NUTRIENT CONTENT OF BIOMASS COMPONENTS OF HAMLIN SWEET ORANGE TREES

The knowledge of the nutrient distribution in trees is important to establish sound nutrient management programs for citrus production. Six-year-old Hamlin orange trees [Citrus sinensis (L.) Osb.] on Swingle citrumelo [Poncirus trifoliata (L.) Raf. x Citrus paradisi Macfad.] rootstock, grown on a sandy Entisol in Florida were harvested to investigate the macro and micronutrient distributions of biomass components. The biomass of aboveground components of the tree represented the largest proportion of the total. The distribution of the total tree dry weight was: fruit = 30.3%, leaf = 9.7%, twig = 26.1%, trunk = 6.3%, and root = 27.8%. Nutrient concentrations of recent mature leaves were in the adequate to optimal range as suggested by interpretation of leaf analysis in Florida. Concentrations of Ca in older leaves and woody tissues were much greater than those in the other parts of the tree. Concentrations of micronutrients were markedly greater in fibrous root as compared to woody roots. Calcium made up the greatest amount of nutrient in the citrus tree (273.8 g per tree), followed by N and K (234.7 and 181.5 g per tree, respectively). Other macronutrients comprised about 11% of the total nutrient content of trees. The contents of various nutrients in fruits were: N = 1.20, K = 1.54, P = 0.18, Ca = 0.57, Mg = 0.12, S = 0.09, B = 1.63 x 10 Cu = 0.39 x 10, Fe = 2.1 x 10, Mn = 0.38 10, and Zn = 0.40 10 (kg ton). Total contents of N, K, and P in the orchard corresponded to 66.5, 52.0, and 8.3 kg ha , respectively, which were equivalent to the amounts applied annually by fertilization.


INTRODUCTION
There is an increasing need to review information on crop nutrition for adequate establishment of nutrient requirements and fine-tuning rates of fertilizer used in agricultural systems because of the risks of adverse environmental impact of excess nutrient applications.This study, conducted in 1999, was part of an ongoing project on best management practices (BMPs) for citrus growing in Florida.Smith (1966a) and Chapman (1968) discussed data on mineral composition of citrus trees published between the 1930's and the 1960's.These authors reported results of several chemical analyses of important components of citrus tree biomass, which allow a broad understanding of amounts, proportions, and distribution of nutrients in the various compartments of the plant.In Brazil, a comprehensive study on mineral composition of fruit (peel, pulp + juice, and seeds) of sweet oranges, acid lime, grapefruit, tangor, and mandarin was presented by Scientia Agricola, v.60, n.1, p.155-160, Jan./Mar.2003 Bataglia et al. (1977).Later work of Amorós Castañer (1995) and Jackson et al. (1995) also reported data on amounts of nutrients removed per ton of oranges and mandarins.There is an agreement that calcium, nitrogen, and potassium are the dominant constituents of citrus tree biomass.Phosphorus, magnesium, and sulfur represent a smaller proportion (~10%) followed by micronutrients (<1%).However, the proportion of individual nutrients may vary among different cultivars, tree age, and horticultural practices in the orchard.
The removal of nutrient by plant biomass has been used as one of the components to estimate fertilizer requirements for annual grain crops.The same approach, on the basis of nutrient removal of fruits, is valid for establishing nutrient management programs for citrus trees; i.e. fertilizer recommendations for citrus in Brazil rely on fruit yield criteria (Quaggio et al., 1996).However, unlike the annual crops, in the case of citrus a large pool of nutrients present in the structural framework of trees represent a large storage which is carried from year to year, and provides nutrients for fruit production during deficiency of applied nutrients (Kato, 1986; Feigenbaum  et al., 1987; Legaz et al., 1995).
In an attempt to tailor BMPs to specific production conditions, research has also focused on modeling biological processes to assess citrus nutrient requirement and to incorporate information into computer-based decision support expert systems (Bustan et al., 1999).Patterns of nutrient accumulation and removal by the crop are of great interest to modelers.The objective of this study was to determine the amount and distribution of nutrients in different parts of 6-year-old citrus trees.

