Plants and establishment of the experiment

For accurate field results, this experiment was carried out under field conditions on the El Cortijo feijoa production farm, located at 4°55'9.59" N and 74°17'26.69" W in the municipality of San Francisco (Cundinamarca, Colombia) at an altitude of 2,283 m a.s.l., with an average temperature of 13.2°C, RH between 79 and 84%, and annual precipitation of 1,400 mm. The plants were in an area protected by a 6-gauge plastic sheet, covering only the roof and protecting the plants from direct rainfall. Inside the structure, the temperature reached a maximum of 22°C at noon.

The 'Quimba' variety (clone 41) was used. Seventy-two 9-month-old plants (including the border plants) from the farm's nursery were used, propagated with cuttings, with an average height of 1.25 m. The plants were transplanted into 10 kg plastic bags filled with inert substrate, pale gray quartz sand (grain size of 1.2 mm and CE of 0,011 dS m^-1). To promote root emergence after transplant, a 2-month (June 30-August 31) adaptation period was used. The treatments started on August 31 and lasted until December 11. The plants were arranged in three rows, with 0.25 and 0.4 m between rows and plants, respectively.

Quantitative variables were evaluated every 15 d starting at day 39 of the treatments for apical growth, production of apical and basal shoots and leaf chlorophyll content, along with symptoms determined for each nutritional deficiency, including chlorosis, marginal necrosis, meristem wilt, and retarded growth. Chlorophyll content index (CCI) units were determined with a portable chlorophyll meter (CCM-200; Opti Sciences, Tynsgboro, MA). The procedure was done in triplicate for each replica in leaves from low, medium and high levels. The growth rate (cm d^-1) was calculated based on ^{Hunt (1990}), determining the change in the length of lateral shoots per time unit. A description of the symptoms manifested in the plants was carried out with a photographic record. This activity was done every 8 d starting at 20 d after starting the experiment.

Following ^{Martínez et al. (2009)}, a modified Hoagland and Arnon solution was applied for each treatment (Tab. 1), suppressing one element in each treatment (Tab. 1). Element doses were used based on guava extraction levels and foliar analyses adapted from feijoa (^{Fischer et al., 2003}; ^{Natale et al., 2009}; ^{Gómez, 2012}). The prepared mixture received 1 L concentrations, applying them twice a week with irrigation water.

The irrigation water was validated with a physicochemical analysis to prove its suitability for the experiment. The compounds were: calcium nitrate (Ca(NO₃)₂), ammonium phosphate ((NH₄)₃PO₄), urea (CH₄N₂O), dipotassium phosphate (K₂HPO₄), potassium nitrate (KNO₃), potassium sulfate (K₂SO₄), calcium chloride (CaCl₂), magnesium sulphate (MgSO₄), sodium borate (Na₂B₄O₇·10H₂O), and chelated compounds of EDTA for minor nutrients (Fe, Zn, Cu, Mn, Mo, Co).

Data analysis

A completely randomized design (CRD) was used, defining one plant as an experiment unit. Five treatments with nine repetitions were set up. Six samplings (five non-destructive and one destructive) were done at the end of the experiment. The data were analyzed using ANOVA and Tukey tests with a significance of 5% and SAS v. 9.2 for quantitative variables.

Symptoms

Once 60 d of the treatments had passed, symptoms were detected, which were consistent for most nutrients. Only Ca had an earlier response, after 30 d of treatment, but it’s singular leaf symptom only appeared after 75 d. Mg and B deficiencies were associated with malformations in shoot and leaves. Additionally, B deprivation resulted in structural loss that was evident in the leaves, which became weary and faint. Figure 1 shows one representative plant for each treatment. Interestingly, most Mg deficient plants that were treated with water (negative control) released most of their leaves and become nearly naked. In the -Ca and -B plants, this also occurred, but only in the upper part, indicating a mobility effect with these two nutrients (^{Marschner, 2012}). Citrus trees may defoliate completely under a Mg deficiency and high fruit production (^{Zekri and Obreza, 2013}). New biomass production has a high demand for Mg, and a serious Mg deficiency culminates in early leaf shedding (^{Pallardy, 2010}). In this experiment, early leaf shedding occurs specifically in Mg repetitions with no new shoots. Defoliation was also noted in the negative control. ^{Dussán et al. (2016)} reported that deficiencies had an effect on distribution matter in guava (Psidium guajava L.), where the Mg deficiency limited leaf production and resulted in the poorest treatment for foliage generation. Thus, similarities with other crops, not only with myrtaceous, were observed for feijoa.

Figure 1. A. Control treatment plant. B. Negative control plant (only irrigated with water). C. Plant deficient in Ca. D. Plant deficient in Mg. E. Plant deficient in B. 

