Fertilizer-nitrogen Management in an Onion and Tropical Pumpkin Rotation in Puerto Rico (2024)

Table 1.

Soil fertility characterization [0–15 cm (5.9 inches)] of plots in Guánica, PR before onion planting.^z

Treatment descriptions.

In onion, we evaluated fertilizer-N levels of 140, 196, and 252 kg·ha⁻¹. Pre-plant fertilization to the raised beds consisted of broadcasting 22 kg·ha⁻¹ of N, 10 kg·ha⁻¹ of phosphorus (P), and 33 kg·ha⁻¹ of potassium (K). The difference between the amount of preplant N fertilization and the fertilizer-N total of each treatment was applied by fertigation at 10 weekly intervals starting 2 weeks after planting (WAP). Onion fertigations were initiated on 12 Nov. 2012 and ended on 14 Jan. 2013. The sources of fertilizer-N used were a combination of urea, ammonium sulfate (AS) and potassium nitrate (PN), with N ratios (urea-N:AS-N:PN-N) of 2:1:1 for 140 kg·ha⁻¹, 2.7:1.7:1 for 196 kg·ha⁻¹, and 3.5:2.5:1 for 252 kg·ha⁻¹ treatments. All treatments received the same P level. A total of 139 kg·ha⁻¹ of K was applied and potassium sulfate was used as the K source for those treatments that needed supplemental K fertilization. Variable fertilizer-N rates were applied for each of the 10 fertigations. In general the fertigation treatment consisted of a 15-min irrigation, followed by 30 min of fertilizer injection, and a final 15-min irrigation.

The tropical pumpkin fertilizer-N treatments were 112 and 280 kg·ha⁻¹ by fertigation. The low N treatment was applied to tropical pumpkin in plots previously used for onion at the lowest N level (140 kg·ha⁻¹) and the high N treatment was applied to plots previously planted in onion with the highest fertilizer-N treatment (252 kg·ha⁻¹). Eight fertigations were applied over a 10-week period. For both onion and tropical pumpkin, the irrigation water had 4 mg·L⁻¹ NO₃-N and was estimated to contribute 1 kg·ha⁻¹ of N for every inch of water applied.

Plant measurements.

For onion, plant density, number of leaves, plant height, sap NO₃, and leaf color were measured at 7 and 10 WAP. Within a plot, observations were made on 10 randomly selected plants within a randomly placed 16-ft² quadrant. The date at which 50% or more of the plants had bulbs was recorded. Leaf color (plant greenness) was assessed on a scale of 1 to 4 using the IRRI Leaf Color Chart (Witt et al., 2005). Fifteen plants from each plot were sampled for leaf nutrient indicators on 25 Jan. 2013 [95 d after planting (DAP)]. Leaves were analyzed for total elemental analysis [N, P, K, calcium (Ca), magnesium (Mg), sulfur (S), iron (Fe), manganese (Mn), zinc (Zn), copper (Cu), and boron (B)] by AgSource Laboratories (Bonduel, WI). Data from leaf nutrient indicators was combined across all fertilizer-N treatments and used to calculate 95% confidence limits (CL). At maturity, plant vegetative and fruit biomass were estimated for each plot. Above-ground plant biomass from 10 plants was cut at ground level. All vegetative material was weighed and a subsample was analyzed for moisture to determine the vegetative biomass dry weight. Onion bulbs were harvested 1 Mar. 2013. Bulb tops were cut, and bulbs were lifted and allowed to dry for several days. Bulbs were then harvested by hand from a 20-ft segment of one of the middle beds of a nine-bed plot. Harvested bulbs were separated into size classes of extra small (< 1.75-inch diameter), small (1.75- to 2-inch diameter), medium (2- to 3-inch diameter), large [or jumbo (3- to 3.75-inch diameter)], and colossal (>3.75-inch diameter), counted and weighed. The mature air-dried aboveground vegetative plant material and the bulbs were analyzed for total elemental analysis as specified previously. A subsample of bulbs was analyzed for moisture to estimate bulb dry weight. Nutrient extraction was calculated based on the nutrient concentration and the bulb dry matter, and nutrient uptake was calculated based on the nutrient concentration and vegetative dry matter. Onion had a mean moisture content of 919.1 g water per kilogram fresh weight.

A random sample of 15 tropical pumpkin leaves per plot were taken 8 July 2013 (85 DAP) and analyzed for total elemental analysis in the same manner as for onion. Fruit from all seven beds of tropical pumpkin plots were harvested, counted and weighed 19 July 2013. Vegetative biomass dry weight was estimated by harvesting plant material in a 1-m² quadrant and correcting for percentage moisture. Fruit biomass dry weight was estimated from a subsample of all harvested fruits.