MATERIAL AND METHODS
Six-year-old Hamlin orange trees [Citrus sinensis (L.) Osb.] on Swingle citrumelo rootstock [Poncirus trifoliata (L.) Raf.x Citrus paradisi Macfad.]planted at 7.6 x 4.6 m spacing on a Candler fine sand (hyperthermic, uncoated Typic Quartzipsament) (National Cooperative  Soil Service, 1990) along the Florida central ridge USA, were used in this study.This soil unit is deep well-drained sand (sand content of 967 g kg -1 ) with no confining horizons, low organic matter content, and low water holding capacity.
The experiment was a randomized design with three replications and a single tree assigned to each experimental plot.Trees were irrigated using under-thetree low volume emitters, using one emitter per tree covering about 7 m 2 area under the tree canopy with a delivery rate of 50 L h -1 .During the course of the experiment, irrigation was scheduled when available soil moisture content depleted by 33% in the top 40 cm depth soil, which represented the effective rooting depth of young citrus trees.Soil moisture content at various depths under the tree canopy was monitored using capacitance probes (Alva & Fares, 1998; Fares & Alva, 1999).The trees received annual application rates (g per tree) of N = 230, P = 35, and K = 200 (Ferguson et al., 1995) as ammonium nitrate, triple superphosphate, and potassium chloride, respectively.The fertilizers were split in three applications.Boron, zinc and manganese were applied by foliar spray using a solution prepared with soluble salts during the spring (Ferguson et al., 1995).
The trees were harvested in December 1999 for dry matter evaluation of different tree components and subsampled for determination of nutrient concentration as described by Mattos Jr. (2000).The aboveground portion of the tree was divided into: (i) summer/fall 1999 leaf flush (<6-month-old); (ii) spring 1999 + older leaves; the latter component was made up mostly by summer/ fall 1998 flushes; (iii) twigs >1.5 cm in diameter; (iv) twigs ≤1.5 cm in diameter; (v) trunk; and (vi) fruit.Two opposing quadrants (NW and SE) of 1.75 × 1.75 m each were marked on the soil surface, and which had the tree trunk in a common vertex.This procedure was taken in order to facilitate root collection, since it was assumed that root distribution was uniform among all soil quadrants.The soil was dug from each quadrant to 45 cm depth and sieved (0.2 cm sieve) to collect the roots, which were separated into: (i) fibrous roots <0.2 cm diameter; (ii) woody roots, 0.2 to 1.0 cm diameter; and (iii) woody roots >1.0 cm diameter.The taproot was also separated from the soil.
Samples from each tree parts placed in sealed plastic bags to prevent water loss, and fresh weights were recorded.These samples were washed, and dried at 65°C for 72 hours and dry weight was recorded.Total dry weight of roots was estimated by multiplying the values obtained for both excavated soil quadrants by 2. A subsample of dry matter from each tree part was ground to pass 40 mesh sieve and concentration of N was determined with an automated Roboprep C/N analyzer (Europa Scientific LTD., England) by dry combustion (Mulvaney, 1993).Concentrations of Ca, Mg, K, Cu, Fe, Mn, Zn were determined by ICP spectrometry, and those of P, S, and B by UV-Vis spectrophotometry, after digestion of tissue with concentrate HNO 3 +HClO 4 (2:1 v/ v ratio) (Bataglia et al., 1983).
Total nutrient content of tree biomass components was calculated by multiplying the dry matter of components by the respective nutrient concentration.An analysis of variance (ANOVA) was used to test the hypothesis that nutrient content of biomass components are equal (P = 0.05) using the GLM procedure of the SAS® system (SAS Institute, 1996).The least significant difference (LSD; P = 0.05) procedure was used for making pairwise comparisons among means of nutrient concentration and nutrient content of tree components.

