Zinc Fertilizers and Additives for Foliar Fertilization of Cocoa Seedlings

The research is financed in part by the Coordination of Improvement of Higher Education Personnel-Brazil Code Abstract Foliar fertilization is an interesting strategy for nutrition with micronutrients in perennial plants; among the micronutrients, zinc (Zn) deficiency is the most frequent in cocoa trees ( Theobroma cacao L.). The present study aimed to evaluate the efficiency of Zn sources through foliar application for the cocoa crop. The experiment was carried out in a randomized block design, with 10 treatments and four replicates. Treatments were: foliar fertilizations containing 1 g L -1 of Zn using two inorganic sources (chloride and sulfate), in the presence or absence of additives (urea and sucrose); two organic sources (Zn-EDTA, and from chloride and sulfate); soil fertilization with 8 mg dm -3 of Zn, and a control (without addition of Zn). Foliar fertilizations with Zn were monthly applied for five months, and the experiment was conducted for 210 days. The results were subjected to analyses of variance and contrast. Zn fertilization, regardless of the form of application, increased Zn contents and accumulations in the leaves. Zn fertilization in the soil, at planting, led to a recovery rate by the plant similar to the mean value caused by foliar fertilizations. ZnCl 2 caused higher Zn contents and accumulations in the leaves and was more efficient than sulfate and EDTA; addition of urea to the ZnSO 4 solution increased Zn accumulation in the leaves, whereas addition of sucrose to the ZnCl 2 solution reduced Zn content in the leaves. and and Zn complexed by EDTA (1 g L of g L of EDTA); two control treatments, one with Zn fertilization in the soil at dose of 8 mg dm of per in the form of ZnSO 4 O, applied before the seedlings and in the entire soil and an absolute control (without Zn fertilization). prepared and agitation for 24 h before utilization

It is important to highlight that soil fertility, nutrition and fertilization are essential components to construct an efficient production system. Therefore, deficiency of one or more nutrients, including those required in lower amounts (micronutrients), will affect several physiological processes of the plant.
Zinc (Zn) is the micronutrient most frequently deficient in cocoa trees in Brazil (Chepote et al., 2013). In perennial crops, foliar fertilization with micronutrients has been a strategy used to supply them, especially when there are restrictions to the absorption by roots, such as: soils with low water availability, low root activity during the reproductive phase and for situations in which soil fertilization has lower efficiency (Wójcik, 2004;Faquin, 2005;Eichert & Fernández, 2012;Fernández, Sotiropoulos, & Brown, 2013).
For cocoa, the most used Zn fertilizer for foliar fertilization is ZnSO 4 (Malavolta, 1987;Chepote et al., 2013), but studies with other crops have demonstrated (Boaretto, Muraoka & Buoratto, 2003) or not (Teixeira, Borém, Silva, & Kikuti, 2008) higher efficiency of Zn absorption when the inorganic source is ZnCl 2 ; other studies also indicate that the efficiency of Zn absorption by the leaves increases when KCl is added to the ZnSO 4 solution (Coutinho Neto et al., 2010).
Chelates, organometallic complexes, can be less adsorbed by the cuticle and cell wall than inorganic sources of Zn (Franco, Martinez, Zabini, & Fontes, 2005) and can be more mobile within the plant, but that does not guarantee that the chelated element has higher absorption in comparison to inorganic salts (Wójcik, 2004). The effect of chelates on leaf capacity to absorb nutrients can be related to some of their properties such as: complex's molecular weight, dissociation constant and complex's stability at solution pH (Reed, Lyons, Júnior, & Mceachrn, 1988).
Thus, this study aimed to evaluate the efficiency of Zn sources and additives for foliar application in cocoa seedlings.

Material and Methods
The experiment was conducted in a greenhouse at the State University of Santa Cruz (UESC), in the municipality of Ilhéus, Bahia state, Brazil.
Cocoa (Theobroma cacao L.) seedlings were produced by cuttings at the Cocoa Biofactory Institute (Instituto Biofábrica de Cacau-IBC), cultivated in commercial substrate and coconut fiber at 1:1 volumetric proportion, in 288-dm 3 tubes, for a period of approximately 120 days. The experiment used the clone CCN51, which is widely known as highly productive and resistant to witches' broom.
