Export and nutrient partitioning in organic onion

Early season onion crop, despite lower yield, is an opportunity for farmers to explore market and better prices in southern Brazil. Knowing the amount absorbed and the distribution of nutrients in the plant is essential for adequate management of fertilization. However, little information on this matter is available for onion, especially for organic farming and nontraditional periods in southern Brazil. The objective of this study was to evaluate the nutrient absorption and partitioning in open-pollinated onion cultivars grown in organic system. The experiment was conducted in the Canguiri-UFPR experimental organic farm, Curitiba-PR, with sowing in January. The experiment was arranged in a completely randomized design with three replications. Cultivars Franciscana IPA-10 (purple), Vale Ouro IPA-11 and Brisa IPA-12 (from Empresa Pernambucana de Pesquisa Agropecuária - IPA), Alfa Tropical (from Embrapa Hortaliças), Alfa São Francisco (VIII cycle) and Alfa São Francisco-RT (thrips resistant– genotype under test; from Embrapa Semi-árido), and BR-29 (Topseed-Agristar) were selected for the study. Chemical analyses were performed for shoot and bulbs collected at harvest. The production potential of cultivars varied, and the most productive ones were the most efficient in converting the nutrients absorbed in bulb yield. The order of nutrient contents in the shoots was K > N > P > Ca > Mg > Fe > Zn > Mn = B > Cu, whereas in the bulbs it was K > N > P > Ca > Mg > Fe > B > Zn > Cu > Mn. Nutrients were required in the following order of amount K > N > P > Ca > Mg > Fe > Zn > B > Mn > Cu, per ton of fresh bulbs, and accumulated in greater quantity in the shoot, except Zn, which had higher concentration in the bulb.


INTRODUCTION
Onion (Allium cepa) is a very important commercial crop in Southern Brazil. Currently, the region accounts for 52% (747,133 ton) of the total produced in the country (1,444,146 tons), with 376,603 tons produced in Santa Catarina, followed by Rio Grande do Sul with 207,089 tons, and Paraná with 163,441 tons (IBGE, 2012). With a cultivated area of 7637 ha, the state of Paraná is the sixth largest producer, with 65% of production coming from the Metropolitan Region of Curitiba (RMC), equivalent to approximately 130,200 ton (SEAB/DERAL, 2009).
However, because of the off-season period from June to September, the state of Paraná needs to import onion from São Paulo and Santa Catarina, as well as from other countries, mainly Argentina (40,500 ton) (CEPA, 2012). A way to meet this demand is the introduction of new cultivars adapted to a nontraditional crop period.
In addition, since much of the RMC is currently considered environmental protection area (State Law No. 12,248/98), any cultivation must be in the organic system, which demands greater adaptation to climate, nutrient absorption capacity and disease resistance of the crops to be implemented.
The production of organic onions is an alternative to be considered, aiming to meet the growing consumers' demand for organic farming products, with higher benefit and attractive features such as the white and purple varieties, with specific market niches (Boeing, 2002).
According to Melo & Ribeiro (1990), the onion germplasm is made up of local populations and cultivars developed over centuries to adapt to different latitudes, farming systems and local consumer preferences. Assessing these characteristics, the Instituto Agronômico de Pernambuco -IPA (Pernambuco Agronomic Institute) in partnership with Embrapa Semi-Árido Unit, developed onion breeding programs (Buzar et al., 2007), resulting in the IPA cultivars. Another national breeding program, developed by Embrapa Semi-Árido, along with Embrapa Hortaliças Unit (Embrapa Vegetables Crop Unit) produced the Alfa Tropical cultivars, and more recently the Alfa São Francisco.
When introducing new cultivars in a region, besides evaluating their adaptability to environmental conditions and the farming system, one also should consider their nutritional requirements, as these traits are genotypically different and reflect the growth and final yield of the species (Vidigal et al., 2003).
Partitioning or distribution of nutrients in different parts of the plant is important to estimate the export and return of nutrients to the soil (May et al., 2008). The little information on nutrient uptake by onion grown in Brazil indicates that K followed by N and Ca are the most absorbed nutrients (Pôrto et al., 2006, Vidigal et al., 2010, and that the nutrients N, P, K and S accumulate preferentially in the bulb, while Ca and Mg in the shoots (Pôrto et al., 2006). Regarding micronutrients, Vidigal et al. (2000) reported that the bulb accumulated most of Zn, Fe and B, while Alloway (2004) found that the onion is demanding in Zn, Mn, Cu and Mo, but little susceptible to B deficiency. Vidigal et al. (2010) reported that the cultivar Alfa Tropical had the following order of nutrient absorption: Fe > Mn > Cu > Zn, in Minas Gerais. It is known that both the absorbed amount and distribution of nutrients in the plant varies among cultivars (Santos et al., 2007).
