AMMONIA VOLATILISATION FROM UREASE INHIBITOR-TREATED UREA APPLIED TO SUGARCANE TRASH BLANKETS

Legal restrictions from burning sugarcane prior to harvest are causing a sharp increase in acreage which is harvested as green cane. The presence of a thick sugarcane trash mulch left after harvest makes it difficult to incorporate fertilisers in the soil. Since large losses of ammonia may occur when urea is surface applied to trash, it is important to find ways to improve urea-N use efficiency. The urease inhibitor NBPT slows down urea hydrolysis and thus may help decrease ammonia losses. Ammonia traps were set up in seven sugarcane fields covered with trash and fertilised with ammonium sulfate or ammonium nitrate, urea, and NBPT-treated urea. All N fertilisers were surface-applied at rates of 80 or 100 kg N ha. Very little N was lost when ammonium nitrate or ammonium sulfate were used. However, volatilisation losses as ammonia from the urea treatments varied from 1% (rainy days after fertilisation) to 25% of the applied N. The percentage of reduction in volatilisation due to NBPT application ranged from 15% to 78% depending on the weather conditions during the days following application of N. Addition of NBPT to urea helped to control ammonia losses, but the inhibitor was less effective when rain sufficient to incorporate urea into the soil occurred only 10 to 15 days or latter after fertiliser application.


INTRODUCTION
Urea accounts for about 60% of the N fertilisers used in the Brazilian agriculture.Other important N sources are ammonium nitrate (AN) and ammonium sulfate (AS).However, in view of the relatively high cost per unit of N in AS and the transport restrictions imposed on AN as a result of the likely use of the latter in the manufacture of explosives, urea (UR) is becoming an increasingly important source of N to sugarcane in Brazil.The main restriction in the use of UR is the possibility of high ammonia volatilisation losses when surface-applied to soils (Terman, 1979), whereas this risk is negligible with AS and AN in the predominantly acid soils of Brazil.
Fertiliser incorporation is a common practice in burnt sugarcane, a condition in which UR is usually as efficient as other N sources not subjected to NH 3 losses (Freney et al., 1992;Cantarella et al., 1999).However, recent Brazilian laws requires sugarcane growers to stop burning.In several important growing areas, more than 60% of the fields are harvested and managed without burning.Surface application of urea to soil covered with sugarcane plant residues may cause NH 3 volatilisation losses between 20 and 40% or more of the N applied (Prammanee et al., 1989;Denmead et al., 1990;Cantarella et al., 1999), but fertiliser incorporation in soil under trash is a difficult and expensive operation.Thus, there is an increasing interest in evaluating the efficiency of surface-applied N fertilisers.
One of the possibilities to control or decrease NH 3 losses is to inhibit the rate of urea hydrolysis in the soil.NBPT [N-(n-butyl) thiophosphoric acid triamide] is now the most promising urease inhibitor and has been commercialised in the USA since 1996.NBPT may inhibit urea hydrolysis for periods between 3 and 14 days depending on soil temperature and other environmental conditions, and has been tested in various countries with generally satisfactory results (Watson et al., 1994;Keerthisinghe & Blakeley, 1995;Cantarella et al., 2005).However it is not known if NBPT can be effective when used in fertilisers applied during the dry winter season (June to September) when part of the ratoon sugarcane fertilisation is done in Southwest Brazil (from May to December).
The objectives of this study were to quantify NH 3 volatilisation losses from surface applied urea and to evaluate the effectiveness of NBPT-treated urea when surface-applied to sugarcane soils covered with trash.
Semi-open chambers to trap NH 3 were used to measure the gaseous N loss from fertilisers, following the model described by Nömmik (1973) and modified by Cantarella et al. (2003).The chambers consisted of PVC hollow cylinders, 20 cm in diameter and 40 cm in height, containing two polyethylene foam discs, impregnated with H 3 PO 4 + glycerol solution in order to absorb the NH 3 released from the soil.The disc placed in the top position was used to trap potential contamination of the atmospheric NH 3 .The lower position disc trapped the NH 3 volatilised from the soil and was used to measure the losses.The upper end of the cylinder was covered with a plastic cap to allow for air circulations.
The PVC chambers were fitted on top of PVC bases, which were cylinders 19 cm in diameter and 20 cm in height and were partially inserted into the soil, leaving around 15 cm of their length above ground.Sugarcane residues were cut with a sharp knife around the bases so that the trash remained inside.Each experimental plot consisted of one chamber and ten bases.The fertiliser treatments were weighed and placed on top of the trash within the PVC bases.The polyethylene foams were replaced periodically, at 2-day intervals in the first two weeks and 3 to 4-day intervals afterwards.Since the chambers do not allow rain or wind to reach the fertiliser placed inside them, and in order to minimise the microclimate effect within the chamber, always when the foams were replaced the chambers were moved to the adjacent PVC base in the same plot.Thus, the NH 3 loss assessment of the following period was done with a fertiliser treatment that had been exposed to the same conditions (rain, temperature, wind etc) as the remaining field.The chambers were also moved to the next base after any raining event along the sampling interval.NBPT was added to urea about two days before its use in the field.The rate of application was 1,325 mg NBPT per kilogram urea in the trial laid down at Iracemapolis in 2003 whereas, at the other six study sites, 530 mg NBPT per kilogram urea was applied.The urease inhibitor used was in a solution containing 200 g kg -1 NBPT in a mixed solvent consisting of Nmethyl-2-pyrrolidone and inert ingredients, supplied by Agrotain International, L.L.C (St. Louis, MO, EUA).Discs collected in the field were sealed in plastic bags and stored in a refrigerator at 5°C until analysis.Ammonia trapped by the foam discs at the lower position in the chamber was extracted by leaching, followed by squeezing, with five consecutive portions of 100 mL of 1.0 mol L -1 KCl solution.An aliquot of the extract was steam-distilled with 10 mL of 5.0 mol L -1 NaOH solution, and then titrated with 0.025 mol L -1 H 2 SO 4 for NH 4 + -N determination (Bremner, 1965).The results are reported as percentage of the N added to the chamber bases.

