Accumulation of N, P, and K in the tubers of potato (Solanum tuberosum L. ssp. andigena) under contrasting soils of the Andean region of Colombia Received for publication: 16 November, 2016. Accepted for publication: 15 March, 2017. Doi: 10.15446/agron.colomb.v35n1.61068 1 Department of Agronomy, Faculty of Agricultural Sciences, Universidad Nacional de Colombia. Bogota (Colombia). igomezm@unal.edu.co Agronomía Colombiana 35(1), 59-67, 2017 Accumulation of N, P, and K in the tubers of potato (Solanum tuberosum L. ssp. andigena) under contrasting soils of the Andean region of Colombia Acumulación de N, P y K en tubérculos de papa (Solanum tuberosum L. ssp. andigena) bajo suelos contrastantes en zona andina de Colombia Manuel Iván Goméz1, Stanislav Magnitskiy1, Luis Ernesto Rodríguez1, and Aquiles E. Darghan1 ABSTRACT RESUMEN The relationship between tuber growth and demand for NPK in andigena group of potato (Solanum tuberosum L. ssp. andigena) is poorly documented under conditions of the An- dean region of Colombia. It is necessary to establish a specific nutrient management for high yields to changes in edaphic- environmental supply and improve the production of Capiro and Supreme cultivars. Twelve treatments were evaluated at different stages (75-100-125-150 days after planting) of tuber growth using a repeated measures design with three factors: two cultivars (Capiro, Suprema); three locations with contrasting soils (Subachoque, Facatativá, and Choconta) and two levels of fertilization, F0 (unfertilized) and F1 (fertilized) of incomplete nature by differential fertilization by soil type. A positive cor- relation between fresh weight, dry weight and extraction of N, P and K (kg ha-1) by the tubers, beside harvest extraction index (HEI) was determined. Under optimal conditions of fertilization significant differences (P<0.001) between factor interactions cultivar x phenology x location for accumulation of N and P were detected, where Suprema was less demanding in the HEI of N (1,92 kg t-1 harvest) and HEI of P (0.38 kg t-1 harvest) than Capiro and was better adapted to acid soils of low fertility, in contrast the HEI of K Between 5.28 and 5.34 kg t-1 harvest did not show differences between cultivars due to the genotypic characteristics in the accumulation of dry biomass and starch that make them suitable for industrial use, in addi- tion it was verified that the nutritional extraction depends on the genetic potential determined by the interaction with the environmental and edaphic supply. La relación entre el crecimiento del tubérculo y la acumu- lación NPK en cultivares de papa (Solanum tuberosum L. ssp. andigena) está poco documentada bajo condiciones de la zona andina colombiana y es necesaria para establecer un manejo específico de la fertilización en la obtención de altos rendimien- tos de los cultivares Capiro y Suprema ante las variaciones de oferta edáfico-ambiental. Se evaluaron 12 tratamientos en diferentes etapas el crecimiento del tubérculo (75-100-125-150 días después de siembra) mediante un diseño en medidas repe- tidas con tres factores: dos cultivares (Capiro, Suprema); tres localidades con suelos contrastantes (Subachoque, Facatativá y Chocontá) y dos niveles de fertilización, F0 (no fertilizado) y F1 (fertilizado) de naturaleza incompleta por la fertilización diferencial realizada por tipo de suelo. Se presentó correlación positiva entre el peso fresco, peso seco y extracción N, P y K (kg ha-1) del tubérculo además se determinó el índice de extracción de cosecha (IEC). Bajo condiciones óptimas de fertilización se presentaron diferencias significativas (P<0,001) entre las inte- racciones de los factores cultivar x fenología x localidad para la acumulación de N y P, donde Suprema fue menos exigente en el IEC de N (1,92 kg t-1 cosecha) y el IEC de P (0,43 kg t-1 cosecha) que Capiro, además se adaptó mejor a suelos ácidos de baja fertilidad, en contraste el IEC de K entre 5,28 y 5,34 kg t-1 cosecha no presentó diferencias entre cultivares debido a las características genotípicas en la acumulación de biomasa seca y almidón que los hacen adecuados para uso industrial, además se comprobó que la extracción nutricional depende del potencial genético determinado por la interacción con la oferta edáfico-ambiental. Key words: nutrient extraction, nutritional requirements, macronutrients, Solanaceae. Palabras clave: ex tracción nutriciona l, requerimientos nutricionales, macronutrientes, Solanaceae. Kumar et al., 2013). Potato stays among the crops of the highest demand for mineral nutrients per kg of dry matter produced (Sierra et al., 2002; White et al., 2009; Kumar et al., 2013). The andigena potatoes are well adapted to soils of low fertility, but respond favorably to NPK fertiliza- tion achieving high yields (Ríos Quinchoa et al., 2010; Hernández et al., 2012). According to Grandett and Lora Introduction The growth of potato plants depends on the environ- mental conditions and genotype-environment interac- tion (Corchuelo, 2005; Cabezas, 2013), where absorption, translocation, and accumulation of essential nutrients in the tuber allow starch accumulation (Corchuelo, 2005; http://http://dx.doi.org/10.15446/agron.colomb.v35n1.61068 60 Agron. Colomb. 