The soil was previously sieved through a 5-mm mesh and corrected with calcium and magnesium carbonates (Analytical Reagents-AR), Ca: Mg ratio of 4:1, to increase base saturation to 80%. Fertilization at planting was performed with the nutrients P, N, K, S, Mn, Cu, B and Mo, at doses of 400, 169, 60, 26, 6, 2, 0.8 and 0.15 mg dm -3 , respectively, in the form of purified monoammonium phosphate (MAP) and the AR salts potassium sulfate, manganous chloride, copper sulfate, boric acid and ammonium molybdate. All fertilizers, except MAP, were applied through a nutrient solution in the entire soil volume.
The experimental design was randomized blocks, with two blocks and each one with two replicates, totaling 40 experimental plots. The experiment had 10 treatments, formed by a (2 × 3) + 2 + 2 factorial scheme, which corresponded to two inorganic sources of Zn, zinc chloride (ZnCl 2 ) and zinc sulfate (ZnSO 4 ·7H 2 O), applied through the leaves at dose of 1 g L -1 of Zn, with and without the presence of two additives (urea at 10 g L -1 and sucrose at 3 g L -1 ); plus four additional treatments: two treatments also with foliar application of Zn, but using organic sources of Zn, from chloride and sulfate, and Zn complexed by EDTA (1 g L -1 of Zn + 6 g L -1 of EDTA); two control treatments, one with Zn fertilization in the soil at dose of 8 mg dm -3 of Zn (88 mg of Zn per pot), in the form of ZnSO 4 ·7H 2 O, applied before transplanting the seedlings and in the entire soil volume; and an absolute control (without Zn fertilization). Zn + EDTA solutions were prepared and subjected to agitation for 24 h before utilization in the sprays.
Seedlings were irrigated with rainwater, collected from the greenhouse roof and stored in polyethylene tanks. Every two weeks, 30 days after transplantation, all treatments received soil fertilization with 20 mg dm -3 of N, 10 mg dm -3 of K and 10 mg dm -3 of P, with urea, potassium nitrate and purified MAP. Ninety days after transplantation, these doses were doubled.
Monthly, 30 days after transplantation, treatments with foliar fertilization were applied and the seedlings were sprayed with the respective solutions, trying to cover all leaf surface. During the application, the soil was covered to avoid the contact with the solution, and plant sides were protected with a cardboard barrier to avoid the contact of the solution with the adjacent plots.
Six foliar applications were performed; the first two and last four used 50 and 100 mL of solution for the four plots of each treatment, respectively, due to the increase in leaf area of the plants. These doses were equivalent to a total dose of 125 mg of Zn per plant along the experiment.
At 210 days of cultivation, plants were harvested and separated into leaves, stem and roots. Leaves were counted and leaf area (LA) was measured with a benchtop LI-3100C meter. Then, the leaves were washed in distilled water, followed by 30 mL L -1 HCl solution and, lastly, two additional washings with distilled water. This procedure was adopted to remove the Zn deposited on leaf surface and adsorbed to the cuticle.
After harvest, all plant parts were dried in oven at 65 ºC for 72 hours and weighed to obtain leaf dry matter (LDM), stem dry matter (StDM), root dry matter (RDM), shoot dry matter (ShDM = LDM + StDM) and total dry matter (TDM = ShDM + RDM). The following parameters were calculated: shoot/root dry matter ratio (S/R = ShDM/RDM), mean leaf area per leaf (MLA = LA/number of leaves) and specific leaf area (SLA = LA/LDM) The soil of each plot was homogenized, sampled and analyzed, and the Zn contents were extracted by Mehlich-1. Zn contents were also determined in the leaves, by atomic absorption spectrophotometry (EMBRAPA, 2009). Zn accumulation in the leaves was obtained by multiplying its leaf content by LDM.
The recovery rate (RR) of the applied fertilizer was calculated by the following equation: RR = (ZnA_L -ZnA_F C ) × 100/ToDZn; where, ZnA_L: Zn accumulation in the leaves, ZnA_L C : mean Zn accumulation in the leaves in the absolute control, ToDZn: total dose of Zn, equal to 88 mg of Zn per plant, for the treatment with soil fertilization, and to 125 mg of Zn per plant, for the eight treatments which received foliar fertilizations.