In Brazil, data on nutrient export in onions are scarce, particularly micronutrients and especially in organic farming.
The present study aimed to determine the absorption and nutrient partitioning between shoot and bulb in openpollinated onion cultivars in organic system, off the traditional growing season.
The experiment was arranged in a was completely randomized design with three replications and seven treatments with the open-pollinated onion cultivars Franciscana IPA-10 (purple), Vale Ouro IPA-11, Brisa IPA-12, Alfa Tropical, Alfa São Francisco, Alfa São Francisco-RT, and BR-29. Seeds were sown in January and plants were harvested in August, 2009. Seedlings were obtained by the sowing and transplanting method. Sowing was carried out off the time recommended for the region, between June and July, and harvest between November and January (Mendonça et al., 2004).
Four 30-cm-apart rows were prepared per raised bed, plants were spaced 15 cm apart, in 1.20 x 4.05 m plots with 104 plants. The plot harvest area consisted of the two central rows by 3.5 m. Seedlings were transplanted on March 20, 2009 (55 days after sowing). At 40 days after transplantation, 80 kg ha -1 of potassium sulfate (50% K 2 O and 17% S) was applied to the soil.
Harvest was carried out at 99 days after transplanting (DAT), when about 75% of the plants reached the popping stage. Yield was expressed in kg per hectare of bulb fresh matter. The bulbs were separated from shoots at leaf insertion, dried in an forced air circulation oven at 60 °C for 72 h, and weighed. The results were extrapolated to hectare. The harvest index was calculated by dividing bulb dry matter by mass of total dry matter (shoot + bulb).
At the harvest day, four plants were collected from the plot harvest area, washed in deionized water, separated into shoot and bulb without roots, and dried in a forced air circulation oven to constant weight. After drying and weighing, the material was ground for chemical analysis. Chemical analyses of tissue were performed at the Biogeochemistry Laboratory of the Department of Soil Science and Agricultural Engineering -UFPR, according to Martins & Reissmann (2007), except for boron, which was performed according to Bataglia et al. (1983). N content was determined by dry combustion (975 ºC), using the Vario El III CHNOS Elemental Analyzer.
Extraction and export of nutrients were obtained by multiplying the nutrient content by the dry matter yield of each part of the plant. For extraction, it was considered the sum of the nutrients present in the bulbs and shoots, while for export, it was considered only the bulbs.
Data variances were tested for homogeneity using the Bartlett's test and the treatment means were compared by the Tukey's test at 5% probability. Data was analyzed by the M-STAT software, version 2.11 (Michigan State University, 1989).

RESULTS
The highest bulb mean yields were recorded for cultivars Alfa São Francisco-RT followed by Alfa São Francisco and Alfa Tropical, and the lowest yields for cultivars BR-29 and IPA-10. Shoot yield ranged from 13,394 to 21,880 kg ha -1 , with cultivars BR-29 and Alfa Tropical higher than IPA-12 and IPA-10, but without differing from the others (Table 1). Bulb moisture ranged from 92.68 to 93.96%, and was close to shoot moisture, which was between 92.17 and 93.71%. The harvest index ranged from 0.23 for cultivar BR-29 to 0.48 (more than twice) for Alfa São Francisco -RT, and this index was directly proportional to yield (Figure 1).
The content of N and P did not vary among onion cultivars, both in the shoot and in the bulbs (Table 2), while the K content in shoot was highest in Alfa São Francisco-RT, followed by BR-29 and Brisa IPA-12, and lowest in Franciscana IPA-10. In the bulbs, K levels were highest in Alpha São Francisco-RT and lowest in Alfa Tropical (Table 2).
Ca contents in onion shoots were highest in Alfa São Francisco-RT and lowest in BR-29. In bulbs, Ca contents were highest in Franciscana IPA-10 and Brisa IPA-12 and lowest in Alfa São Francisco-RT, Alfa São Francisco and BR-29. The Mg content in the bulbs were similar among the cultivars, except for BR-29, which had the lowest content (Table 2).
Comparing shoots with bulbs, no difference was observed between the content of N and P, except for cultivar Alfa São Francisco-RT, which had the highest content of P in the bulbs. K, Ca and Mg contents were higher in the shoots, except for cultivar IPA 12, which had higher Mg content in bulbs ( Table 2).
