Losses of nitrate from a sandy loam soil under corn: lysimeter experiment

LOSSES OF NITRA TE FROM A SANDY LOAM SOIL UNDER CORN: L YSIMETER EXPERIMENT. F. Cabrera, A. Reyes, E. Femández-Boy, 1.A. Cayuela. 1.M. Murillo and F. Moreno. Instituto de Recursos Naturales y Agrobiología de Sevilla. (CSIC). Apartado 1052. 41080 Sevilla. Additional index words: irrigation, Kjeldahl-N, drainage water, soil water content. Abstraet Losses of nitrate from an undisturbed sandy loam soil contained in two monolith lysimeters, Lo and L¡, (1 m diameter, 1.5 m deep) under com and irrigation (628 mm) were studied. The crop was fertilized with O (Lo) and 800 (L¡) kg N ha-l. Water draining was collected periodically and analysed for nitrate content during the experimental period (20 March 91-23 March 92). Mean values of nitrate concentration in the drainage water were 3.1 and 41.2 mg NOrN 1-1 in Lo and Ll respectively, the latter much higher than the limit imposed by the EC for potable water (11.3 mg N l-I). Three periods were distinguished during the experiment: i) Crop season, in which the total volume of water drained was 310 and 47 mm respectively in the Lo and Ll, while the nitrate 10sses were of the same order in both lysimeters ( 21 and 17 kg N hal, respectively); ii) Dry season, in which there was no water drainage; and iii) Rainy season, in which the total drained water was 157 and 139 mm, and the nitrate los ses 5 and 91 kg N hal in Lo and L¡ respectively. The greater amount of water drained in Lo is due to the lower crop development (grain yield 4.4 and 21.3 t hal; total nitro gen plant uptake 69 and 400 kg N hal, in Lo and Ll respectively). Total nitrate losses in Ll account for 13% of the nitro gen fertilizer applied. Most of these losses occur in the rainy season. At the end of the experimental period a decrease of the soil nitrate content was observed in both lysimeters. Introduction Production of heavy yields depends on the use of large quantities of inorganic fertilizers, which can imply an environmental risk: the pollution of groundwater by nitrate. Whatever form of nitro gen fertilizer is used, the end product in the soil is nitrate, which is vulnerable to being washed out of the soil by rain or irrigation. The most decisive factors determining the magnitude of nitrate leaching are nitrogen fertilization, c1imatic conditions, type of soil, soil management and type of crop (Gustafson, 1983; Duynisveld et al., 1988; Addiscott et al., 1991). Measurement of nitrate los ses under field conditions is difficult. Five methods for measuring the concentration and fluxes of nitrate leaving the soil were outlined by Addiscott (1990). A good approach is the use of monolith lysimeter containing undisturbed soil (Dowdell et al., 1984; Owen, 1990, Bergstrom and Johansson, 1991). The aim of this study was to evaluate timing and quantity of nitrate leaching in a highly fertilized soil under corn and irrigation using monolith lysimeters. These are preliminary results of a wider study lasting four years. 2 Material and methods Two monolith lysimeters (1 m diameter, 1.5 m deep) were constructed without disturbing the soil profile and reinstalled in situ in a plot of 0.1 ha. The soil was a sandy loam XEROCHREPT (pH(H2 0) 7.2 and 7.1; CaCOJ 5.2 and 3.1 %; O.M. 0.88 and 0.69 %; Kje1dahl-N 599 and 454 mg kg-I; NO)-N 7.5 and 10.4 mg kgI at O-50 and 50-100 cm respeetively. These are mean values of 45 samples). Lysimeters were provided with a system to colleet drainage water (1.2 m deep) and with aeeess for neutron probe. The 1ysimeters were sown with corn (170000 p1ant haI, 5 April 1991) and fertilized with O (Lo) and 800 kg N haI (LI ). Fertilizer was applied in three steps: a deep fertilization before sowing (22 March 1991) with 1000 kg haI with a 15-15-15 complex NPK fertilizer, and two top dressings with 400 kg haI (24 May 1991) and 1000 kg haI (7 June 1991) of urea (46 %) respectively. Lysimeters were periodically irrigated (EC 2.0 dS m-I; SAR 2.0 meqlfl I-Ifl; NO)-N 10.5 mg 1-1) receiving a total amount of water equivalent to 628 mm. Water draining was collected periodically and analysed for nitrate during the experimental period (20 March 1991-23 March 1992) (figure 1). Nitrate in water was determined by ionie chromatography using a mixture of borate-gluconate pH 8.5 as eluent. Soil water content and rainfall were also monitored throughout the experimental period (total rainfal1 equivalent to 293 mm (figure 1». Soil samp1es were collected at 0-30,30-60 and 60-90 cm at the end of the experiment by means of a thin auger. Soil Kje1dahl-N was determined by the method deseribed by Hesse (1971). Soil NOrN was extracted by 0.2% Ca(OH)z suspension (Sims and Jackson, 1971) and the extracts analysed for NOrN by the method of Scheiner (1974). Plant height was measured periodically. Yield, mean weigth of ears and total Kje1dah1-N in kernel were determined at harvest. 