Soil measurements.

Soil was sampled within each plot before onion planting, during the onion growing period, at the end of the onion harvest, and at the end of the tropical pumpkin harvest. Soils were sampled pre-plant and post-harvest at 0–15, 15–30, 30–60, and 60–100 cm. On occasion, additional samples were taken when a soil textural discontinuity was observed. Soils were stored in plastic bags and transferred to the laboratory and air-dried and then sieved to pass a 4-mm mesh. Air-dried and sieved soil samples were stored in the bags in the laboratory until analysis.

The surface 0 to 15 cm soil interval was analyzed for soil fertility parameters by AgSource Laboratories (Lincoln, NE). Profile soil samples were analyzed for 1m potassium chloride (KCl) extractable ammonium-N (NH₄⁺-N) and NO₃-N (inorganic N). Inorganic N concentrations were expressed on a kilogram-per-hectare basis after correcting for the corresponding sampling depth interval 1 and a soil bulk density of 1.35 g·cm⁻³. Immediately available N was estimated as the amount of inorganic N within the 0- to 15-cm and 15- to 30-cm depth intervals. Potentially leached N was estimated as the inorganic N within the 30- to 60-cm and 60- to 100-cm depth intervals. Total profile N was estimated as the depth integrated NO₃-N concentration to a depth of 100 cm.

Climatologic data and irrigation.

Precipitation and air temperature data were gathered from a nearby weather station. Each raised bed contained two drip irrigation lines (Chapin™; Jain Irrigation Systems, Maharashtra, India) with a flow of 0.152 gal/min per 100 ft at 10 psi with 12 inches between emitters. Soil water in selected beds was monitored during the onion growing season with tensiometers (Model “R”; Irrometer Co., Riverside, CA) placed in the seed row at a depth of 20 cm. The field was irrigated 24 times during the growing season for a total irrigation depth of 18.54 cm. The field was irrigated when tensiometers indicated soil moisture of less than −30 kPa.

Soil-solution composition evaluation.

Eight suction-cup lysimeters (Irrometer Co.) were installed at 6- and 12-inch depths. Four lysimeters were placed in the lowest fertilizer-N treatment plots during onion production and four were placed in the highest fertilizer-N treatment plots, each at two soil depths, 6 and 12 inches. Data were collected from three fertigation events during onion production. To obtain a soil solution, a tension of about −80 kPa was applied, and between 8 and 12 h later, soil solution samples were obtained. Samples were taken before, during, and after fertilizer application. Samples were stored frozen until analysis for electrical conductivity and NO₃ and NH₄ (AgSource Laboratories).

Economic analysis.

Final harvest yields were expressed as fresh onion or tropical pumpkin weight per hectare. Estimated onion gross return per hectare was calculated based on harvest prices of $783/t for jumbo and colossal, $712.8/t for large, $557/t for medium, and $402.6/t for small bulbs. The harvest price of tropical pumpkin was $0.35/kg of fruit (C. González, personal communication). The only variable cost was N fertilizer depending on the fertilizer level applied. All other costs were considered fixed.

Statistical analysis.

Analyses of variance (completely randomized design) were performed using InfoStat (Di Rienzo et al., 2014) to test treatment effects. All statistical comparisons were made at P < 0.05 and means separation was done with Fisher’s least significant difference test when appropriate.

Onion agronomic indicators.

Mean onion plant density was low due to poor seed germination ranging from 304,287 to 326,142 plants/ha with no differences among N treatments. The highest fertilizer-N treatment resulted in faster bulb formation (bulb formation was defined as swelling has begun) at 10 WAP (P < 0.05) with 55% of the plants having bulbs as compared with 29% and 33% with the 140 and 196 kg·ha⁻¹ N treatments, respectively. One week later, 69% of the plants had bulbs, and there was no significant difference among treatments. Previous reports from Puerto Rico suggest that high N fertilization rates are associated with prolongation of bulb formation (UPR-AES, 2012a).

At 7 and 10 WAP, there were some differences (P < 0.05) in onion plant height and number of leaves among the three N fertilizer rates (Table 2). However, the trends were not in a consistent direction, and there were no differences in plant height and number of leaves between the low (140 kg·ha⁻¹) and high (252 kg·ha⁻¹) fertilizer-N treatments.

Onion leaf color (an indication of N status) did not vary among N treatments at either 7 or 10 WAP (Table 2). At both 7 and 10 WAP, onion sap NO₃-N was significantly greater (P < 0.05) for the highest fertilizer-N treatment (252 kg·ha⁻¹) compared with the intermediate N rate (196 kg·ha⁻¹).

