Institutional Repository
Thesis Issued 2026-09-06 EN

Prediction of Nitrogen Mineralization in Sub-Tropical Paddy Soils of Bangladesh

Author: Md. Mosaraf Hossain

Abstract

Understanding the factors controlling nitrogen (N) mineralization in paddy soils and the development of useful indicators is essential to improve N use efficiency (NUE) in paddy rice production and thereby reduce the application of relatively costly N fertilizer for rice production. The present study aimed to assess NUE for paddy rice production, determining the ability of some chemical and biological N mineralization indices to predict N mineralization and subsequently validate newly developed N mineralization predictors. To achieve these objectives, A series of field, laboratory and pot experiments were carried out to achieve these objectives. Varietal and seasonal variations in NUE were determined through a field experiment at BAU farm, Mymensingh with 11 popularly grown rice varieties. After that, soil variations in NUE were examined by field trials with T. aman rice (cv BINA dhan-7) in eight locations representing eight soil series. For determining N mineralization, a set of 18 paddy soils were incubated for 14 weeks under saturated condition at 25°C. For predicting N mineralization, various chemical indices such as Hot KCl extraction, Phosphate-borate buffer extraction (PBB), acid and Alkaline KMnO4 extraction, and biological indices viz. Bglucosaminidase, Arylamidase, Protease, ẞ-glucosidase, Phosphomonoesterase, Lglutaminase and Urease activity of soil were tested. This experiment was validated using another set of soils (nine soil series) through laboratory N mineralization and bioassay test (N uptake study) in pot culture and field conditions. Results revealed a wide variability in N use efficiency among rice varieties, seasons and soils. There was no correlation between total N in soil and N mineralization indicating that currently used total N is not a good predictor for determining soil N availability in paddy soil. A significant variation was observed among the soil series of both the soil sets for the chemical and biological N mineralization indices. The combination of both chemical (Acid KMnO4 extractable N) and biological indicators (ẞ-glucosidase and Arylamidase) significantly improved the N mineralization prediction ability in the larger soil set. However, Acid KMnO4 extractable N appeared to be a better predictor for determining N availability for both the paddy soil sets and formulating N fertilizer recommendation in paddy soils.