35(1) 2017 (1979), Villamil (2005), and Gómez and Torres (2012), the macronutrients of the highest requirement in andigena potato plants are K, N, and P followed by Mg, S, and Ca. In potato, N is the element essential for protein synthesis, respiration, and growth of tubers (Westermann, 2005; Kav- vadias et al., 2012). In soil conditions of mountain tropical regions, N availability to the plants could be affected by low mineralization rate, low temperatures, clay soils, and low organic matter content (Castro and Gómez, 2013), al- though the excess of N may result in luxury consumption and reduce tuberization (Ruza et al., 2013). Deficiency of N caused a reduction in dry matter, reduced leaf area and fewer leaf lets that provides less light interception and a lower rate of photosynthesis (Balemi, 2009; Marouani et al., 2015; Li et al., 2016). Phosphorus promotes root growth (Aguilar-Acuña et al., 2006), rapid formation of tubers and starch synthesis (Perrenoud, 1993). In strongly acid soils with pH lower when 5.5 of the Andean regions with culti- vated potato, P is fixed and forms precipitates with iron and aluminum, which decreases its availability (Hernández et al., 2012; Castro and Gómez, 2013). Potassium is essential for translocation of sugars to the tubers and starch syn- thesis, a fundamental processes in the tuber growth and filling (Reis and Monnerat, 2000). Low soil supply of K in sandy soils, excess of Ca (Kavvadias et al, 2012), salinity and clay minerals of 2:1 type (Castro and Gómez, 2013) limit availability of K to potato crop. The relationship between tuber growth and demand for macronutrients NPK in andigena potatoes is scarcely documented, especially as affected by the changes in edaphic-environmental conditions of Colombian Andean region. This relationship is important to establish a spe- cific nutrition management and efficient use of nutrients, and optimize fertilization moments in order to improve productivity and sustainability of cultivars Supreme and Capiro as economically important for fresh consumption and industrial use. The aim of the study was to evaluate the accumulation of N, P and K in two cultivars spp andigena, under different stages of tuber growth in contrasting soils of the Cundiboyacense plateau. The second objective of the research was to analyze the effect of fertilization on the growth and yield of tuber in order to establish an efficient and specific management of nutrition as dependent on soil type and plant phenology. Materials and methods The study was done in 2013 and 2014 in soils contrasting in the level of fertility in the Andean region of Colombia in representative locations in Cundinamarca province char- acterized with high yield potential (> 50 t ha-1) of potato and different environmental and soil fertility conditions (Tab. 1). Twelve treatments were evaluated using a repeated mea- sures design with three factors: two cultivars ssp. andigena (Capiro, Supreme), three locations (Subachoque, Facatativa and Choconta), and two levels of fertilization, F0 (unfertil- ized) and F1 (fertilized). The experimental design was of incomplete nature by differential fertilization by soil type (Tab. 2), with a within-subjects factor associated with four phenological critical stages tuber growth adapted from Segura et al. (2006). These stages were stage II, 70-75 days after planting (dap) (formation of secondary stems and tuberization initiation); stage III, 90-100 dap (bloom, maximum tuberization, and beginning of tuber filling); stage IV, 120-125 dap (final f lowering, tuber filling); stage V, 150-160 dap (senescence, maximum filling, and matura- tion of tubers). In each locality, divided plot design with three replicates was used, where the main plot corresponded to the cultivars and subplots corresponded to the two levels of fertilization. The contribution of mineral nutrients, fertil- izer sources and fractionation of fertilization (F1) levels in each location were done according to table 1. Planting was done in experimental units of 50 m2 (135 plants/ unit) with a row spacing of 1 m and 0.37 m between the plants, and useful harvest area of 36 m2 with a density of 27,000 plants/ha. At each stage of tuber growth four plants were taken from each experimental unit and a destructive analysis of tubers and aerial part was carried out. At each stage, fresh weight of tubers (FWT) (kg ha-1) was measured. For dry weight of tubers (DWT), 200 g of fresh plant mate- rial were dried to constant weight in oven at 70°C for 72 h and the concentrations of N, P, and K was determined in tubers (IGAC, 2006). The methods used for elemental analysis were Micro-Kjeldahl volumetric for total nitrogen (%), calcination at 475°C colorimetric with molybdate and ammonium vanadate for total phosphorus (g kg-1), calcination at 475°C for K (g kg-1). Using the accumulation of macronutrients per dry weight values the nutritional extraction (kg ha-1) and nutrient (g kg-1) were quantified per growth stage of tubers adapted from Cabalceta et al. (2005) and Gómez and Torres (2012). With the above mentioned variables harvest index (HI) and the rate of nu- trient extraction at harvest (EI) (kg of nutrient extracted per ton of harvested tuber) were evaluated. 