We also looked at leaf K, Ca, Mg, S, Fe, Mn, Zn, Cu, and B (data not shown). Due to the general lack of clear treatment effects, we used the means averaged over the three N fertilizer treatments to calculate the ± 95% CL for each nutrient (Table 3). These CL can be used as a guide to improve the local database in regards to leaf-based sufficiency levels in onion.

Onion nutrient uptake.

Onion bulb biomass and bulb N represented 57% and 54% of the total biomass and N uptake, respectively (Table 4). The mean harvest index across treatments was 0.57 and unaffected by fertilizer-N levels. Of all nutrients taken up by the crop (biomass plus bulb), the highest rate of nutrient uptake occurred for K. An estimated 171 kg·ha⁻¹ of K was taken up, of which 45% was removed in the harvested portion (onion bulb) of the plant. Fertilizer-N uptake was 123 kg·ha⁻¹, which was lower than the lowest fertilizer-N rate applied. Of the total biomass N removed, nearly 46% was expected to be returned and recycled to the soil for the next crop.

Onion size classification.

Onion size classification was unaffected by fertilizer-N levels (data not shown). The colossal size class was less than 3% of all marketable onions and the small-size classification was less than 12% of all onions. As expected for the variety planted, an average of between 54% and 61% of the onions were in the medium-size class and between 21% and 29% were in the large-size class.

Onion yield.

All of the harvested material was free from visible pathogens. Only 5% or less of the total harvest was not commercially acceptable. Fertilizer-N rate did not affect the number of marketable bulbs (mean of 173,942 bulbs/ha) or marketable onion yield (mean of 40,168 kg·ha⁻¹) (Table 5). Marketable yields in this study were similar to typical yields for onions for the growing conditions of Puerto Rico, estimated at 39,282 kg·ha⁻¹ (UPR-AES, 2012a). The experimental field had fertilizer-N applications of 175 kg·ha⁻¹ of N during the 2 years before the study. These rates of fertilizer-N added may have resulted in large amounts of residual soil N, part of which could have been available during the 2012–13 onion growing season. Other experiments in Puerto Rico and other areas have shown no response to fertilizer-N as a result of previous N fertilization or in the case of Halvorson et al. (2008) when onion followed soybean in rotation. The mean preplant soil NO₃-N (0 to 30 cm) was 95.7 kg·ha⁻¹ of N (Fig. 1), and could explain lack of response above the 140 kg·ha⁻¹ of N fertilizer treatment.

The results suggest that onion growers can take a conservative approach to fertilizer-N application for onion production in situations similar to that of the experimental location of in Guánica, Puerto Rico, including similar historical rotations and nutrient management programs. N fertilizer rates of less than 140 kg·ha⁻¹ can result in optimum onion yields, compared with the usual over 200 kg·ha⁻¹ of N used by many vegetable producers in Puerto Rico.

Tropical pumpkin nutrient uptake.

Tropical pumpkin produced an average total dry weight of 8692 and 249 kg·ha⁻¹ of N, respectively; fruit dry matter and fruit N represented 67% and 69% of the total, respectively (Table 4). Mean harvest index across treatments was 0.67 and unaffected by fertilizer-N levels. N uptake was highest for the 280 kg·ha⁻¹ of N treatment, which removed 279 kg·ha⁻¹ of N, of which 67% was removed in the harvested portion (pumpkin fruit) of the plant. In plots of the lower fertilizer-N treatment, tropical pumpkin fruit removed 70% of the total N uptake.

Tropical pumpkin yield.

Both the number and yield of marketable fruit of tropical pumpkin was about 50% greater in the high fertilizer-N treatment (280 kg·ha⁻¹) (Table 5). Yields were similar to those previously reported for ‘Soler’ tropical pumpkin (Wessel-Beaver, 2005). About 4.6% of the total harvest was not commercially acceptable. The preplant tropical pumpkin profile (0 to 100 cm) of NO₃-N was greater than 200 kg·ha⁻¹ of N (Fig. 1), and in spite of this, there was a crop response to fertilizer-N.

Economic analysis for onion and tropical pumpkin production.

Fertilizer-N and complementary nutrients (P, K, and micronutrients) costs ranged from $632/ha to $945/ha for onion and from $570/ha to $1110/ha for the additional fertilizers added to the following tropical pumpkin crop. Estimated production costs (at 2012 prices) for onion and tropical pumpkin were $20,258/ha (UPR-AES, 2012a) and $6200/ha, respectively (UPR-AES, 2012b). Fertilizer-N used in this experiment represented 3.1% to 2.8% of estimated production costs in onion and 9.2% to 17.9% of the estimated production costs in tropical pumpkin. The gross profit obtained for onion production was typical for similar locations in Puerto Rico (UPR-AES, 2012a). The gross profits estimated based on yields obtained in our experiment were at or near the farmer’s production costs estimated at $19,690/ha, using 195 kg·ha¹ of N (C. González, personal Communication). In tropical pumpkin, for each additional $1.00 invested above 112 kg·ha⁻¹ of N there was a benefit of $11.30 for 280 kg·ha⁻¹ of N, respectively (Table 6).