3 Results and Discussion Volumes of water drained by Lo are greater than those drained by L¡, and both are re1ated with the vo1umes of water supplied by irrigation and rainfall (figure 1). The total eomu1ative volume of water drained by Lo is also greater than that by L¡ (figure 2). Concentrations of NOrN in drainage water ranged from <1 to 41.9 and <1 to 155 mg NO)-N 11 respeetively in Lo and L¡. Their mean values were 3.1 and 41.2 mg NOr N I-¡ (figure 1), the latter being much higher than the 'maximum admissible coneentration' imposed by the EC for potable water (11.3 mg NOJ-N 1-1). Three seasons can be distinguished during the experimental period: i) Crop season (20 March 1991-S August 1991; ii) Dry season (6 August 1991-30 September 1991); iii) Rainy season (1 October 1991-23 March 1992). During the 'crop season' the lysimeters received a total of 628 and 55 mm of water by irrigation and rainfall respectively. In this season· total water drained was 310 and 47 mm respectively in Lo and L1, and total nitrate losses were of the same order in both lysimeters (21 and 17 kg NOrN ha-¡ respectively). Analysing the hydric profiles of the lysimeters during the maximun development of the crop (figure 3), it can be observed that on 19 July 1991 (a few hours after the beginning of an irrigation), water content decreased throughout the soil profile down to 20 and 70 cm depth respectively in Lo and L¡. Afterwards water content remained approximately constant at 0.227±O.OO8 and 0.151±O.004 cm3 cm3 (mean±SD) respectively in Lo and L¡. Four days later (23 July 1991), water content at 0-80 cm in Lo and 0-60 cm in L¡ decreased compared with the situation of 19 July 1991, and the shape of the hydric profile were similar for both lysimeters down to 70 cm depth. From 80 cm down in Lo and 70 cm in L¡, water contents were nearly constant, 0.226±O.011 cm3 cm3 in Lo and 0.141±O.01 cm3 cm3 in L¡, values similar to those found at the same depth on 19 July. Therefore, during this irrigation event water content in the soil profile of Lo is always greater than in that of L¡. The same situation was observed in all the irrigation events throughout the 'crop season'. As total volume of water drained was greater in Lo than in L¡ (figure 2), it is supposed that water depletion in L¡ is mainly caused by the higher water uptake of the crop in this lysimeter because of the higher plant development. In fact, table 1 shows that corn plants of L¡ were toller than those of Lo, and that the yield, mean weight of ears and total Kjeldahl-N in kernel, was also greater in L¡. , In the 'dry season' there was no water drainage and conséquently no los ses of nitrate in either of the lysimeters. During the 'rainy season', total water drained in both lysimeters was similar, 15Tand 139 mm in Lo and L¡ respectively, 66 and 58 % of the total rainfall in this periodo However, in the same period nitrate losses in the drainage water were much higher in L¡ (91 kg NOrN ha-¡) than in Lo (S kg N0 3-N ha-¡). Nitrate los ses in L¡ during the 'rainy season' were the 11 % of the N applied as fertilizer. Total nitrate leached in L¡ was 108 kg N0 3-N ha-¡, equivalent to 13% of the total nitro gen applied as fertílizer. This quantity is higher than that reported by Croll and Hayes (60-80 kg NOrN ha-¡) for spring-sown cereals, and would be sufficient to constitute a serious threat to the maintenance of the EC-recommended level in groundwater (Foster et al., 1982). At the end of the 'rainy season' it was found that the content of Kjeldahl-N throughout the profile decreased in Lo (5733 kg N ha-¡) and increased in L¡ (6396 kg N ha-¡) with respect to the average initial content of the soil (61S7 kg N ha-¡), although neither of the differences was statistically significant (P<0.05) (figure 4). At the same time, nitrate content in the soil profile decreased significantly (P<Ü.OS) with respect to the average initial content of the soil (104 kg N0 3-N ha-¡) inboth Lo (23.8 kg NOrN haI) and Ll (38.9 kg N0 3-N ha-I) (figure 4). Figure 5 shows a