Biometric variables, Zn contents in soil and leaves, and Zn accumulation in the leaves were subjected to analysis of variance, with follow-up test of treatment degrees of freedom by contrasts at 0.05 probability level by F test.

Results and Discussion
Mean values of biometric variables, Zn contents in soil (ZnC_S) and leaves (ZnC_L), Zn accumulation in the leaves (ZnA_L) and recovery rate of the applied fertilizer (RR), as well as their coefficients of variation (CV), are presented in Table 1. The contrasts analyzed are presented only for variables which showed significant differences (Table 2). Table 1. Biometric variables¹ of cocoa seedlings (clone CCN 51) and zinc in soil and leaves² at 210 days of cultivation, as a function of treatments (T)³  Note. 1: Biometric variables: height (H); diameter (D); leaf area per plant (LA); mean area per leaf (MLA); specific leaf area (SLA); masses leaf dry matter (LDM), stem dry matter (StDM), root dry matter (RDM), shoot dry matter (ShDM) and total dry matter (TDM) and shoot/root dry matter ratio (S/R).
4: The recovery rate (RR) of the applied fertilizer.

5: Coefficients of variation (CV), %.
The biometric variables analyzed had lower experimental variability than Zn contents in soil and leaves, since the CV of the former oscillated between 10% and 25%, whereas the CV of the latter varied from 32 to 53%, and ZnA_L showed highest variability (Table 1).
There were significant effects for Zn content in the soil (ZnC_S) and especially for its content (ZnC_L) and accumulation (ZnA_L) in the leaves. In comparison to the control, Zn fertilizations significantly increased its contents in the soil and in the plant and, consequently, ZnA_L (Table 2, contrast 1), which had mean increment of 467% compared with the control (Table 1).
Both forms of fertilization resulted in similar mean contents and accumulations in the leaves (Table 2, contrast 2), indicating that the dose applied in the soil, 88 mg of Zn per plant, led to Zn contents and accumulations in the leaves similar to those caused by the dose applied through the leaves, 125 mg of Zn per plant, a fact confirmed by the fertilizer recovery rate (RR), which was equal to 12.4% for soil application and mean of 12.6% for the eight treatments which received foliar fertilization (Table 1). However, Zn fertilization at planting in the entire soil volume, compared with the mean of the eight treatments with foliar fertilization, significantly increased its content in the soil (  Without vs with urea for chloride T3 vs T4 ns ns ns ns 7 Without vs with urea for sulfate T6 vs T7 ns ns ns 9,1 * 8 Without vs with sucrose for chloride T3 vs T5 -34.7 * ns -76.7 * ns 9 Without vs with sucrose for sulfate T6 vs T8 ns ns ns ns Note. 1: Only significant contrasts were calculated, which were calculated by the difference between the means of the second and the first group of treatments (** and * = significant at 1 and 5% by the F test, respectively; ns = not significant).
2: Mean area per leaf (MLA); zinc contents not only (ZnC_S) and in the leaves (Znc_l) and total content of zinc in leaves (ZnA_L).
The sources of foliar Zn did not affect its content in the soil but significantly influenced its contents and accumulations in the leaves (Table 2). Fertilizations with ZnCl 2 , in comparison to ZnSO 4 , significantly increased the mean ZnC_L by 72.5 mg kg -1 and mean ZnA_L by 7.0 mg per plant (Table 2, contrast 3). Subtracted from the contribution of root absorption by the control (treatment not fertilized with Zn), these values result in mean increments of 58% and 49% of ZnC_L and ZnA_L, respectively, i.e., the mean RR of ZnCl 2 was approximately 50% higher than that of ZnSO 4 , for foliar fertilization in cocoa trees (Table 1). These results are consistent with those observed by Boaretto, Muraoka & Buoratto (2003) for the coffee crop. However, Teixeira et al. (2008) found no significant difference in ZnC_L of common beans due to fertilizations with Zn chloride or sulfate.