Calcium

Symptoms started 30 d after applying treatments. As an early symptom, young leaves shriveled and struggled to expand. Early necrosis occurred, advancing from the leaf tips to the leaf base, encompassing the distal half (Fig. 2). This also occurred in young leaves at the plant base. The second symptom occurred after 75 d, with a visible symptom throughout all plants in the treatment group, reddish and black veins in young, mature leaves, advancing from the apical portion towards the central portion (Fig. 3).

Figure 2. Evident symptoms associated with a calcium deficiency in feijoa. A. Distal necrosis in young leaves. B. Wilted tips in basal shoots. C. Wilted apical leaves. 

Figure 3. Evident symptoms associated with a calcium deficiency in Feijoa. A. Intravenous reddish in young, fully-expanded leaves. B. Intravenous reddish in young, fully-expanded leaves. C. Reddish veins in foliar adaxial view. D. Reddish veins in foliar abaxial view. 

Tip necrosis occurs mainly because Ca is not sufficiently mobile in plants; hence, when limited, cell wall production in distal zones of young organs is impaired by a lack of cell-to-cell adhesion, and necrosis occurs (^{Marry et al., 2006}), confirming its crucial role in strengthening cell walls and plant tissue (^{Marschner, 2012}; ^{Prabhakar et al., 2016}). Additionally, Ca is only transported via xylem, and organs with a low transpiration rate, such as young leaves, are especially vulnerable to Ca deficiencies, leading to necrosis (^{Dwivedi and Dwivedi, 2012}). Equally, in cape gooseberry, new leaves of Ca deficient plants crinkled and shriveled (^{Silva et al., 2017}). Interestingly, visibly symptoms were broadly expressed in young and fully-expanded leaves; thus, its expression is an earlier indicator of Ca deficiencies. In crops, this symptom is easily noted and can be used as an indicator to correct Ca fertilization (Fig. 3) or add an extra foliar application. Unfortunately, these symptoms have not been reported before in similar crops, such as guava, or in eucalyptus, another member of the Myrtaceae family, but the apical bud died in both eucalyptus and feijoa (^{Silva-Cardoso et al., 2014}).

Magnesium

Magnesium symptoms were mainly black spots in the leaves (Fig. 4 and 5A) and malformations at growth points. In the early treatment stage, the leaves exhibited interveinal chlorosis (<30 d), similar to avocado (^{Hurtado et al., 2019}), followed by black spots and malformations and necrosis spreading to different growth points. Interestingly, several leaves were not able to expand properly and developed an invagination in mature leaves. This might have occurred because Mg is essential to ribosomes integrity, so cell division and protein biosynthesis disruption result in an inability to complete organogenesis (^{Pallardy, 2010}). This was also observed in eucalyptus (^{Silva-Cardoso et al., 2014}). Figure 5A shows a common response to meristematic death, which in turn produces an overproduction of shoots downward to the apex. These symptoms have been reported with a B deficiency by ^{Shireen et al. (2018)}.

Figure 4. Evident symptoms associated with a magnesium deficiency in Feijoa. A. Black spots in an old leaf. B. Black spots advancing acropetally. C. Malformations caused in young leaves. D. Reddish veins in foliar adaxial and tip malformations. 

Another symptom was observed in mature leaves: swelled central tissue with a chlorotic leaf edge (Fig. 5B). This particular symptom might be associated with an incapacity to transport substance between tissue. Mg is critical to loading sucrose into phloem for long distant transport (^{Cakmak and Yazici, 2010}). Given its ligand property with other compounds, Mg facilitates several communication functions (^{Marschner, 2012}). For instance, the Mg/ATP and H⁺-ATPase substrate drive sucrose export from foliar cells to the phloem (^{Tränkner et al., 2018}). A Mg deficiency sees sucrose and starch accumulation in the leaves and reduced CO₂ fixation (^{de Bang et al., 2021}). Hence, an accumulation of certain substance might indicate a symptom.

Figure 5. Evident symptoms associated with a magnesium deficiency in Feijoa. A. Slight curled malformation associated. B. Leaves with interveinal chlorosis and swelled tissue in the inward portion. C. Intervein chlorosis in fully-expanded leaf. 