RESULTS AND DISCUSSION
In the experiment of Iracemapolis (Figure 1A), a 30 mm rainfall had occurred before the experiment was setup but, when the fertiliser treatments were applied, the trash was almost dry.Temperatures were relativelly low (14°C) for this time of the year.Except for small amounts of rain, the weather remained relatively dry for about one month.Volatilisation was dependent on weather conditions.Initially the volatilisation losses were low because of the windy and dry weather in the first days following N application.A 2.6 mm rainfall on the fifth day probably triggered urea hydrolysis and, therefore, NH 3 volatilisation.Other small rainfalls (1.4 mm) 25 and (7.2 mm) 30 days after fertilisation caused an additional slight increase in NH 3 losses but it seems that, at this time, the potential for losses was already small.NBPT-treated urea only started to have volatilisation losses seven days after fertiliser application.Volatilisation proceeded at a slow pace until about the 35 th day and then subsided.At the end of the Iracemapolis experiment, losses of NH 3 from AN treated plots were negligible (0.26% of applied N) but attained 25% and 15.2% of the N applied when UR and UR-NBPT were used respectively (Figure 1A).All means were different (Tukey p ≤ 0.05).
In Araras, the weather was dry except for a light rainfall (0.6 mm) on 4th Nov so that the cane mulch was dry when the fertiliser was applied on the 6th Nov. Later in that same day a sudden change of weather, typical of Brazilian spring and summer, took place around the experimental area and a fast but intensive shower (13.8 mm) occurred.
Volatilisation losses measured 20 h after fertiliser application were negligible indicating that little urea hydrolysis took place in that interval.However, on the second day after N application, NH 3 losses with UR reached 2.5% of the N added and peaked in the fourth day (3.5% of N applied).Volatilisation losses with the UR treatment followed the same pattern observed in previous experiments but with a lower intensity, probably due to the rain that occurred after fertiliser application (Figure 1B).
The 13.8 mm rainfall was not sufficient to incorporate most of the UR into the soil but it may have caused a decrease in volatilisation losses.Ten to 20 mm rain or irrigation is considered sufficient to stop or drastically reduce NH 3 volatilisation when urea is surface-applied to bare soil (Terman, 1979;Hargrove,   1988).However, because of the presence of trash, a greater amount of rain may be required and, as reported by Haysom et al. (1990), a 10 mm rainfall was found to be insufficient to effectively reduce NH 3 losses when urea was applied to an unburned sugarcane field.Freney et al. (1991) observed that more than 16 mm rainfall were needed to cause reduction in NH 3 volatilisation whereas Calcino & Burgess (1995) concluded that 23 mm were not sufficient to dissolve all the urea and incorporate it into the soil.In Brazil, Oliveira et al. (1999) observed that there were still some NH 3 losses after 38 mm of rain.The higher amounts of water needed to incorporate UR in trashcovered soil could be because the water may flow through preferential channels formed within the plant residue cover so that the rain does not efficiently dissolve and leach the urea into the underneath soil (Freney et al., 1994) In Araras, small amounts of rain fell in subsequent days but volatilisation losses only stopped when a 49 mm rainfall occurred on the 12 th day (Figure 1B).The cumulative losses of NH 3 reached about 12% of the N applied as UR.For UR-NBPT, volatilisation losses started on the 4 th day and continued slowly until the 49 mm rainfall event took place; the gradual hydrolysis caused the rates of loss to be relatively low as opposed to that of untreated urea which tend to have volatilisation concentrated in a short period of time due to the fast break down of the urea molecule.Cumulative losses of NH 3 with UR-NBPT reached 7% of the N applied (Figure 1B).The same pattern was observed in the five other experiments with NH 3 volatilisation and the total amount of N lost being strongly dependent on the climate (Table 1).
As expected, volatilisation losses with AS or AN were very small (Table 1).For UR, NH 3 losses varied from 1.1% of the applied N in a place where heavy rain fell soon after fertiliser application, to about 25% under less favourable conditions.The addition of NBPT caused a reduction in volatilisation losses that varied from 15% to 78%, compared with those of untreated UR.In the case of the higher reduction of N loss (Pirassununga), even the volatilisation observed with the UR treatment was small due to the occurrence of rain soon after fertiliser application so that the 78% N loss reduction was of little practical significance.

