The variability of soil as the basis for determining fertilization recommendation for rice in North Sumatra, Indonesia

Soil fertility and soil units at a detailed scale are needed to determine site-specific nutrient management. The objective of the study was to describe the variability of lowland rice fertility and fertilizer recommendation in North Sumatra. Field research was investigated at Pasar Miring Experimental Station, Deli Serdang District, North Sumatra Province (latitude: 3°, 29', 57''N, longitude: 98°, 54', 23'' m.a.s.l.) puddled soil with a land area of 15.33 ha for the period August-December 2020. The study was preceded by a work desk, field survey, analysis of soil samples in the laboratory, data processing, and reporting. GIS applications are used for data processing. The soil series is distinguished by the type and arrangement of horizons, colors, textures, structures, consistency, soil reactions of each horizon, other soil chemical properties, and mineral properties of each horizon. The results showed that (1) Soil in the study area was tuff acid dacite with above alluvium cover material was sand and loam that it required lime and organic material to improve soil structure, especially on a bit rough to rough textures (2) Soil classification in the experimental area was Inceptisols ordo, Aquepts suborder, Endoaquepts group, Typic Endoaquepts subgroup, and Kotasan 1, Kotasan 2, Kotasan 3, Jaharum B1, and Jaharum B2 series; (3) Soil fertility map scale of 1: 2000 at depth 0-20 cm and 20-40 cm were divided into three categories i.e. inherent potential of soil fertility status (IP) with high to very low grade, organic matter and soil N status (OM) with very high to very low class, as well as P availability status (AP) with very high to medium grade; (4). Applied mineral soil + ash empty fruit bunches of palm oil increased rice production up to 1.3 t ha-1 (15%).