Most of the cultivars showed macronutrient content in the following order: K > N > P > Ca > Mg, both in shoots and bulbs. The exceptions were Franciscana IPA 10 and BR 29, which showed higher content of N than K, and Alfa São Francisco-RT that concentrated more Ca in shoots, with content of this nutrient greater than P in the shoot and the lowest content among the other cultivars in the bulbs. Rev. Ceres, Viçosa, v. 63, n.5, p. 683-690, set/out, 2016 Cu contents in onion shoots were highest in Vale Ouro IPA-11 and lowest in Franciscana IPA-10 and Alfa Tropical. In bulbs, Cu contents were highest in BR-29 and lowest in Vale Ouro IPA-11. Mn content in shoots was highest in Vale Ouro IPA-11 and lowest in Franciscana IPA-10, while in bulbs, this nutrient was highest in cultivar BR 29 and Vale Ouro IPA-11 and lowest in BR 29 (Table 2).
Fe was the micronutrient found in the highest content in onion, both in shoots and bulbs. In shoots, Fe contents were in highest in Franciscana IPA-10 and Alfa São Francisco-RT, and lowest in Alfa São Francisco and BR-29 (Table 2). In bulbs, this nutrient content divided the cultivars into two groups: cultivars Franciscana IPA-10, Vale Ouro IPA-11, Brisa IPA-12 and Alfa Tropical showed on average 2.5 times higher Fe content in this part of the plant than cultivars Alfa São Francisco, São Francisco Alfa-RT and BR-29.
The Zn content in shoots did not vary among cultivars. In bulbs, Vale Ouro IPA-11 showed higher content than the Brisa IPA-12. Unlike Zn, B content varied among cultivars only in shoots, being the lowest in cultivar BR-29 (Table 2).
In shoos, the decreasing sequence of micronutrient content was Fe > Zn > Mn = B > Cu in all cultivars, while in bulbs, the sequence was Fe > B > Zn > Cu > Mn, except for Vale Ouro IPA-11 with Mn content higher than Cu content.
In relation to the plant nutrient content, the least productive cultivars (Table 1), and especially BR 29, showed the largest amounts of nutrients extracted per ton of fresh bulb (Table 3). Likewise, the most productive cultivars (Table 1), especially Alpha São Francisco RT, accumulated the smallest amounts of nutrients per ton, therefore they were more efficient in nutrient use. The export of nutrients (content in bulbs) varied less than extraction among cultivars, especially macronutrients (Table 3), with no relation to bulb yield. Considering the average of all cultivars, about 36,41,32,28,31,47,27,47,68 and 31% of N,P,K,Ca,Mg,Cu,Mn,Fe,Zn and B,respectively, were allocated to the bulbs, that is, except for Zn, all nutrients are concentrated in greater amounts in the shoots of onion. In general, both the level (Tables 2 and 3) and the content (Table 3) of nutrients were higher in the shoots of onion. However, the micronutrients Cu, Fe and B showed higher levels in the bulbs, but higher contents in the shoots, the opposite was true with Zn, which was the only nutrient that even with the highest level in the shoots showed higher content in the bulbs.
Regardless of the cultivars, the sequence of nutrients in relation to the amounts extracted (whole plant) was K > N > P > Ca > Mg > Fe > Zn > B > Mn > Cu. The amount considered as export (bulbs) also showed the same sequence for most cultivars, changing only for higher Cu values compared to Mn. The exceptions were cultivars Franciscana IPA 10 and BR 29, which had N export higher than K; however with very similar values. Cultivar BR 29 showed the following export sequence: Cu > B > Mn; and Vale Ouro IPA 11 had Mn export higher than Cu export (Table 3).

DISCUSSION
Variation of yield potential among cultivars was observed for organic and off-season farming in the southern region of Brazil, with cultivar Alfa São Francisco-RT showing the greatest yield potential. This higher yield potential is associated with the higher harvest index, i.e., greater portion of bulb mass relative to the total mass accumulated by the plant. Therefore, the higher bulb yield was not followed by shoot yield, suggesting that the leaf area in all cultivars was sufficient to meet demand  Bettoni et al. (2012), who obtained yields between 15 and 18 ton for cultivars Alfa São Francisco and Alfa São Francisco-RT. The genetic variability among the cultivars did not affect N and P contents in the shoots and bulbs, however, significantly influenced the amount accumulated in the plant. Although the most productive cultivars extracted larger amount of N and P per area, they used less N and P for each ton of bulb produced. Despite the variation between cultivars in total content, the amount of N and P in the bulbs was not changed. This suggests that there is a control by the plant for the amount of these nutrients accumulated in the bulb, which was kept in balance with the shoot. This is also observed for Mg and K, which are also mobile nutrients in the plants.