CONCLUSIONS
The urease inhibitor indeed has contributed to reduce NH 3 losses but, in the experiments under dry conditions, the effect of NBPT tended to be small because, after the inhibitory effect on urease subsided, the fertiliser still remained unincorporated in the soil and, therefore, still subject to volatilisation losses.Furthermore, the NH 3 losses that occurred under long periods of dry weather during the winter months (Jun to Aug) seem to be relatively low: less than 20% of applied N at most sites.The results obtained indicated that, with AN and AS, there were negligible NH 3 losses and that these fertilisers would be the preferable N sources for surface application in unburned sugarcane.However, the lower cost and availability can make urea a more competitive N source particularly in situations where losses are expected to be low as in very dry and cool months.Depending on cost, NBPT, although only partially effective in reducing NH 3 volatilisation losses, can help to increase UR fertiliser efficiency.

Figure 1 -
Figure 1 -Cumulative NH 3 losses from urea (UR), NBPT-treated urea (UR-NBPT) and ammonium nitrate (AN) following surface application to a trash-covered sugarcane soil.(A) Iracemapolis and (B) Araras experiments.Figures above arrows indicate the amount (mm) and position of arrows indicate the time of rain events.Experiments started on 25 th Sep (end of dry season) in Iracemapolis and on 6 th Nov in Araras.

Table 1 -
Ammonia losses by volatilisation from seven field experiments comparing ammonium nitrate (AN) or ammonium sulfate (AS) with urea (UR) or NBPT-treated urea surface-applied to sugar field covered with trash.