ZnCl 2 was also a source superior to Zn-EDTA because it led to significant increments in mean ZnC_L and ZnA_L of 143.3 mg kg -1 and 14.2 mg per plant, respectively (Table 2, contrast 4). Subtracted from the contribution of root absorption of the treatment not fertilized with Zn, these values represented mean increments of 150% and 135% in ZnC_L and ZnA_L, respectively, i.e., the chelation o Zn from ZnCl 2 reduced its RR by 57% (Table 1). Lower efficiency of the source Zn-EDTA, compared with ZnCl 2 , has also been observed in the coffee crop (Boaretto, Muraoka, & Buoratto, 2003).
Nonetheless, the sources ZnSO 4 and Zn-EDTA did not differ significantly with respect to ZnC_L and ZnA_L (Table 2, contrast 5), indicating a similar efficiency for foliar fertilization in cocoa trees and lower RR (Table 1). Franco et al. (2005), studying translocation and compartmentalization of Zn applied in coffee and common beans, under Zn sufficiency and deficiency, also observed similarities between the Zn contents effectively absorbed by the leaves when the sources were ZnSO 4 and Zn-EDTA, except for common bean plants under Zn sufficiency, in which foliar absorption of Zn was higher when the source was Zn-EDTA. Ojeda- Barrio et al. (2009) also found no significant difference in ZnC_L in pecan trees sprayed with ZnSO 4 and Zn chelates (EDTA and DTPA).
Although ZnSO 4 is a source of lower efficiency for foliar fertilization than ZnCl 2 , the addition of 10 g L -1 of urea significantly increased ZnA_L, by 9.1 mg per plant (Table 2, contrast 7), which represents an 86% increment in the total Zn accumulated in the leaves, indicating that urea is a good additive for foliar absorption of Zn by cocoa trees, when the source is ZnSO 4 . Addition of urea to the ZnCl 2 solution did not have the same beneficial effect observed when the source was ZnSO 4 , because it did not significantly affect ZnC_L and ZnA_L (Table 2, contrast 6), which may have occurred because the foliar absorption of Zn 2+ is already originally facilitated by Cl -( Table 2, contrasts 3 and 4).
Addition of 3 g L -1 of sucrose to the 1.0 g L -1 Zn solutions reduced ZnC_L when the source was ZnCl 2 , but did not significantly affect ZnC_L when the source was ZnSO 4 ( Table 2, contrasts 8 and 9, respectively).
Although fertilizations with Zn increase its absorption by plants, they did not result in increments of either growth or biomass since no significant effect was observed for almost all biometric variables (height, diameter, LA, SLA and dry matter of the different plant components), possibly due to the original Zn content in the soil, which guaranteed a minimum nutritional condition for plant growth, even for the absolute control, which had mean ZnC_L of 32.1 mg kg -1 . Compared with sufficiency range for diagnostic leaves of adult cocoa trees, this value is within the ranges proposed by Raij et al. (1997) and Salgado-García et al. (2006), but is below the ranges considered by Malavolta (1987) andSouza Júnior et al. (2012). Absence of response of growth or production variables to foliar fertilizations with Zn have also been observed for common beans (Teixeira et al., 2008) and silage corn (Coutinho Neto et al., 2010). However, increase in production (Zoz et al., 2012) and even in the nutritional value of the crops, due to foliar fertilizations with Zn, have been found when there is deficiency of the nutrient (Ghasemi et al., 2013).
The mean area per leaf (MLA) was the only biometric variable significantly affected by the treatments, being lower for foliar fertilization with Zn than for its application in the soil ( Table 2, contrast 2), possibly due to the saline effect of the fertilizer solutions, which would also explain the reduction in MLA caused by the addition of sucrose in the ZnCl 2 solution ( Table 2, contrast 8). Nonetheless, such restriction to leaf expansion was compensated by the increase in the number of leaves since the leaf area per plant (LA) was not significantly affected by the different solutions used in the foliar sprays.

Conclusions
(1) Zn fertilization in the soil or leaves increased Zn contents and accumulations in the leaves of cocoa seedlings.
(2) The recovery rate by the plant for Zn fertilization in the soil, at planting, was similar to that caused by foliar application.
(3) For foliar Zn application, chloride was more efficient than sulfate and EDTA, and these latter sources were similarly efficient.
(4) Addition of urea increased ZnSO 4 efficiency, but the addition of sucrose did not increase the efficiency of foliar fertilizations with either Zn chloride or sulfate.