Several leaves had black spots in expanded leaves and necrosis in folded leaves (Fig. 6). Black spots in expanded leaves appeared to be a l symptom of particular to the Mg deficiency in feijoa, different from many other fruit species, and could also be related to the light intensity in the growth chamber of the experiment (^{Cakmak and Yazici, 2010}). Necrosis has also been reported in avocado but with a different pattern (^{Barnard et al., 1991}) that does not reach unfolded leaves but has small necrotic spots on the tip of fully-expanded leaves. During the last part of the experiment, several Mg deficient leaves started to fall, leaving the plants naked. ^{De Bang et al. (2021)} also reported results where a Mg deficiency induced root stress via sugar deprivation, producing ethylene and leading to leaf detachment. Ethylene production has been reported as a consequence of Mg deficiencies, inducing root hair growth more than defoliation (^{Liu et al., 2017}). In this experiment, only the -Mg plants and negative control presented strong leaf loss; thus, ethylene overproduction might not be a suitable explanation because of its gaseous form; no other treatment plants experienced this (^{Adams and Yang, 1981}). Recently, Mg was associated with leaf senescence and loss, oxidative damage causing changes in redox states, metabolic alterations and photosynthesis disruption that underpins defoliation (^{Tanoi and Kobayashi, 2015}).

Figure 6. Evident symptoms associated with a magnesium deficiency in feijoa. A. Wilted tips in young leaves. B. Black spots in fully-expanded leaves and reddish veins associated with a Ca deficiency in upper leaf. C. Black spots in matured leaf. 

Boron

A boron deficiency predominantly damages actively growing organs, such as shoot and root tips, so the whole plant may be stunted (rosetting) as ^{Han et al. (2008)} observed in boron-starved citrus cuttings.

The first symptoms associated with a B deficiency are related to chlorosis and disruption of leaf expansion. Leaf turgid was ameliorated, and chlorosis occurred in several young leaves. In mature, young leaves, chlorosis occurred in the second-half, veins and margins (Fig. 7). Interveinal chlorosis was noted, with some necrosis at the distal part in young leaves. Interestingly, leaves became wilted and thinner than in other treatments, which might confirm the role of B as a structural element in the membrane and in lignification (^{Brown et al., 2002}; ^{Cakmak et al., 2023}). During the study, there was no effect on growth in the control plants; perhaps the experiment was not long enough to induce a stunting effect.

Figure 7. Evident symptoms associated with a boron deficiency in Feijoa. A. Chlorosis in, overall and margins, young leaves. B. Intervein chlorosis in fully-expanded, young leaves. C. Chlorosis and degraded green in fully-expanded, young leaves. D. Uncommon pattern associated with a B deficiency. 

In guava, B deficiencies are expressed differently, with stunted growth and buds that fail to open, while others open but then shrivel and die (^{Kumari et al., 2018}). In other crops, such as blackberries, major disorders were mentioned, including multiple architecture shoot disorders with a higher production of rosette-shaped shoots in the entire plant (^{Castaño et al., 2008}). Although this abnormal growth has been noted in many production sites, during the experiment, this result did not occur with the B deficiency. Instead, it was seen with the Mg deficiency, which might be explained by small quantities of B that remained in cuttings from earlier growth, which overcame the deficiency limitations throughout the experiment (^{Fischer et al., 2003}; ^{Fischer et al., 2020}). Young leaves require boron because of its role in cell division and elongation (^{Shireen et al., 2018}). In feijoa, new leaves also showed strong chlorosis, as noted in Figure 7A, and showed misshaped leaf blades (^{Cakmak et al., 2023}).

Additionally, an unusual pattern was noted in B deficient plants (Fig. 8), with a mosaic of integrating patterns. Marginal and central vein chlorosis were observed after black patches in the leaf base (Fig. 8B). This was due to B’s functional role in membranes. Studies have indicated alterations of H⁺, K⁺, PO₄ ³⁻, Rb⁺, and Ca²⁺ ions with a B-deficiency in Vicia faba (^{Robertson and Loughman, 1973}), which produced a camouflaged deficiency of those elements. Also, B is involved in phenolic metabolism (^{Cakmak et al., 2023}), and, as mentioned, the feijoa plant is a high phenol producing plant, not only in the fruits but also in the leaves (^{Zhu, 2018}). Thus, black patches are probably the result of oxidation caused by elevated polyphenol concentration. This substance is metabolized by the B-sugar cis-diol complex, which regulates phenol accumulation and maintains its proper metabolism. Hence, because of the absence of B, this metabolism is disrupted, and oxidation damage appears (^{Wimmer et al., 2019}). ^{Marschner (2012)} reported that a B deficiency decreases usage of absorbed light in photosynthesis, generating phenolics oxidation and hindering the defense mechanism of antioxidants in plants, causing a higher susceptibility of plants to excessive light intensity and ROS formation.

Figure 8. Evident symptoms associated with a boron deficiency in Feijoa. A. Uncommon pattern associated with a deficiency, central and edge chlorosis with intervein necrosis. B. Uncommon pattern during first stage. 