61Goméz, Magnitskiy, Rodríguez, and Darghan: Accumulation of N, P, and K in the tubers of potato (Solanum tuberosum L. ssp. andigena) under contrasting soils... TABLE 1. Edaphoclimatic conditions in the locations of the study. Climatic conditions1 Subachoque Facatativa Choconta Altitude (m a.s.l.) 2,680 2,520 2,780 North Latitude 4°57´50.1” 4°49´26.9 ” 5°5́ 30.37” West Longitude 74°09’28.1” 74°22’29.7” 73°43´2.04” Annual precipitation (mm) 870 951 1,295 Max. temperature (ºC) 18.7 18.1 16.9 Min. temperature (ºC) 6.3 7.0 4.4 Average temperature (ºC) 12.5 12.6 10.6 Soil characteristics2 Typic Hapludand Andic Eutrudept Humic Dystrudept Soil fertility3 Low High Low pH 5.16 6.4 5.5 Al (cmolc kg -1) 0.59 0.0 0.1 MO (g kg-1) 171 166.7 67.7 CEC (cmolc kg -1) 8.31 31.95 9.52 Texture Loamy Loamy Clay loamy N (g kg-1) 8.5 8.3 3.3 P (mg kg-1) 24.8 39.64 18.18 K (cmolc kg -1) 0.1 3.14 0.68 Ca (cmolc kg -1) 6.11 24.26 7.20 Mg (cmolc kg -1) 1.29 4.34 1.57 S (mg kg-1) 23.0 29.53 11.58 Saturation K (g 100 g-1) 1.2 9.82 7.14 Saturation Ca (g 100 g-1) 75 77 75 Saturation Mg (g 100 g-1) 15 13.17 16.4 Ca/K 62.52 7.72 10.51 (Ca+Mg)/K 71.1 9.1 12.8 1 Environmental data were taken and calculated from IDEAM (2012-2013). 2 Physico-chemical characterization of soils in arable layer (0-30 cm) was done according to methodology IGAC (2006). The soils were classified according to USDA classification system (Soil Survey Staff, 2010). 3 Chemical fertility potential was assessed in the topsoil 0-30 cm (Castro and Gómez, 2013). TABLE 2. Doses of mineral nutrients applied with fertilization in the stu- died locations. Nutrient1 (kg ha-1) Subachoque (F1s) Facatativa (F1f) Choconta (F1ch) N 198 171 192 P2O5 374 261 340 K2O 380 180 348 CaO 40 110 45 Mg 55 70 56 S 37 74 80 B 2,8 3,4 1,2 Zn 5,6 5,6 2,4 Mn 7,0 7,0 3,0 Cu 1,4 1,4 0,6 Fe 2,8 2,8 1,2 1 Fertilizers applied according to the method of soil-plant balance (Castro and Gómez, 2013) and fractioned according to historical management in areas, where the high yields have been obtained (>50 t ha-1): N, 60% sowing, 40% dap 45-50; P, 70% sowing, 30% 45-50 dap; K, 30% sowing and 70% 45-50 dap. The sources of granular fertilizers were: N-P, DAP; K, KCl (0-0-60) potassium sulfate (0-0-50); Ca, calcium nitrate (25% CaO); Mg, Kieserite; Nutricomplet®, a complex micronutrient source B, Zn, Cu, Mn and Fe, based on sulfates. Principal component analysis was used to perform an exploratory analysis and reduce the dimensionality of the set of variables. Two components were analyzed with a cumulative variability highly significant 87.55% as a result of statistical analysis; the weights of the components were used as new variables for which the multivariate analysis of variance, MANOVA (Ravindra and Dayanand, 1999) was performed. The Pearson correlation matrix of qualitative variables was analyzed and the regression curves were fit- ted; additionally, the graphs of the interaction of factors that describe the behavior and distribution of nutrients were compared. SAS statistical software version 2014 and Stat graphics version 2010 were used for the graphics. Results and discussion Accumulation of N, P and K in tuber phenology At the beginning of tuberization, 75 dap, Supreme took up 51 kg N ha-1, and 100 kg K ha-1, while Capiro took up about 30 kg N ha-1 and 84 kg K ha-1, with similar uptakes of P 62 Agron. Colomb. 35(1) 2017 (Fig. 1) in Humic Dystrudepts (Choconta) with increased extraction and higher yields than other soil evaluated. These significant differences between the cultivars for N y K (P<0.05) could be attributed to early and stepped tuberization in Supreme and to highly bearing phenotype of this cultivar. These plant features, apparently, promoted higher demand for assimilate f low into the tubers at initial growth stages in soils of low fertility, such as of Choconta. Suprema presented better uptake in soils of low fertility Humic Dystrudepts and Typic Hapludands (Fig. 1) with reduced availability of N, P, and K (Tab. 1), while accumulation of Capiro depended directly on increased availability of P in the soil, where the highest uptake of P was obtained in Andic Eutrudepts slightly acidic soil (Fig. 1, Tab. 1). Therefore, it should be performed earlier and gradual fertilization management of Supreme as related to the fer- tilization of Capiro that presented accumulation of NPK 75 dap after the beginning of tuberization (Fig. 1B, D and F). After this growth stage, Capiro y Suprema increased FIGURE 1. Accumulation of NPK in tubers during the growing season in andigena potato, cv. Supreme (left) and cv. Capiro (right) in Typic Hapludands (Subachoque, u); Humic Dystrudepts (Choconta, n) and Andic Eutrudepts (Facatativa, p) with fertilization (fer1) in the altiplano Cundiboyacense. dap (day after planting). Error bars indicate standard error. U pt ak e K ( kg h a- 1 ) F dap, Capiro U pt ak e K ( kg h a- 1 ) E dap, Suprema U pt ak e P ( kg h a- 1 ) D dap, Capiro U pt ak e P ( kg h a- 1 ) C dap, Suprema U pt ak e N ( kg h a- 1 ) B dap, Capiro U pt ak e N ( kg h a- 1 ) A Subachoque Choconta Facatativa 0 20 40 60 80 100 120 140 160 180 200 0 25 50 75 100 125 150 175 dap, Suprema 0 25 50 75 100 125 150 175 0 50 100 150 200 250 0 25 50 75 100 125 150 175 0 25 50 75 100 125 150 175 0 5 10 15 20 25 30 35 0 25 50 75 100 125 150 175 0 25 50 75 100 125 150 175 0 5 10 15 20 25 30 35 40 45 50 0 50 100 150 200 250 300 350 400 450 500 0 50 100 150 200 250 300 350 400 450 63Goméz, Magnitskiy, Rodríguez, and Darghan: Accumulation of N, P, and K in the tubers of potato (Solanum tuberosum L. ssp. andigena) under contrasting soils... the demand of assimilates that coincided with increased nutrient transport to tubers and a localized tuberization at this phenological stage (Figure 1). This is consistent with reports Villamil (2005) for Capiro and comparable with the data found by Sierra et al. (2002) for the cultivar Desiree in Chile and White et al. (2009) for spp tuberosum; where the authors reported increase in NPK accumulation after the beginning of the tuberization. The uptake should be synchronized with the application of NPK fertilizer to promote initial growth associated with the growth of stolons, leaves and stems, increased number of tubers and protein accumulation. In this sense, it is also important to consider factors, such as hormonal balance, low tempera- tures and photosynthetically active radiation (Ruza et al., 2013; Marouani et al., 2015). In both cultivars, the maximum extraction of N, P, and K by the tubers presented at the tuber filling stage 150-160 dap, except cv. Capiro in Subachoque that presented a maximum extraction at 125 dap (Fig. 1B, D and F). Com- pared with Facatativa and Choconta, Subachoque likely had more adverse conditions in the availability of N and K (Tab. 1) and water deficit in this locality, which acceler- ated ripening and affected proportionately losses in yield potential because tuberization of Capiro was grouped into more defined stages 70-100 dap, with greater emphasis on nutrient availability at flowering stages. Therefore, it should be avoided stress affecting organs and metabolism sources for filling of tubers towards the end of the cycle, such as nutritional and hydric deficits, phenomena that have been reported by Monneveux et al. (2013). Equally, cv. Ica Purace presented the earliest maximum accumulation of nutrients at 120 dap (Grandett and Lora, 1979), cv. Pimpernel at 130 dap (Sierra et al., 2002) regarding accumulation at the end of the cycle for cv. Desiree (Sierra et al., 2002) and cv. Atlantic (Coraspe-León et al., 2009). The differences in nutrient accumulation were directly related to soil fertility and tuberization dynamics per cultivar, whether determi- nate or indeterminate, so it is important to consider these conditions in fertilizer management. The P accumulation had significant differences (P≤0.05) between localities and cultivars mainly for Capiro, prob- ably because of its higher nutrient requirement and lower extraction on soils poor in P (Typic Hapludands and Humic Dystrudepts) (Fig. 1D). Regarding Supreme, it presented a better response in these soils of low fertility and without differences in extraction in studied soils. Thus, Supreme showed a better adaptation of this cultivar to soils low in P and a lower requirement of P applied with fertilization and probably for its further exploration at the root and mechanisms of adaptation to acid soils by exudates of root organic acids that improve the availability of P fixed. These mechanisms of differential absorption have not yet been investigated for these cultivars. Phosphorus was the mineral nutrient of less accumulation in tubers as compared to N and K in evaluated cultivars characterized with 17-40 kg ha-1 uptake P (Fig. 1C and D), which coincided with local study by Villamil (2005) for Capiro. The importance of P should not be underestimated, because its deficiency in potato crops dramatically reduced the size of tubers, yields (Aguilar-Acuña et al., 2006; Barben et al., 2007; Hernandez et al., 2012) and tuber quality (Fer- nandes et al., 2015). Phosphorus deficiency slowed down the apical growth, resulting in small plants and reduced formation of starch in tubers, which manifests itself with necrotic spots distributed in the tubers (Hernández et al., 2012). Several reports indicated that potato plants growing in soils with low P develop less leaf area, which affected light interception, reduced plant growth and differentiation, and generated less stolon bearing tubers (Barben et al., 2007). Results obtained by Balemi (2009) in three genotypes of ssp. tuberosum showed that a low supply of P affected dramati- cally the plant growth and production of dry matter, also the importance of P in starch accumulation in potato was comparable with that of K (Perrenoud, 1993). Accumulation of K in tubers of Capiro was similar in three locations, with a maximum extraction of 373 kg ha-1 in soils matching high availability of this element in Choconta and Facatativa (soil saturations with K >7%). Supreme presented the maximum extraction of 429 kg ha-1 in Choconta fol- lowed by Subachoque (335 kg ha-1) and Facatativa (232 kg ha-1); this cultivar evidenced a differential extraction (P≤0.05) curves according to location and that were similar to those of N (Fig. 1A and E). The most uptake rate of K by the tubers of this cultivar was more associated with adequate environmental supply with lower temperatures <12°C and greater efficiency in the use of K y N in these soils that allows adequate sink- source management and potassium fertilization in soils of low fertility balance. In soils of high fertility, such as Andic Eutrudepts, potato plants could present luxury consump- tion of N y K, as it was evidenced in Facatativa, where a similarly extraction was obtained but with greater accu- mulation of N and K in the aerial parts and similar yields in Choconta, suggesting lower doses of these elements in these soil conditions. In addition, the higher average air temperature in Facatativa over Choconta may cause a higher photosynthetic rate that matches high growth of the 64 Agron. Colomb. 35(1) 2017 aerial part and could encourage higher activity of enzymes related to assimilation of K and, therefore, a high demand for this element. Importantly, Capiro in soils of low fertility of Subachoque with low concentrations of native soil K (0.1 cmolc kg-1) (Fig. 1F) presented an earlier accumulation of K in the tubers (125 dap) unlike 150 d in other locations (Fig. 1F). These data contrasted in this locality with an accumulation of K at the end of the cycle (150- 160 dds) for Supreme, where tuberization and tuber filling was evident by persistent vegetative development and tuber growth. This implicates a need for more precise handling of supplementary fertiliza- tion in soils of low offer of K. It could be suggested after 100 dap to keep photosynthetically active and “green” plant until the end of the cycle avoiding early ripening and to make liquid or foliar fertilization at the late stages of the crop. These technologies have been evaluated by Horvat et al. (2014) as an alternative in the supplementary fertiliza- tion management of this crop. The steps of tuber filling and accumulation of K after 100 dap were related directly to the partition of the dry mat- ter in tubers and were higher than the rest of the plant, which coincided with high harvest index > 75%, which favors high yield potential. Similar rates of harvest were found by Sierra et al. (2002) and Sauceda Acosta (2010) for ssp. tuberosum cv. Desireé with yields close to 90 t ha-1 and Santos et al. (2010) for cv. Capiro and cv. Suprema with yields close to 50 t ha-1. In soils Andic Eutrudepts of Facatativa, the harvest indexes (HI) for Supreme were lower than 55% and lower than those obtained in low fertility soils with HI >75%, which corroborates a low sink-source capacity in soils of high fertility associated with high availability of N and K. Accumulation of N, P, and K among cultivars The demand for N, P, and K in the cultivars was directly related to the accumulation of FWT and DWT. A se- quence of accumulation of mineral nutrients K> N> P in both cultivars was observed, similar to that reported by Westermann (2005); Coraspe-León et al. (2009); Subra- manian et al. (2011) and Fallas and Bertsch (2014) in ssp. tuberosum, and by Grandett and Lora (1979), Villamil (2005); Gomez and Torres (2012) and Lefèvre et al. (2012) in andigena potato. The significant relationship between nutrient extraction and FWT allowed estimating the harvest extraction index (HEI) in kg of nutrient extracted per ton of harvestable organ (Fig. 2), and is important to determine nutrient requirements in fertilizer recommendations according to performance targets locals as reported Ciampitti and García (2008) and Fallas and Bertsch (2014). HEI of N and P showed significant differences among cultivars with higher yields at 50 t ha-1 (P≤0.05) and HEI higher for Su- preme than for Capiro in a positive linear model (Fig. 2A and B). For Supreme HEI was 1.92 kg N t-1, 0.38 kg P t-1; 5.3 kg K t-1, while for Capiro HEI was 2.27 kg N t-1, 0.47 kg P t-1; 5.3 kg K t-1, when compared with results reported by Sierra et al. (2002) and Ciampitti and Garcia (2008), when in po- tato ssp. tuberosum HEI was 3.5 kg N t-1, 0.7 kg P t-1; 5.4 kg K t-1 with greater differences for N and P and equal for K. Capiro required 32% more of N and 48% more of P for the yields superior than 50 t ha-1 more than Supreme with significant differences in beta coefficient according to a positive