Residual soil NO₃-N.

To examine if there was an increase in soil profile inorganic N as a result of fertilizer-N application, we performed a partial soil and crop-N budget for the onion-tropical pumpkin rotation in the low (140 kg·ha⁻¹ for onion + 112 kg·ha⁻¹ for tropical pumpkin = 252 kg·ha⁻¹) and high (252 kg·ha⁻¹ for onion + 280 kg·ha⁻¹ for tropical pumpkin = 532 kg·ha⁻¹) fertilizer-N plots. Before onion planting, there was similar soil inorganic N at 0–30 cm (available N) but greater potentially leached N at the 252 kg·ha⁻¹ N treatment (Fig. 1). At onion harvest, similar amounts of N were removed in the onion bulb and vegetative dry matter in both fertilizer-N treatments. If applied fertilizer-N, irrigation-N, and N in the soil at 0 to 30 cm were all considered as available to the crop, then 25% and 47% of available N was removed in the onion bulb and total crop, respectively, at the lowest fertilizer-N treatment, and 21% and 37% of available N was removed in the same crop pools at the highest fertilizer-N treatment. The lower crop-N use efficiency under the highest fertilizer-N treatment is reflected in the greater amount of soil N at 0 to 30 cm (316 kg·ha⁻¹) remaining after onion harvest and before tropical pumpkin planting in plots receiving 252 kg·ha⁻¹ N, compared with plots receiving 140 kg·ha⁻¹ N (which had 158 kg·ha⁻¹ soil N at 0 to 30 cm). When tropical pumpkin was harvested, 66% of the available N was used by the crop (fruit and vegetative) for the low fertilizer-N treatment while only 42% of the available N was used by the crop in the high fertilizer-N treatment.

When the soil was re-sampled nearly 1 year after onion was first planted, similar soil N amounts were found in the soil profile at both 0–30 and 30–100 cm (Fig. 1). Thus we presume that the difference in the N use efficiency between the 252 and 532 kg·ha⁻¹ fertilizer-N treatments resulted in N losses by leaching, volatilization, denitrification, and any runoff that occurred after tropical pumpkin planting and before soil sampling. The estimated amount of N losses can be calculated based on the differences between the amounts of N applied and the N removed in the crop and not measured in the soil. We assumed an annual soil-N mineralization rate of 80 kg·ha⁻¹ of N for both treatments (D. Sotomayor, unpublished data). In the low fertilizer-N treatments a net gain of 24 kg·ha⁻¹ of N to the soil was estimated, with 124 kg·ha⁻¹ of N remaining in the soil profile below the rooting zone at 30 to 100 cm. In the high fertilizer-N treatment a potential net loss of 175 kg·ha⁻¹ of N from the soil was estimated with 126 kg·ha⁻¹ of N remaining in the soil profile below the rooting zone.

Soil solution N.

The observed increase in soil profile N with increasing fertilizer-N addition should also have been reflected in soil solution measurements taken during the course of the onion growth period. Each fertigation event delivered an estimated 14 and 28 kg·ha⁻¹ of N for the low and high fertilizer-N treatments, respectively. Based on the irrigation water volume and soil moisture percentage, the theoretical solution concentration was calculated as between 73 and 150 mg·L⁻¹ of N for the low and high fertilizer-N treatments, respectively. Fifty percent of the N was in urea form and 25% and 14% of the N was in NO₃ form for the low and high fertilizer-N treatments, respectively. Soil solution NH₄-N concentrations were less than 1 mg·L⁻¹ N in most cases, even though NH₄-N in fertigation solution was between 25% and 36% of the total N applied. This suggests that N in NH₄⁺ form was rapidly nitrified. No significant differences (P > 0.05) were observed in the amount of soil solution inorganic N in samples collected before fertigation and after the fertigation events which suggests that inorganic N is persistent in the soil solution, even days after fertigation (Fig. 2). However, during two fertigations inorganic N concentrations increased immediately after fertigation. Mean inorganic N concentrations were 66 and 58 mg·L⁻¹ of N at 15 cm and 73 and 61 mg·L⁻¹ of N at 30 cm for the 140 and 252 kg·ha⁻¹ of N, respectively.