Materials and Methods

This study was developed in the irrigation and drainage area of the Federal University of Viçosa, in Viçosa, Minas Gerais State, Brazil (23 K, 722569.09 m E; 7701897.59 m S, UTM). The climate is classified as Cwa, with an average annual temperature of 20.6 °C. The culture was installed in the field using a conventional planting system, that is, plowing, harrowing, and operation with a rototiller for seedbed preparation. Planting fertilization consisted of the application of 20 kg·ha⁻¹ of N, 300 kg·ha⁻¹ of P₂O₅, and 200 kg·ha⁻¹ of KCl. The cultivar Ito was planted on 7 May 2018, in three double rows per seedbed, disposed of longitudinally, spaced 0.1 × 0.1 × 0.4 m (between cloves on the line, between rows, and between the double rows). In each double row of planting, 32 cloves were planted. During data collection, the two central lines were considered useful, and 0.3 m of the extremities were neglected.

The experiment consisted of three depths of irrigation (60%, 80%, and 100% of the soil field capacity) and three depths of fertirrigation with N (60%, 80%, and 100% of the recommended dose) applied during the period of clove differentiation. Topdressing was performed via fertigation at 8, 15, 22, 60, 70, 80, and 90 d after planting with 30 kg·ha⁻¹ of N and 20 kg·ha⁻¹ of K with formulated fertilizer 30–00–20. The x and y directions of the cartesian coordinate system were defined; at the end of the garlic phenological cycle (15 Sept. 2018), the experimental grid was staked in plots spaced at 1.6 m. Each experimental grid comprised three transects of 48 × 1.6 m.

Irrigation of the garlic was performed by the drip irrigation system using a distribution line in the middle of the double row of garlic planting, with drippers spaced 0.20 m apart. Determination of the water application rate of the drippers was performed following the method described by Bernardo et al. (2006). The need for water application by the garlic culture was monitored using Irrigâmetro (UFV, Viçosa, Brazil) installed in the experiment area. The readings of the Irrigâmetro were performed daily at 8:00 am. The water level applied during each irrigation was enough to restore the soil moisture content to the field capacity.

At 51 d after planting, the water depth applied to all seedbeds was controlled so that at the beginning of the controlled water deficit period, the water in the soil was at field capacity. At 82 d after planting, irrigation with water depth and N variation was started. A total of four irrigations with variations were performed at 82, 86, 101, and 107 d after planting.

From the total of 90 sampling points (beds), 66 of these points were irrigated with 100% of the water requirement of the crop (without water restriction), estimated by the Irrigâmetro. Of the remaining 24 sampling points, 12 points were irrigated with 80% and 12 points with 60% of the crop water requirement until the garlic harvest phase. The last irrigation was performed 15 d before harvest.

The experiment was harvested at 130 d after planting, during the senescence phase of the plants, when they had approximately six green leaves. The determined production components were individually collected in the sample area, which was composed of a double line with a length of 1.0 m, totaling 20 plants. Garlic bulb yield (BY) was determined by weighing all harvested bulbs and values were expressed in kg·ha⁻¹. Irrigation (IR) with restriction applied in the formation of the bulbs was determined by measuring the amount of water applied (mm) at all sampling points throughout the development cycle of the crop. Nitrogen applied via fertirrigation with a restriction on bulb formation was determined by measuring the amount of nutrient applied in the topdressing (kg) at all sampling points. The LSG was obtained using the percentage of plants with secondary bulb growth at each sampling point. The bulb diameter (BD) and bulb length (BL) were measured with the aid of a caliper, and the values were expressed in mm. The leaf area index (LAI) was measured on the day of harvest using the PAR/LAI Ceptometer model CP-80 (Decagon Devices, Inc., Pullman, WA). The bulb volume (BV) was calculated as a function of BD and BL and the values were expressed in cm³.

Descriptive analysis was performed. Semi-variogram adjustments and semi-variance estimation were performed to estimate the coefficients of the theoretical model for the semi-variogram called the nugget effect (C₀), sill (C₀+C), and range (A₀). The data were interpolated by kriging to allow visualization of the spatial distribution patterns of the garlic production components through maps. se maps of kriging prediction were generated. Cross-validation is a tool used to evaluate alternative models of simple and crossed semi-variograms.

Results and Discussion

With the exception of LSG, all variables showed cv values less than 20% (Supplemental Table 1). The IR, NI, and BV components fit the spherical model, whereas the components BY, BD, and LAI fit the exponential model (Table 1). The LSG and BL components fit the Gaussian model. According to Cambardella et al. (1994), the values of the NI and LAI components presented strong spatial dependence evaluation (SDE), whereas BY, IR, LSG, BD, BL, and BV presented moderate SDE.

Table 1.

Estimated parameters for the simple semi-variogram of garlic crop production components.

View Table

The semi-variograms of garlic production components and external components are shown in Supplemental Fig. 1. The simple kriging maps of the garlic crop production components are shown in Supplemental Fig. 2. These maps allow visualization of the spatial variability of each of the components in the study area. Supplemental Table 1 shows the cross-validation parameters for garlic yield kriging and plant phenological indexes. The increasing ratio of these, as analyzed by the magnitude of the correlation coefficient (r), was as follows: BD (0.14) > BV (0.23) > BL (0.35) > BY (0.35) > LAI (0.45) > LSG (0.53) > IR (0.67) > NI (0.83).

When the cross-semi-variogram was performed between plant phenological indices, positive spatial correlations (Supplemental Table 2) for the BY = f(BL), BY = f(BD), BD = f(BL), LSG = f(BY), LSG = f(IR), BD = f(IR), LSG = f(NI), BD = f(NI) were found to have SDE values of 34.7%, 31.4%, 50.0%, 100.0%, 89.1%, 95.0%, 78.4%, and 99.9%, respectively. These results show that there is a direct relationship between the yield, height, and diameter of the bulb, and that the increase in the diameter of the bulb implies an increase in yield, and that the percentage of garlic LSG increases with the bulb yield. Another important observation was the direct relationship observed between garlic LSG and irrigation and N; the increase of irrigation depth and fertigation in the clove formation phase implies higher percentages of plants with garlic LSG.

Negative spatial correlations were obtained for BL = f(IR), BV = f(IR), BL = f(N), and BV = f(N), with SDE values of 86.7, 97.9, 80.0, and 97.1 respectively. When analyzing these indices, it was concluded that the increase in irrigation and fertigation during the formation phase of the cloves imply lower volumes of bulbs. These results will be very useful for garlic producers because with the reduction of irrigation and fertigation during the formation phase (differentiation) of cloves, the producer will have bigger bulbs and a lower incidence of garlic.

Conclusions

Garlic yield is directly related to the height and diameter of the bulb. Garlic LSG is directly related to yield. Water depths and N doses applied during the differentiation phase of cloves are directly related to the garlic LSG. The increases in irrigation and nitrogen fertigation during the formation phase of the cloves imply lower volumes of bulbs.