RESULTS AND DISCUSSION
Characteristics of trees and biomass production are outlined here with data adapted from Mattos Jr.
Scientia Agricola, v.60, n.1, p.155-160, Jan./Mar.2003 (2000).The mean height of the trees was 2.5 ± 0.01 m with a canopy diameter of 2.5 ± 0.14 m.Trunk height was 41.4 ± 3.9 cm, and perimeter was 24.5 ± 1.5 cm, above the bud union.Table 1 shows that the dry matter in the aboveground parts of the tree represented >70% of the total tree biomass.The largest proportion of total tree biomass was that of fruit, which represented 30%.The average fruit yield was 15.04 ± 0.11 ton ha -1 (fresh weight basis).Leaves accounted for about 9.7 and roots accounted for 27.7% of the total dry weight of the tree.Fibrous roots accounted for the greatest proportion in the belowground portion (35.6%) after the taproot (38.4%).Our data are comparable to the values reported for citrus (Nadir, 1973; Dasberg, 1987;  Feigenbaum et al., 1987; Alva et al., 1999), since >70% of the total tree biomass was found on the aboveground portion.Deviations in the partitioning of the dry matter of various tree parts are attributed to differences in combination of scion and rootstock varieties, tree age, type of tree part sampled, fruit load and horticultural practices.Lea-Cox et al. (2001) observed that same age Redblush grapefruit trees on Volkamer lemon were larger than trees on sour orange rootstock, and dry weight distribution of tree parts was affected by N fertilization and soil condition.A competitive effect of fruit load on the dry weight allocation to the spring flush of these trees was also observed.
Concentration of macro and micronutrients in the leaves were within the range of optimal concentrations as suggested by the guidelines for interpretation of leaf analysis based on 4-to-6-month-old spring flush from nonfruiting terminals (Tables 2 and 3) (Hanlon et al.,  1995).The Ca concentration was greater in the older flush (>6-month-old) leaves compared to that in the young flush (<6-month-old).A similar pattern was also observed for micronutrients concentrations.This trend agrees with that reported by Smith (1966b).Moreno &  Garcia-Martínez (1984) described the biochemical process of N mobilization in citrus trees (proteolysis), in which total protein content of old leaves decreased progressively from the beginning of February (spring flush) to June.Retranslocation of nutrients from mature to younger tissue occurs to a lesser extent for Ca and micronutrients compared to that for N, K, P, and Mg (Marschner, 1995).
The concentration of K in the fruit (10.7 g kg -1 ) was within the K concentration range for leaves, i.e., 10.3 to 13.8 g kg -1 (Table 2).This value is also comparable to the range of K concentration in the pulp + juice of fruit of orange varieties (9.0 to 13.4 g kg -1 ) reported by Bataglia  et al. (1983).
Among the macronutrients in the tree, the content of Ca was the greatest (273.8 g per tree), followed by N and K (234.7 and 181.5 g per tree, respectively) (Figure 1).Other macronutrients (P, Mg, and S) comprised about 11% of total nutrient content in the trees.The pattern of nutrient distribution within tree biomass components was distinct, with greater amount of Ca being allocated in the framework structure of trees such as twigs.Nitrogen and K contents were markedly removed by the fruit component, which presented a N to K ratio of 1:1.28, which compares with the value of 1:1.40 reported by Jackson et al. (1995).The total P content was a small portion of nutrients stored in the trees (Figure 1) what is in accordance with data reported for non-bearing Hamlin orange trees in Florida (Alva et  al., 1999).This correspondence occurs because of fruit did not represent a strong sink for absorbed P. On the other hand, the proportion of N in the tree parts will greatly vary with fruit load (Lea-Cox et al., 2001).Macronutrients content of 19-year-old grapefruit trees presented by Chapman (1968) far exceeded our values because of the greater amount of total dry matter of trees (273.15kg).
The content of macronutrients in the citrus orchard, as per area basis, decreased in the order: Ca > N > K > Mg > P > S (Table 5).Total content of N, K, and P in the orchard was equivalent to the amounts applied annually by fertilization.However, this relationship cannot be understood as high fertilizer use efficiency because of assimilation of nutrients occurs from the soiltree system on an annual basis.For example, the 15 N recovery in the whole citrus tree of applied labeled N fertilizers during the spring summed up to 40% (Mattos  Jr., 2000).Table 4 also shows that nutrient export by fruits, on the basis of whole tree nutrient content, agrees with data reported by Nadir (1973) for 20-year-old Valencia orange trees, and which results were N = 20%, K = 40%, and P = 30%.

Table 2 -
Macronutrient concentrations in dry matter from various parts of 6-year-old Hamlin sweet orange trees on Swingle citrumelo rootstock.

Table 3 -
Micronutrient concentrations in dry matter from various parts of 6-year-old Hamlin sweet orange trees on Swingle citrumelo rootstock.†

Table 4 -
Micronutrient contents in various parts of 6-year-old Hamlin sweet orange tree on Swingle citrumelo rootstock.
†Numbers within parenthesis represent the standard error of the mean.