lineal model in both cultivars (Fig. 2A and B). These variations in N and P uptake in Capiro can be explained by a genotype of low stature and tuberization more synchronized with better adaptation to fertile soils and highest genetic potential; in addition this was also as- sociated with accumulation of organic and mineral forms that determine the nutritional characteristics of the tubers. This confirms the differences in accumulation of NPK between potato genotypes reported by White et al. (2009), Karam et al. (2009) and Wekesa et al. (2014). It coincides with the differences in accumulation of these nutrients in cultivars ssp. tuberosum made by Tekalign and Hammes (2005) and White et al. (2009) in identical environments and andigena cultivars in the Bogota Plateau made by Grandett and Lora (1979). Significant differences (P≤0.05) in extracting P in favor of cultivar Capiro regarding Supreme (Fig. 1B and 2B) can be explained by the most concentrated tuberization, the better translocation of assimilates and increased demand of sinks than that in Supreme. This features in Capiro facilitate uniform ripening with a defined harvest, which requires a greater demand for ATP and may be associated with organic and inorganic forms of P. These forms of P could serve as a reserve and will explain responses in the dormancy period of the tubers that lasts up to 4 months in this cultivar and is longer than that in Supreme that presented an earlier period of up to two months. The differences in P suggest a higher concentration of phytate complex for Capiro than that in Suprema, which should be evaluated in future studies to test this hypothesis. In relation to P in the tubers Subramanian et al. (2011) mentioned that this element was metabolized from organic bonds orthophosphate-ester in starch that accumulate early 65Goméz, Magnitskiy, Rodríguez, and Darghan: Accumulation of N, P, and K in the tubers of potato (Solanum tuberosum L. ssp. andigena) under contrasting soils... HEI for K showed no significant differences between the cultivars and this was the element that had the highest correlation (r2 >0.95) with both yield and DWT compared to N and P (Figure 2 c). According to the results for K, HEI presented independent effects between cultivar, locality and fertilization, suggesting that the K extraction in these cultivars of ssp. andigena was mainly depended on the genetic potential of the cultivars. The later corroborate the direct effect of K in transport, storage, and conversion of carbohydrates in ssp. andigena, therefore, the importance of studied cultivars for the process of industrial transforma- tion whit high dry matter in tuber that was corroborated in this study. The K compared to N and P is the nutrient that had the highest correlation in Capiro (r2 0.95) and Supreme (r2 0.99) respect to DWT. The relationship between K and tuber growth matched the one reported by Coraspe-León et al. (2009), Karam et al. (2009), Sarkar et al. (2010), Kav- vadias et al. (2012) for potato ssp. tuberosum and Villamil (2005) and Pérez (2015) for potato ssp. andigena. The ac- cumulation of K was directly related to accumulation of dry matter (% DWT) and starch in the tubers (Perrenoud, 1993; Westerman et al., 1994; Reis Jr and Monnerat, 2000) of 19-28% DWT for ssp. andigena (Jiménez et al., 2009). DWT % evaluated in this study (20-27%) was similar to those reported by Jiménez et al. (2009) and Ñústez (2011). Conclusions It is necessary a differential management of mineral nutri- tion in cultivars according to their phenology and soil type and considering the differences in nutritional extractions with higher yields to 40 t ha-1 in both cultivar. It is sug- gested the management Supreme fertilization in soils of low fertility applying 30% of total extraction NPK before stage II, beginning of tuberization (75 dap) in a uptake ratio N:P:K of 10-2-12 N as related to Capiro. Capiro requires about 15% of the total extraction NPK in the uptake ratio N: P: K 10 2-18; which shows a higher uptake ratio of K in early stages, so its best response was in soils Humic Dys- trudepts of Chocontá and Andic Eutrudepts of Facatativá where saturation K soil was >7%. For tuber filling stages, before stage III (maximum tuberization and start filling), 70% (Supreme) and 85% (Capiro) of the total NPK required by the tuber with an 8-1-20 uptake ratio for Supreme and 12-2-25 uptake ratio for Capiro should be applied. These recommendations can improve the physiological efficiency of these nutrients by taking into account the timing be- tween doses, ratio, and age. Actually, it is fractionated in two applications before 50 dap and is not applied according to the nutrient extraction. FIGURE 2. Relationship between the uptake in tubers of N (A), P (B) and K (C) (kg ha-1) and yield (FWT) in cv. Capiro and cv. Suprema in soils of Cundiboyacense plateau. ** The value of coefficients b were significant at the 0.05 probability level. ns, not significant the value of coefficients b. y=0,0023x-10,99 