balance of the nitro gen inputs and outputs iri Lo and L¡. It can be observed that at the end of the experimental perlod it was possible to determine the fate of most of the nitro gen of the system. In fact, in Lo the difference between the initial and final states (6327 and 5853 kg N hao! respectively) was not significant (P<0_05) and accounts for sorne 7% of the nitrogen in the initial state. In L! total nitrogen in the initial state was 7121 kg N hao! and in the final one 6958 kg: N hao!, the difference being non-significant (P<Ü.05) and equal to sorne 2% of the nitrogen in the initial state. These differences, though non-significant, can be explained by the inherent error of the methods, losses by ammonia vaporization from urea, by denitrification, or even by the releas e of gaseous nitrogen (mainly arnmonia) by the plants themselves. Acknowled~ment Research carried out in the framework of contract No. STEP-CT90-0032-C (DSCN) oftheCE. References Addiscott, T.M. 1990. Measurement of nitrate leaching: a review of methods. In: Nitrate-Agriculture-Eau (R. Calvet). INRA. Paris. Addiscott, T.M., Whitmore, A.P., and Powlson, D.S. 1991. Farming, fertilizers and the nitrate problem. C.A.B. International. Wallingford. Bergstrom, L, and Johansson, R. 1991. Leaching of nitrate from monolith lysimeters of different types of agricultural soils_ J. Environ_ Qua!. 20: 801-807. Croll, B.T., and Hayes, C.R. 1988. Nitrate and water supplies in the United Kingdom. Environ. Pollul. 50: 163-187. Dowdell, R.J., Webster, c.P., Hill, D., and Mercer, E.R.1984. A lysimeter study of the fate of nitro gen in spring barley crops grown on a shallow soil overlying Chalk: crop uptake and leaching los ses. J. Soil. Sci. 35: 183-190. Duynisveld, W.H.M., Strebel, O, and Bottcher, J. 1988. Are nitrate leaching from arable land and nitrate pollution of ground water avoidable? Ecol. Bull. 39: 116125. Foster, S.S.D., Cripps, A.C., and Smith-Carrington, A. 1982. Nitrate leaching lo groundwater. Phil Trans. Roy. SOCo 18: 103-116. Gustafson, A., 1983. Leaching of nitrate from arable land into groundwater in Sweden. Environ. Geol. 5: 65-71. Hesse, P.R. 1971. A textbook of soil chemistry analysis. John Murray Pub. Ltd. London. Owens, L.B. 1990. Nitrate-nitrogen concentration in percolate from lysimeters planted to a legume-grass mixture. J. Environ. Qual. 19: 131-135. Scheiner, D. 1974. A modified version of the sodium salicylate method for analysis of wastewater nitrates. Water. Res. 8: 935-840. Sims, J.R., and Jackson, G.D. 1971. Rapid analysis of soil nitrate with chromotropic acid. Soil Sci. Soco Amer. Proc. 35: 603-606. Table 1 - Mean values of corn plant height and results of corn performance at harvest. Lysimeter Plant Plant Yield Ear Kernel height height 18 Jun 18 Jul mean Total weight Kjeldahl-N cm cm kgha1 g % 70a 136b 218 a 247 b 4415 21295 543 a 187.3 b 1.22 1.53 Values followed by different letters in the same column differ significantly (P<O.05). 30 e 20 e ~ 10 111 .... r--::::c=-r-=-=-,1 200 l'IU'_ -l'lU 150 ; 0~~~~~~~~~--~-40 L -t 1 200 z· :'30 ySlme er 1_ water 1 1 111 "! .• NOs-N. 150 0" .5 20 • Z 111 '"' "tl 10 100 50 El 58 1----r"-...¡~"-r-_j::.JII....4I'-......,.-+_'___l O e 40 §30 :;3 11120 bO .í:: 10 .!:: O e 50 8 40 .....;'30 'iii 'j320 .¡¡¡ 10 '"' 0f--'L-r-...L.r---,---'~---flL--'r--"""~ O 100 200 300 400 Time (days (rom 20 March 91) Figure 1 - Drained water, nitrate in drainage water, and irrigation and rainfall data. El El 400 • 111 e ;:1 'O 300 > 111 > :;¡ 200 '5 e 8 100 OI---'4""""'t----+-+--+-+---+----l '111 120 ..c:: bO ~ 100 Z 1., 80 O Z 111 60 > :;1 I\l '5 40 8 ;:1 U 20 Nitrate losses • Lysimeter o o Lysimeter 1 OL__ E......-L.._-'-----'-_...L----L---.J O 100 200 300 400 Time (days (rom 20 March 91) Figure 2 - Cumulative volume of drained water and cumulative nitrate losses. 3 -3 water content, cm cm 0.0 0.1 0.2 0.3 0.4 0.5 O r-'--r-'--~-r~--r-,--r-, • LO 19 JUL .. LO 23 JUL o L1 19 JUL v L1 23 JUL 90 Figure 3 - Hydric profiles of the lysimeters. . 140 D initial lS:Sl final LO 120 ~ final L1 C.L. 95% 100 i' <ti 80 .c: C.L. 95% ~60 40 20 O L-~~~~L-_~2~-L~~~--~ Kjel.-Nxl0 N0 3 -N Figure 4 - Contents of Kjeldahl-N and N0 3-N in the soils of the two lysimeters. mm! soil N03 -N rn soil Kjeldahl-N _ irr. water N03 -N ~ plant uptake N D fertilizer N rs:s:J drainage water N03 -N «3 ..r:::: t:ID .!S: 10000 5000 1000 Z 500 O initial tinal (Ti rri I L . O inilial tinal state state • III1111 Figure 5 - Nitrogen balance.