmingo^® Farbaly, located in Chiesuola, Russi (Ra), Italy at the Zani Agricultural Company (44°20′11.56″ N, 12°02′28.13″ E). The trees were 8 years old at the beginning of the trials, the training system was a central leader, with a canopy height 3.9 m, spacing row, 1.9 × 3.9 m (1.350 tree ha⁻¹). At mechanical thinning time, 26th April 2019 (104 days from fruit ripening), the young fruits (45 DAFB) had an average weight of 14.5 g, longitudinal diameter of 35 mm and equatorial diameter of 25 mm. Comparative tests were carried out simultaneously with the ordinary hand thinning (control), by recording the working times per row, the number of plants for rows, and the workers employed, thus calculating the manual finishing time per plant and the fruit load left at the end of mechanical and manual thinning.

The original fruit load was estimated by splitting four plants into four sectors in height per side and proceeding with the relative counting of the fruits before the thinning operations. Within the rows, three groups of two plants were identified (replication) for a total of six plants per row. Ten representative branches of each plant were identified and for each of them the diameter and length were measured. The branches were grouped into four classes according to their length: <30 cm, 30–60 cm, 60–90 cm and >90 cm. Branch inclination on the plant, radial or longitudinal with respect to the row, were evaluated. Fruits number before the thinning, after the first and the second machine intervention were counted. Previous CREA’s experiences [26,27] showed the possibility that flowers or fruits, may be damaged, and subsequently fall, hence the plants were monitored a few days after the treatment. Fruits in branches were counted three days after the mechanical thinning and after the hand finishing. Furthermore, damages of remained fruits were assessed, and twenty fruits for each tree were collected and characterized. The measurements were carried out in CREA’s lab with a precision balance (Mettler Toledo, model PM460, Milan, Italy) and a digital caliper (Borletti, model CDJB15 Antegnate (Bg), Italy), for size (weight, length, and width), to evaluate side effects of the working elements.

2.2. Thinning Machine

The trial was carried out with a rear semi mounted thinning machine. The machine is composed by a main metal frame with the couplings at the front for connection to the three-point linkage of the tractors, a pair of support at the bottom from recovery and the idle tool holder rotor in a central vertical position with hydraulic system for setting. The free wheel thinning machine does not require any other power system and shows a single vertical rotor (Figure 1).

The machine is composed of a vertical rotor on which 1797 flexible working tools (rods), with a length of 1.37 m, are radially inserted with a nut spring system. The rods are available in three main version characterized with materials, dimensions, and colors (blue, green, yellow). The yellow rods are slightly stiffer than the green ones. The first machine was equipped with blue rods in version ECLAIRVALE 2500 with 2808 rods; the second machine has a working height of 1.6 m, to reduce the overlapping in the central part of the canopy.

The rods are made of flexible glass fiber with soft plastic end caps. They have the function to penetrate the foliage and to detach fruits with a transverse movement of friction. The machine can be also utilized for flower and fruit harvest for industrial purpose (juice and pulp). The speed of rotation is defined by the rods’ contact with the tree canopy. The selected working speed, as defined by previous tests, was of 1.61 m s⁻¹.

Two hydraulic connections with 10 L min⁻¹ flow and at least 190 bar pressures are required to set the position of the machine. The thinning effect is mainly caused by rods input/output to the canopy for direct contact with flowers and fruits. The machine can be adapted to be both semi-mounted and mounted, and to adequately perform the thinning operation, the total mass of the tractor must guarantee the stability of a 3 m rear overhang. The thinning machine has a working surface of 1.6 m, 0.9 m less than the first CREA’s test, to avoid double or multiple passages, the relief areas mainly concerned the upper, middle, and bottom part.

The trial was conducted in two passages of machine and manual finishing thinning. Hand finishing operation was carried out with the help of a self-propelled fruit truck with extendable platforms; however, for this trial, the hand working time was monitored excluding the hourly cost of the support machine.

Machine performance was evaluated by measuring working time in accordance with the ASAE machinery management [35,36]; other references include the “Commission Internationale pour l’Organisation Scientifique du travail en Agriculture” (CIOSTA) and the Italian Society of Agricultural Engineering (AIIA) 3A R1 [37]. The effective field capacity was determined and expressed by area capacity as in the previous experience [27].

At harvest time, the total and marketable yield per row were evaluated according to the current commercial standard.