R2=0,9458** y=0,002x+0,9887 R2=0,80512** y=0,0053x+2,9223 R2=0,95218 ns y=0,0052x+5,6299 R2=0,9612 ns y=0,0005x - 0,6094, R2=0,84685 ** y=0,0004x+2,7127 R2=0,85868** K U pt ak e (k g ha -1 ) C FWT (1,000 kg ha-1) P U pt ak e (k g ha -1 ) B FWT (1,000 kg ha-1) N U pt ak e (k g ha -1 ) A P-Capiro P-Suprema K-Capiro K-Suprema FWT (1,000 kg ha-1) 0 50 100 150 200 250 0 10 20 30 40 50 60 70 80 90 N-Capiro N-Suprema 0 10 20 30 40 50 60 70 80 90 0 5 10 15 20 25 30 35 40 45 50 0 10 20 30 40 50 60 70 80 90 0 50 100 150 200 250 300 350 400 450 500 in tuber formation and from mineral forms of complexes of phytic acid (myo-inositol hexakisphosphate) towards maturation of tuber and sprouting and is necessary for proper formation of seeding tubers. The latter complex may reach between 0.1 and 0.27% DWT with differences between cultivars ssp. tuberosum according to Phillippy et al. (2004). 66 Agron. Colomb. 35(1) 2017 Acknowledgements The authors express their gratitude to Ingeplant SAS, Fedepapa and Unal for funding and technical support of the agronomists of Ingeplant SAS Paola Torres, Liliana Arévalo, Elías Silva, and Andrea Barragán. We also thank growers Walter Guzmán - Biogenética, Yovanny Pulido, Ricardo Rojas and Carlos Acero for their support in the development of research. Literature cited Aguilar-Acuña, J.L., J. Martínez-Hernández, V. Volke-Haller, J. Etchevers-Barra, H. Mata-Vázquez, and M. Hernández- Martínez. 2006. Distribución del fósforo en suelo, raíces y materia seca de tubérculos de papa cultivada con fertirriego. Terra Latinoam. 24(2), 269-276. Balemi, T. 2009. Effect of phosphorus nutrition on growth of potato genotypes with contrasting phosphorus efficiency. Afr. Crop Sci. J. 17(4), 199-212. Barben, S., B. Nichols, B. Hopkins, V. Jolley, J. Ellsworth, and B. Webb. 2007. Phosphorus and zinc interactions in potato. pp. 219-223. In: Western nutrient management conference. Salt Lake City, UT, USA. Cabalceta, G., M. Saldias, and A. Alvarado. 2005. Absorción de nutri- mentos en el cultivar de papa MNF-80. Agron. Costarr. 29(3). Cabezas, M. 2013. Fisiología de la maduración del tubérculo de la papa. Fedepapa, Bogota, Colombia. Castro, H. and M.I Gómez. 2013. Fertilidad y fertilizantes. pp. 231-304. En: Ciencia del suelo - principios básicos. Sociedad Colombiana de la Ciencia del Suelo. 2ª ed. Bogota, Colombia. Ciampitti I. A. and O. García. 2008. Requerimientos nutriciona- les: absorción y extracción de macro nutrientes y nutrientes secundarios II. Hortalizas, frutales y forrajeras. Archivo Agronómico No 12. International Plant Nutrition Institute (IPNI), Buenos Aires, Argentina. Coraspe-León, H., T. Muraoka, V. Franzini, M. De Stefano, and N. Granja. 2009. Absorción de macronutrientes por plantas de papa (Solanum tuberosum L.) en la producción de tubérculo- semilla. Interciencia 34(1), 57-63. Corchuelo, G. 2005. Ecofisiología de la papa. pp. 13-17. In: Memorias I Taller nacional sobre suelos, fisiología y nutrición vegetal en el cultivo de papa. CEVIPAPA, Consejo Nacional de la Papa. Bogota, Colombia. Fallas, R. and F. Bertsch. 2014. Análisis del estado nutrimental del cultivo de la papa en Costa Rica con base en información existente. Agron. Costarr. 38(1), 199-206. Fernandes, A.M., R.P. Soratto L.A Moreno, and R.M. Evange- lista. 2015. Effect of phosphorus nutrition on quality of fresh tuber of potato cultivars. Bragantia 74, 102-109. Doi: 10.1590/1678-4499.0330 Gómez, M.I. and P. Torres. 2012. Absorción, extracción y manejo nu- tricional del cultivo de papa. Revista Papa Fedepapa 26, 20-25. Grandett, G. and R. Lora. 1979. Acumulación de materia seca y de varios nutrimentos en tres variedades de papa cultivadas en la serie Mosquera del Centro Experimental Tibaitata. pp. 98-102. En: Marín M. (ed.). Programa Nacional de Suelos. Informe de Progreso 1978. Mosquera. Instituto Colombiano Agropecuario (ICA), Mosquera, Colombia. Hernández, H.M., F. Forero, D. Otálora-Antolines, and P.A. Serrano- Cely. 2012. Respuesta agroeconómica del cultivo de papa (Solanum tuberosum L.) bajo diferentes fuentes de fósforo en Villapinzon, Cundinamarca. Cienc. Agric. 9(2), 97-104. Horvat, T., M. Poljak, B. Lazarevic, Z. Svecnjak, and K. Hanacek. 2014. Effect of foliar fertilizers on physiological characteristics of potato. Roman. Agric. Res. 31(1), 159-165. Jiménez, M., A. Rossi, and N. Sammán. 2009. Phenotypic, agronomic and nutritional characteristics of seven varieties of Andean potatoes. J. Food Comp. Anal. 22, 613-616. Doi: 10.1016/j. jfca.2008.08.004 Karam, F., Y. Rouphael, R. Lahoud, J. Breidi, and G. Colla. 2009. In- f luence of genotypes and potassium application rates on yield and potassium use efficiency of potato. J. Agron. 8(1), 27-32. Kavvadias, V., C. Paschalidis, G. Akrivos, and D. Petropoulos. 2012. Nitrogen and potassium fertilization responses of potato (Sola- num tuberosum) cv. Spunta. Com. Soil Sci. Plant Anal. 43(1-2), 176-189. Doi: 10.1080/00103624.2012.634711 Kumar, C. V., S.S. Prakash, G.M. Prashantha, M.B.M. Kumar, S. Lohith, and T. Chikkaramappa. 