N and P, respectively, were the second and third nutrients most required by onion, which was also reported by other authors (Pôrto et al., 2006;Vidigal et al., 2010) and other vegetable crops (Filgueira, 2003) .
K was the nutrient most accumulated in the shoots and bulbs of onion (Table 2) and the most extracted one (Table 3). However, the onion response to fertilization with K are generally limited (Magalhães, 1993), contrary to N (Brewster, 1994). Although being the nutrient most absorbed by onion, only about 18% was allocated to the bulbs (Table 3), which makes their content per bulb ton similar to N. This appears to contrast with the K function which, in species that accumulate organic compounds in the bulbs, has a role in the transport of solutes and consequently the expansion of cataphylls, influencing the growth and size of the bulbs (Mógor, 2000). Porto et al. (2006) and Vidigal et al. (2010) also found K to be the most accumulated nutrient in onion.
In general, the cultivar Alfa São Francisco-RT, with the highest bulb yield, showed the highest K, Ca and Mg content in the shoots, and BR 29, with the lowest bulb yield, had the lowest levels these nutrients, especially calcium and magnesium. The highest levels in the most productive cultivar indicate its greater demand for these nutrients. There was little genetic influence on the partition of macronutrients between shoot and bulb, whose contents was higher in the shoot, which is in agreement with other authors (Pôrto et al., 2006;Santos et al., 2009).
Nutrient content can be used to determine the amount of nutrient extracted and exported by the species. In the first case, one considers the nutrient present in the whole plant and an estimate of the nutrient requirement by the species is obtained. In the second case, one can use the information to tailor the fertilizer application according to the productivity of each crop. The most productive cultivars had the lowest nutrient content per bulb ton produced, which shows greater efficiency of these cultivars in using the nutrients absorbed.
The descending order of nutrient levels found in most cultivars for both shoot and bulbs, which was K > N > P > Ca > Mg, differs from other authors' reports. According to Santos et al. (2007), the cultivars IPA-10 and Alfa São Francisco showed the following order of levels in the shoots: N > Ca > K > Mg > P and N > Ca > K > P > Mg, respectively. On the other hand, May et al. (2008) reported nutrient levels as N > P > K > Ca > Mg for cultivars Optima and Superex. Vidigal et al. (2010) observed nutrient levels as K > N > Ca > P > Mg. These variations suggest influence mainly of the crop environment, since the differences among cultivars in the different studies was not significant. In this study, the organic farming techniques with planting in January -off the recommended time for the region -with harvest expected in August, may have influenced the results.
For micronutrient contents, there is no clear definition for differences among cultivars. Fe is the micronutrient with the highest content in all cultivars, which was also reported by Vidigal et al. (2010). Similar to Fe, B levels were Figure 1: Relationship between bulb yield and harvest index (ratio between dry part of harvested bulb and total dry part of the plant -bulb + shoot) of onion cultivars in organic system (R² < 0.01).
The highest concentration of Cu in the bulb observed in most cultivars, except for IPA-11 and Alfa São Francisco. Cu, besides a constituent of several enzymes involved in the processes of photosynthesis, respiration, hormone regulation, N fixing and secondary compound metabolism, also affects the bulb skin color, by intensifying it, and provides increased peel strength and less weight loss during storage (Mendes et al., 2008). The higher Mn content in shoots of onion in relation to bulbs agree with the results reported by Vidigal et al. (2003), but the same behavior observed for Zn was not found by Vidigal et al. (2000), whose results showed a higher concentration of this nutrient in bulbs.
The order of micronutrient levels in the shoot Fe > Zn > Mn = B > Cu in all cultivars differs from that obtained by Vidigal et al. (2010), in which the most absorbed micronutrients by onion were Fe and Mn, being followed by Cu and Zn; B was not evaluated.
Fe, followed by Zn and B, is the most absorbed micronutrient by onion, also present in greater quantities in the bulbs (Table 3). Proportionally, Zn is the most exported nutrient with bulb harvest (68% of the total absorbed), which should be considered in fertilizer replacement of this nutrient, especially in successive crops of onion.

CONCLUSIONS
The most productive cultivars for off-season planting in organic system in the metropolitan region of Curitiba were Alfa São Francisco-RT, Alfa São Francisco and Alfa Tropical.
Yield of onion bulbs has no relation to yield of shoots.
Macronutrient contents in both onion shoots and bulbs was in the following order: K > N > P > Ca > Mg.
Fe was the most absorbed micronutrient by onion, followed by Zn > B > Mn > Cu.
The amount of nutrients accumulated in the shoots was higher than in bulbs, and the most productive cultivars were the most efficient ones, since they provided greater bulb yield per each kg of nutrient extracted.