Corn is the third most cultivated cereal in the world, whose production is currently only behind wheat and rice. The United States is responsible for producing 36%, followed by China with 21% and Brazil, which accounts for 7% of world production [1] . However, several factors contribute to low productivity of the corn crop, among which are aspects related to mineral nutrition, which involves the fertilization with micronutrients, such as boron and zinc, whose deficiencies can limit productivity [2] .

The most common nutrient deficiencies in corn are related to the micronutrients boron (B) and zinc (Zn). The low fertility of some soils, the exportation of crops, and the increasing use of lime and phosphate fertilizers are factors that contribute to the greater insolubility of zinc [3] . However, requirements of the plants are relatively low and the presence of small concentrations in soil potentially provided by plant residues can meet the plant nutrition, therefore there is a need for further studies to verify the effects of these residues on crop yields and the best method of application.

Of all the micronutrients, zinc is that which presents the best productivity responses in the corn crop in Brazilian soils, due to widespread deficiency that occurs, mainly in soils of the Cerrado [4] . Zinc is essential for different enzyme systems of the plant, controlling the production of important growth regulators. Its basic function is related to the metabolism of carbohydrates, proteins, phosphates, and also the formation of auxins, RNA (deoxyribonucleic acid) and ribosomes [5] [6] . Despite the benefits described in literature with the application of zinc to corn, [2] [7] did not encounter increased productivity after applying doses of this nutrient.

There are currently three known methods of zinc application to corn production systems, including application on: the ground (located or incorporated), leaves or seeds. It should be highlighted that incorporated zinc application to the soil may result in important residual effects in corn production [8] . Meanwhile, Galrão (1994) [9] noted that the mode of incorporated Zn application to the soil (broadcasting) had a superior effect on the first crop, where the second crop applications to the soil and also the leaves or seeds resulted in the same effects on crop production. Sakal et al. (1983) [10] found that foliar applications of zinc were similar to soil applications, either by broadcasting or application in the row at planting.

Given that the zinc doses required by crops are small [11] , there are difficulties in even distribution of fertilizer when applied to the soil [12] [13] , and therefore leaf application is an alternative. However, foliar applications have the disadvantage of low zinc mobility in the phloem [14] [15] . On the other hand, studies have indicated another method of Zn application, via the seed, considered viable for providing this micronutrient to the corn crop, especially to meet the initial phase of plant growth [16] .

Thus, the objective of the present study was to evaluate the residual effect of zinc application methods and rates on nutrition and productivity of corn.

The experiment was conducted in the experimental area of the Teaching, Research and Production Farm of the FCAV/UNESP (21˚15'22''S, longitude 48˚18'58''W, altitude of 575 m) in a dystrophic Dark-Red Latosol (dRL) [17] , with very clayey texture and gently undulating relief. Chemical analysis of soil (0 - 20 cm) was performed according to the methods described by Raij [18] and presented the following properties: pH CaCl₂ = 5.5, (O.M.) = 1.6 g∙dm⁻³, P (resin) = 31 mg∙kg⁻¹, K = 0.11 cmol_c∙kg⁻¹, Ca = 2.5 cmol_c∙kg⁻¹, Mg = 1.2 cmol_c∙kg⁻¹, H + Al = 2.2 cmol_c∙kg⁻1, sum of bases (SB) = 3.8 cmol_c∙kg⁻¹, V% = 63, S = 1.0 mg∙kg⁻¹, B = 0.30 mg∙kg⁻¹, Cu = 1.0 mg∙kg⁻¹, Fe = 13.0 mg∙kg⁻¹, Mn = 16.1 mg∙kg⁻¹ and Zn = 0.5 mg∙kg⁻¹. Each experimental unit consisted of four planted rows measuring 5.0 m long, spaced at 0.90 m, totaling 18 m². For assessment of the treatments the two central lines were considered, discarding the 1.0 m from the ends and the other remaining lines which were considered the borders. A simple corn hybrid was used (Dow AgroSciences^® 2B710). Planting was performed in December 2010. The average rainfall during the experimental period from planting to harvest was 1425 mm, and the average temperature for the period was 22˚C.

At the time of planting, basic fertilizer was applied uniformly to all treatments with the following composition: 30 kg∙ha⁻¹ of N, 50 kg∙ha⁻¹ of P₂O₅ and 50 kg∙ha⁻¹ of K₂O, in the form of urea and NPK fertilizer (02-20-20), according to the recommendations of Raij and Cantarella [19] .