Effect on some growth variables

Nutrition in plants is essential to obtaining the desired yield. Plant production requires all essential nutrients to develop all functions that provide a product with known quality (^{Rajendran et al., 2009}). Because each distinct nutrient plays a strategic function in the plant, there is a physiological expression caused by its particular depletion (^{van Maarschalkerweerd and Husted, 2015}). To test growth, several quantitative variables were measured to provide information on plant performance with a deficiency.

The apical, longitudinal growth did not show differences between the distinct treatments; however, decreased growth was recorded for all treatments. Surprisingly, B had the smallest effect, possibly because of the short experiment time (^{Fischer et al., 2020}), followed by magnesium and calcium. Apical growth requires healthy photosynthesis to meet the demand for energy (^{Pallardy, 2010}). With 4.17 cm, Ca showed less apical elongation. ^{de Bang et al. (2021)} reported that a deficiency produces meristem death and necrotic lesions on leaf margins and tips, which was also reported in this experiment. This result is supported by the Ca structural function in budding and rooting (^{Marschner, 2012}).

For apical shoot production, only Mg presented a significantly reduced result, which is underpinned by its role in root growth, transport and carbon allocation, which are linked to photosynthesis (^{Cakmak and Yazici, 2010}; ^{Marschner, 2012}). These functions were disrupted, halting shoot meristems production. Interestingly, neither Ca nor B had a significant reduction in shoot production. Both calcium and magnesium are widely known as cooperative nutrients that have a complementary role (^{Rajendran et al., 2009}). Differences in cell wall chemistry based on Mg and Ca concentrations allow plants to execute distinct stoichiometric equations. Thus, feijoa plants are more sensitive to Mg absence because of its role in carbon allocation (^{de Bang et al., 2021}). The Mg deficiency was observed early despite its property of fast ligand exchange kinetics, which result in easy substitution with several positive ions, including Mn²⁺, Ca²⁺, Fe²⁺, Co²⁺, Cu²⁺ and Zn²⁺ (^{Allnér et al., 2012}). These authors indicated that hampered growth may still occur, mainly because Mg substitution slows down enzyme activity (^{Schmidt and Husted, 2019}).

Basal budding occurred in feijoa; however, in each treatment, different tendencies occurred because of the lack of significance in the results (Tab. 2). Plants without Ca in the nutrient solution tended to produce 19 times more basal shoots than the negative control. In previous experiments, a lack of nitrogen resulted in an almost total remove of basal shoots, while a K deficiency increased the number. This was associated with an osmotic disruption caused by K deprivation that promoted shoots at the plant base (^{Buitrago et al., 2021}).

Chlorophyll content is a common measure used to gain information on the nutrition status of a plant (^{Roveda-Hoyos et al., 2022}). Only the negative control accumulated a significantly lower content in this variable (Tab. 2). Mg is a main component of tetrapyrrole ring in chlorophyll (^{Taiz et al., 2017}); however, its content was not lessened with the Mg deficiency in this experiment. This might have occurred because only 5-25% of the total absorbed Mg was established in the chlorophyll ring (^{de Bang et al., 2021}). The second role of Mg is related to the cation exchange capacity of the cell wall, modulating Ca and Mg and additional signals (^{White et al., 2018}).

The depletion of Ca and Mg strongly and significantly impaired growth rate, which decreased 27 and 22% from the control treatment, respectively (Tab. 2). This reaction occurred because Ca not only belongs to the cell wall as a structural element but also influences cell wall permeability by creating bridges with phosphate and carboxylate groups in lipids and proteins (^{de Bang et al., 2021}). Calcium is important as a signaling substance, sensors include calmodulins (CaMs) and CaM-like proteins (CMLs), which promote protein activation with Ca binding, which in turn induces gene regulation (^{La Verde et al., 2018}). Hormones are also related to Ca signaling. For instance, the ABA action is Ca dependent, via CPKs/CDPKs and calcineurin B-like (CBL) receptors that allow ABA to act on guard cells. Mutants lacking CPK do not react to ABA stimulation or, therefore, to droughts (^{Geiger et al., 2010}). For magnesium, the Mg-starved plants showed not only a significantly lower apical shoot production but also a lower growth rate than the control plants. As previously stated, a Mg deficiency can decrease growth because of the deficient activation of enzymes involved in photosynthesis, respiration, and DNA and RNA synthesis (^{Taiz et al., 2017}; ^{Schmidt and Husted, 2019}). In addition, its role as the central Mg atom of chlorophyll molecules that absorb photons and initiate electron flow (^{Marschner, 2012}) means that any deficiency of this nutrient affects photosynthetic performance. On the other hand, ^{Cakmak and Yazici (2010)} reported a high accumulation of sucrose in Mg-deficient leaves, indicating severe inhibition of sucrose transport out of Mg-deficient leaves, impeding photosynthesis.