2.3. Statistical Analysis

All recorded data were analyzed using analysis of variance procedures (free software PAST version 2.12, Øyvind Hammer Natural History Museum, University of Oslo). Fruit number before and after thinning and fruit removal percentage, were the variables; shoot length classes and orientation were the factors analyzed. Before the analysis of variance, fruit removal data, expressed as percentages, were transformed into arcsine-square-roots values. Mean separations were performed by using the post-hoc Duncan test at p ≤ 0.05. Data are expressed as the mean ± standard deviation (SD).

3. Results

In order to choose the working speed, during both the setting phase and the test phase, the thinner machine worked without clogging or other problems. The main parameters of work performance are shown in

Table 1. Working times and performance of the machine.

Parameter	Unit		Value
Working speed	m s−1		1.61
Theoretical area capacity	ha h−1		1.13
Effective area capacity	ha h−1		0.95
Field efficiency	%		84.40

.

The average fruit load of orchards is quite different, considering the row orientation (from south-west to north-east): 866.45 ± 89 from the south-side and 983.75 ± 39.24 from the north-side for about 1900 fruits per tree, mainly in the upper part.

The effect of mechanical thinning was immediately visible in certain parts of the trees, given that the machine’s effects were diversified, depending on branch length, orientation, and position within the canopy. Regarding branch length (Figure 2), the effect of mechanical thinning was greater on the branches 30–60 cm-long (52% of fruit detected). The smaller branches (0–30 cm-long) were less susceptible to the effect of the thinner machine and hand finishing thinning was greater for this category. The effects on middle-class branches (30–60 cm and 60–90 cm) were diversified; in some cases, the percentage of fruits removed by the machine was comparable to the hand removed, in some cases it removed much more, and in a few cases, it could not remove them, and this is an aspect to be further investigated.

Regarding the branch orientation, those arranged longitudinally were higher to those arranged perpendicularly, and a higher percentage of fallen fruit (50% on average) was detected after the two mechanical interventions in the longitudinal branches (Figure 3). Longitudinal branches compared to perpendicular ones, with respect to the main stem of the plant, were generally more susceptible to the effect of the machine, but no statistical difference was found.

Considering both branch orientation and length classes, the machine detached more fruits, thus being more efficient, in the branches arranged longitudinally of 30–60 cm- and 60–90 cm length classes, an increasing percentage proportional to the length of the branch, until 60–90 cm-length branches. Indeed, in the branches >90 cm-length, there were no significant differences between mechanical and hand thinning. In 0–30 cm-length branches, the effect of the machine was limited both in the longitudinal and in perpendicular ones. The manual thinning on the shorter longitudinal branches was much more marked (Figure 4).

Analyzing the two machine passages and the damaged fruits that dropped 3 days-after the treatment, the machine effect was lower in the first class of branch length and progressively increased proportionally to length (Figure 5). Indeed, fruits arranged in the branches of 60–90 cm-length were more affected by thinning and the first passage was the more effective. In the branches of 30–60 cm length, the second passage showed the best performance.

Examining separately the two mechanical passages and considering the branch insertion on the main stem, the main effect was noticed on the longitudinal branches at the second passage. This effect tends to equalize between first and second passage of the thinner machine in the branches arranged radially (P). The fruits damaged by machine passages and dropped naturally within 3 days were mostly characterized in longitudinal-perpendicular branches, given the lower efficiency of the thinning machine in this case (Figure 6).

The mechanical thinning led to the reduction of 45% of the total fruit load, allowing only 32% for manual finishing. The initial fruit load, after the mechanical and hand thinning, was reduced; of the 77%, only the 23% fruits remained for the harvest. In the area treated only manually, helped by a self-propelled platform, the work needed was 52 h ha⁻¹ while the manual finishing operation after the mechanical thinning needed 18 h ha⁻¹ with a 2.8-fold reduction with respect to only the manual thinning.

4. Discussion

The main novelty of this thinner machine was the new rods, characterized by hardness, diameter, and elasticity in color classes from blue (low), green and yellow (high), with the aim to vary their elasticity and to increase their external friction towards fruits; this aspect seems to allow the use of the machine at higher leaf moisture levels. The previous experience of CREA on peach thinning [27] was implemented with some changes in dimensional aspects, height control system, construction materials, diameter, and length of working bars (rods). The machine was reduced in terms of height and positioning system. The purpose of these variations, mainly concerning the consistency and elasticity of the rods, was to enhance canopy penetration, thus improving the contact with flowers or fruits, to operate the detachment directly and homogeneously in the canopy and the direct fall to the ground without side thrusts. The different work elements were also equipped with greater mass and allowed a more homogeneous penetration capacity in the tree canopy detaching the fruits from the branch directly at the fruit abscission point without damage, that could compromise the development of remained fruits and plant productivity. The major flexibility allows adaptation to the obstacles represented by large section branches. The presence of solicited fruits, which then fall, only after 3 days, mainly concerned the branches perpendicular to the row and with a length of less than 30 cm or up to—60 cm-length (Figure 4). The effect of penetration into the canopy was satisfactory even if no precise reference values were detected, but only linked to the low presence of branches not affected by the work elements of the machine. The crop showed a good yield level, with 39.7 t ha⁻¹ and a percentage of marketable fruits exceeding 80%. At the end of the cycle mechanical thinning completed with manual finishing did not influence quality traits significantly more than other management techniques (i.e., plant protection, irrigation, fertilization).