Nine treatments were distributed as follows: three Zn doses (2, 4 and 8 kg of Zn ha⁻¹) for application in the rows at planting, three Zn doses (6, 12 and 24 kg of Zn ha⁻¹) incorporated with the soil at 0 - 20 cm depth, foliar application (0.4 kg of Zn ha⁻¹), application to the seeds (40 g of Zn kg⁻¹ of seeds) [20] and the control (no Zn application). The treatments applied to the soil and leaf received Zn in the form of zinc sulfate (22.7% Zn and 11% S). In the treatment applied to the seeds Zn oxide was used (79% Zn). Leaf applications of Zn were performed at 15 and 30 days after crop emergence [21] . For zinc applied via the seeds, a sugar solution was prepared and mixed with the seed to ensure better adhesion of the fertilizer. The experimental design consisted of randomized blocks with four replicates.

To establish the doses of Zn located, was adopted as the reference indication Raij and Cantarella (1997), as 4 kg∙ha⁻¹ Zn, and D1: half standard dose; D2: standard dose and the D3: twice the standard dose, then corresponding to 2, 4 and 8 kg∙ha⁻¹, respectively. For doses of Zn incorporated settled triple the dose for localized application, with the default amount of 12 kg∙ha⁻¹ Zn. Thus, treatments were incorporated into zinc consisting of: D1: half standard dose; D2: standard dose and the D3: twice the standard dose then corresponds to 6, 12 and 24 kg∙ha⁻¹, respectively.

The amounts of Zn incorporated in the soil were applied earlier, in December 2009, by means of the assay developed by Puga [22] . In the present study the residual affect was evaluated. Thus, the results of the Zn analysis in soil at the depth of 0 - 20 cm in the localized treatment with 2 kg∙ha⁻¹ of Zn was 0.88 mg∙dm⁻³; in the localized treatment with 4 kg∙ha⁻¹ of Zn was 1.70 mg∙dm⁻³; in the localized treatment with 8 kg∙ha⁻¹ of Zn was 1.88 mg∙dm⁻³; in the incorporated treatment with 6 kg∙ha⁻¹ of Zn was 1.0 mg∙dm⁻³; for the incorporated treatment with 12 kg∙ha⁻¹ of Zn was 2.01 mg∙dm⁻³; in the incorporated treatment with 24 kg∙ha⁻¹ of Zn was 1.68 mg∙dm⁻³; in the leaf treatment with 0.4 kg∙ha⁻¹ Zn the result was 0.55 mg∙dm⁻³; in the seed treatment with 40 g∙kg⁻¹ of Zn the result was 0.78 mg∙dm⁻³; and in the control treatment was 0.47 mg∙dm⁻³.

Weed management was performed by mechanized spraying of the post-emergence herbicides Atrazine (4 L∙ha⁻¹) and Glyphosate (3 L∙ha⁻¹). Broadcast fertilizing was performed by applying 90 kg∙ha⁻¹ of N and 60 kg∙ha⁻¹ of K₂O in the forms of ammonium sulfate and potassium chloride, respectively, uniformly in all treatments.

Two evaluations of growth were conducted, the first at 30 days after crop emergence and the second during the flowering period of the plants, considering that the crop cycle is about 120 days. Factors analyzed were the number of fully expanded leaves, plant height by measuring from the soil surface to the last fully expanded leaf and basal diameter (at the first internode).

To evaluate the nutritional status, leaf sampling was performed according to the indications of Cantarella [23] , in which the central third of the leaf at the base of the corn cob was collected. Determinations of Zn in the leaf were performed according to the methodology proposed by Battaglia [24] .

Corn harvest was performed from plants of the useful area of each plot. The corn ears were separated mechanically into cob and grains. Grains were stored in paper bags identified and weighed to calculate productivity, with moisture adjusted to 130 g∙kg⁻¹. The Zn content in grains was then determined according to the methodology of Bataglia [24] .

Soil sampling was performed in each experimental plot at the depth of 0 - 20 cm, by collecting ten points in the row (fertilization area) of the useful area of the corn crop. In treatments where Zn was applied in a localized form, points near the plant rows were sampled. In the other treatments sampling was done conventionally, in the area of fertilization.

Data was submitted to analysis of variance (ANOVA) according to the experimental design described. Means were compared by the Tukey test (P < 0.05), and for the doses a polynomial regression analysis was also applied, using the statistical program Sisvar [25] .

Foliar spraying with Zn simultaneously promoted increased leaf Zn content and basal diameter of the corn plant.

The Zn content in grains and soil and the grain yield were not affected by Zn application methods or by the applied doses.

The residual effect of the incorporated Zn doses applied was observed only in the basal diameter and the number of leaves at 60 days after emergence.