Sensitivity analysis to investigate the factors controlling the effectiveness of a nitrification inhibitor in the soil

Abstract

Nitrification inhibitors have been proposed as an option to mitigate nitrogen (N) losses from farmland. One of these inhibitors, dicyandiamide (DCD), has been shown to significantly reduce N leaching and N2O emissions from nitrogen enriched soils with no adverse effects to the soil environment. However, the effectiveness of DCD in reducing N losses has been shown to be quite variable. We postulate this variability can be linked to the persistence of DCD in the soil layers where ammonium is located. DCD in soils will be subject to biodegradation and so the persistence of DCD in soils should be inversely proportional to microbial activity, for which temperature is a major driver. Furthermore, DCD is a neutral-charge compound and so should be highly mobile in soils and prone to leaching in direct proportion to rainfall. Thus, after application, the distribution of DCD in the soil profile over time can be affected by temperature, rainfall and the soil's drainage characteristics. We employed the simulation model APSIM with a recently-developed DCD module to analyse how the environmental conditions affect the processes governing the persistence of DCD in soils. The analyses involve data from simulations of typical pasture land in different locations across New Zealand with representative rainfall and temperature regimes and soil types. The simulations considered the application of DCD at different rates and in different seasons, over a 20 years period. Using sensitivity analysis techniques, we studied which processes play the greatest role in the disappearance of DCD. This led to recommendations for experimental protocols, including soil sampling, and data analysis which can be used to discern between leaching and degradation. The persistence of DCD was studied using the solutes' apparent half-life, that is, the time taken for the concentration of the solute to decrease to half of its initial concentration. We use the adjective 'apparent' because the reduction in concentration is a result of a combination of degradation and leaching rather than degradation alone. The results from the simulations with DCD were compared with those from simulations with the application of an inert tracer in which the apparent half-life is affected by leaching alone. Differences in the apparent persistence of DCD and tracer from the soil's top 20 cm layer were used to infer the relative effects of degradation and leaching. The results showed large variability for the values of apparent half-life of both solutes, but the variability was considerably smaller for DCD than for the tracer. The month of application and location (weather) were the two major factors driving the variations in apparent half-life of both solutes, with soil-type and some of the interactions between factors being secondary. However the order of relevance for the factors was different for DCD than for the tracer. Overall the influence of month of application in the apparent half-life of DCD was quite small, whereas it was the most important factor for the disappearance of the tracer. The difference between the apparent half-lives of the two solutes can be used to infer the effect of degradation on DCD persistence; the results show, therefore, that the month of application is very important for determining DCD degradation. The relative effects of leaching and degradation follow a clear seasonal pattern, in summer degradation is the major factor and leaching is of lesser importance, but in winter leaching is much more important than degradation. This offset seasonality for the two processes results in the low overall seasonality for the apparent half-life of DCD. It also emphasizes the challenge of distinguishing the relative importance of leaching and degradation for DCD persistence in the soil. To better determine the causes of variation in residence time of DCD, and thus its effectiveness, experiments should use a tracer applied alongside DCD.