For these evaluations, a gap analysis evaluation [38] of thinner machine improvements, with targeted insights, could be interesting to carry out. It could be probably applied between manual and mechanical thinning, given the strong technical and economic gaps occurring. Various publications cited [4,9,11,13] refer to whole plants thinning; in our case, the comparison at crop level focuses on specific branch characteristics (length and orientation) on the technological level, instead it only affects part of the structure of machine setting and different work elements with rather marked criticalities for this analysis aimed at quantifying factors and results, which are not always shown.

Intensive cultivation systems imply that plants are closer and normally trained at a production height between 3.5 and 3.9 m, so that the reduction of the working front to 1.6 m of the machine used in this test is favorably reflected by the reduction of the overlapping central areas caused by parallel passages. In the case of the previous version [27] of 2.5 m, a forced overlap of over 1.5 m per passage was determined, bearing in mind that the frequent presence of anti-hail nets does not allow the lifting of the machine over 3.2–3.5 m.

This setting of the machine showed interesting improvements in terms of thinning efficiency and of damage to both fruits and branches as a function of class length and insertion point in the main stem. One critical aspect dealt with fruits in short branches, where the machine shows a more reduced thinning effect. In regard to the thinning efficiency, it was always above 37% of the load left after the intervention, and on average close to 44%, as observed by [27], in branches <35 cm-length. Damages were below quantifiable thresholds as actual damage to marketable production or to the structure of the plant; the study was in line with others [3,27,33]. The management of the foliage in terms of regularity, homogeneity, and thickness is an important aspect to be evaluated and improved for future experiences. In fact, all the plant training solutions were able to bring towards a homogeneous distribution of the fruits in the hedge, and in the external part of the canopy, to avoid the unfavorable effects of shading, allowing at the same time positive feedback on the adoption of mechanical thinning interventions. In the view to the progressive integration of these agronomic, physiologic, and mechanical aspects, it will be necessary to provide ad hoc management options not only limited to the thinning operations to which CREA is already operating.

Another aspect that deserves further attention concerns the regularity of positioning/distance of the thinner machine from the row to uniform the level of rod penetration in the canopy. Being closer to the row, the rod lengths of 1.34 allowed to cross the canopy, sometimes generating horizontal oscillations of rods’ terminal tips that can affect fruits, branches, and leaves. A more precise application is needed, by limiting the penetration to canopy midpoint, leaving the other part to the passage on the opposite side. For this purpose, other thinning machines are in the design stage, with shorter rods to improve the homogeneity of canopy penetration, and to reduce the overlapping outside the specific thinning sector.

The thinner machine with the new rods showed a thinning efficiency on apricot of a few percentage points higher than that found in 2016, from 43.6 to 45% [27].

5. Conclusions

The results obtained in this experiment were particularly interesting. The mechanical thinning led to the reduction of 45% of the total fruit load without commercial fruits being damaged, allowing only 32% for manual finishing. In the area treated only manually, the workload was of 52 h ha⁻¹ while the manual finishing operation after the mechanical thinning needed 18 h ha⁻¹ with a strong reduction, with respect to only the manual thinning.

A critical aspect of the machine was dealing with fruits on short branches, where the machine shows a more reduced thinning effect. The thinning efficiency was always above 37% with peaks of 44%, in branches <35 cm-length.

The management of the canopy in terms of regularity, homogeneity, and thickness is an important future aspect for improving the efficiency and setting of the thinning machine.

6. Future Work

At this level of development of the machine, where good efficiency and work capacity have been achieved, targeted integrations between cultivation technique and mechanized construction site may be desirable and necessary to reduce the criticalities found related to the distribution of fruits along the branches and to the length of the branches themselves. Tests of specific management of the foliage will be carried out, for example, by pruning in order to check the development of the branches and the distribution of the fruits on them. Similar experiences have already been successfully undertaken on peach trees.