2013. Dry matter production and yield of potato as inf luenced by different sources and time of fertilizer application and soil chemical properties under rainfed conditions. Res. J. Agric. Sci. 4, 155-159. Lefèvre, I., J. Ziebel, C. Guignard, J.F. Hausman, R.O. Gutiérrez Rosales, M. Bonierbale, and D. Evers. 2012. Drought impacts mineral contents in Andean potato cultivars. J. Agron. Crop Sci. 198(3), 196-206. Doi: 10.1111/j.1439-037X.2011.00499.x Li, W., B. Xiong, S. Wang, X. Deng, L. Yin, and H. Li. 2016. Regula- tion effects of water and nitrogen on the source-sink relation- ship in potato during the tuber bulking stage. PloS one 11(1), e0146877. Doi: 10.1371/journal.pone.0146877 Marouani, A., O. Behi, H.B.H. Salah, and O.A. Quilez, 2015. Estab- lishment of chlorophyll meter measurements to manage crop nitrogen status in potato crop. Com. Soil Sci. Plant Anal. 46(4), 476-489. Doi: 10.1080/00103624.2014.997386 Monneveux, P., D.A Ramírez, and M.T. Pino. 2013. Drought toler- ance in potato (S. tuberosum L.): Can we learn from drought tolerance research in cereals?. Plant Sci. 205, 76-86. Doi: 10.1016/j.plantsci.2013.01.011 Ñústez, C. 2011. Variedades colombianas de papa. Universidad Nacional de Colombia, Bogota. Pérez, J.A. 2015. Efecto de diferentes dosis de N y K sobre el ren- dimiento y fritura en papa (S. tuberosum L.) variedad Pastusa suprema. Trabajo de grado. Facultad de Ciencias Agrarias, Universidad Nacional de Colombia, Bogota, Colombia. Perrenoud, S. 1993. Fertilizing for high yield potato. IPI Bulletin 8. 2nd ed.. International Potash Institute, Basel, Switzerland. Phillippy, B., M. Lin, and B. Rasco. 2004. Analysis of phytate in raw and cooked potatoes. J. Food Comp. Anal. 17, 217-226. Doi: 10.1016/j.jfca.2003.08.001 Ravindra, K. and N.K. Dayanand. 1999. Applied multivariate sta- tistics with SAS® Software. 2nd ed. SAS Institute Inc., Cary, NC, USA. 67Goméz, Magnitskiy, Rodríguez, and Darghan: Accumulation of N, P, and K in the tubers of potato (Solanum tuberosum L. ssp. andigena) under contrasting soils... Reis, Jr, R .d.A. and P. Monnerat. 2000. Nutrient concentra- tions in potato stem, petiole and leaf let in response to potassium fertilizer. Sci. Agric. 57, 251-255. Doi: 10.1590/ S0103-90162000000200009 Ríos Q., J.Y., S.D.C. Jaramillo V., L.H. González S., and J.M. Cotes T. 2010. Determinación del efecto de diferentes niveles de fer- tilización en papa (Solanum tuberosum ssp. andigena) Diacol Capiro en un suelo con propiedades ándicas de Santa Rosa de Osos, Colombia. Rev. Fac. Nac. Agron. Medellin 63(1), 5225-5237. Ruza, A., I. Skrabule, and A. Vaivode. 2013. Inf luence of nitrogen on potato productivity and nutrient use efficiency. Proc. Latvian Acad. Sci. 67(3), 247-253. Santos, C. M., A.M Segura, and L. Ñústez. 2010. Análisis de cre- cimiento y relación fuente-demanda de cuatro variedades de papa (Solanum tuberosum L.) en el municipio de Zipaquirá (Cundinamarca, Colombia). Rev. Fac. Nac. Agron. Medellin 63(1), 5253-5266. Sarkar, D., S.K Pandey, and S. Sharma. 2010. High K+ does not affect potato (Solanum tuberosum L.) tuber induction, but represses its development in vitro. In Vitro Cell Dev. Biol-Plant, 46(6), 569-577. Doi: 10.1007/s11627-010-9307-6 Sauceda A., C.P. 2010. Dosis de nitrógeno y fecha de siembra en la modelación del rendimiento de papa (Solanum tuberosum L.) en el norte de Sinaloa. MSc. thesis. Colegio de Postgraduados, Montecillos,México. Segura, M., M. Santos, and C. Ñústez. 2006. Desarrollo fenológico de cuatro variedades de papa (Solanum tuberosum L.) en el municipio de Zipaquirá (Cundinamarca). Fitotec. Colomb. 6(1), 33-43. Sierra, G., J. Santos, and B. Kalazich. 2002. Manual fertilización del cultivo de papa en la zona Sur de Chile. Instituto de Investiga- ciones Agropecuarias. Santiago, Chile. Soil Survey Staff. 2010. Keys to soil taxonomy. 9th ed. Soil Conser- vation Service. 331. United States Department of Agriculture (USDA), Washington DC, USA. Subramanian, N.K., P.J. White, M.R. Broadley, and G. Ramsay. 2011. The three-dimensional distribution of minerals in potato tubers. Ann. Bot. 107(4), 681-691. Doi: 10.1093/aob/mcr009 Tekalign, T. and P.S. Hammes. 2005. Growth and productivity of potato as inf luenced by cultivar and reproductive growth: II. Growth analysis, tuber yield and quality. Sci. Hortic. 105(1), 29-44. Doi: 10.1016/j.scienta.2005.01.021 Villamil, H. 2005. Fisiología de la nutrición en papa. CEVIPAPA, Consejo Nacional de la Papa. Bogota, Colombia. Wekesa, M., M. Okoth, G. Abong, J. Muthoni, and J. Kabira. 2014. Effect of soil characteristics on potato tuber minerals compo- sition of selected Kenyan varieties. J. Agr. Sci. 6(12), 163-171. Doi: 10.5539/jas.v6n12p163 Westermann, D. 2005. Nutritional requirements of potatoes. Am. J. Pot. Res. 82, 301-307. Doi: 10.1007/BF02871960 White, P.J., J.E Bradshaw, M. Finlay, B. Dale, G. Ramsay, J.P. Ham- mond, and M.R. Broadley. 2009. Relationships between yield and mineral concentrations in potato tubers. HortScience 44, 6-11.