Nutrient and Yield Spatial Variability in a Commercial Malaysian Paddy Field

In recent times, one of the major challenges in paddy plantation is the lack of proper interpretation of yield maps for site-specific management, and the identification and understanding of the causal factors influencing the variability of paddy yields. The ability to find and comprehend the soil factors influencing yield variability of paddy will enable to manage them more efficiently. A study was conducted in a commercial paddy fann at Kahang, Johor, southern part of Peninsular Malaysia with the objectives (i) to quantify and characterize the nature of spatial soil N, P and K variation, (ii) to quantify and characterize the spatial yield and grain moisture variation, and (iii) to establish the inter-relationship between paddy yields and soil N, P and K, so as to identify the ranges of soil nutrients for maximum paddy growth and production for the proposed management zone based on paddy yield variability. Soil samples (within 0-15 cm depth) were collected at 20 m x 20 m grid pattern at the beginning of paddy (MR211 variety) planting and analyzed for total N, available P and exchangeable K in the study plots. A GeoExplorer II was used to record the soil sampling points and boundary of the plots to help interpret the spatial maps in GIS platfonn with proper coordinates. Paddy yields and grain moistures were recorded for harvested paddy sample on a grid pattern of 55 m x 30 m during the harvesting period. Relationship between paddy yields and soil N, P and K were examined using regression analysis. Geostatistical analysis was used to characterize the spatial variation of soil N, P and K, and paddy yields. The coefficient of variation of soil nutrients (N, P and K) for the study area was more than 50% where soil P indicated very high (130%). Significant positive correlationships were found between each pair of soil nutrients (N, P and K). Semivariance analysis showed that the spatial dependence for soil N, P and K was reached at 350 m to 450 m (2 samples ha-I where plot width was 70 m). The kriged soil N, P and K maps showed that a large portion of area (> 80%) in each plot was with soil N lower than 3 g kg-I, soil P lower than 10 mg kg-I, and soil K more than 0.45 cmol(+) kg-I. The paddy yield variation was more than 47% and was significant within and between plots. The semivariance for paddy yield was attained at 380 m, suggesting that the spatial correlation existed within 2.6 ha (70 m x 380 m) where 70 m is width of the plot. The paddy grain moisture showed randomly distributed after a separation distance of 30 m and it has no significant variation within the plots. The yield kriged map illustrated that the middle portion (> 50% of the area) of each plot has lower yield « 3.5 t ha-I) compared to both ends in lengthwise. Significant positive correlations were obtained between each soil nutrient (N, P and K) and yield. In the combined effect of soil N, P and K on yields, soil N and P together could explain 41 % of yield variations where P accounted for 36% and N accounted for 5% of the total yield variation (y = 2.41 + 0.39 XN + 0.20 Xp; R2 = 0.41 ** , n = 159). Based on this interrelationship, a management zone was suggested in yield maps where soil N and P need to be improved in moderate and low ranges, respectively, for obtaining maximum yield (5 t ha-I). Some selected soil properties were measured at different locations and the variations were found to be 4.25 to 4.54 in pH, 0.77 to 0.92 g cm-3 in bulk density, and the soil texture was clay with 75-83% of clay .content. The yield difference between estimated yield from spatial maps and actual yield from the combine ha.-vester was ±10%. To characterize the combine harvester, average machine operation grain loss was determined which was about 2% of total grain yield, and the average machine field capacity was 1.01 ha h-I with 72% of field efficiency. Tlt1J'iJ

Text IB_2001_5_.pdf Download (1MB)

Actions (login required)

View Item