Modelling the relationship between soil color and particle size for soil survey in ferralsol environments
2009
Kone, B. (African Rice Center, WARDA (Cotonou)) | Yao-Kouame, A. (Universite de Cocody-Abidjan (Cote d'Ivoire). UFR des Sciences de la Terre et des Ressources Miniere) | Ettien, J.B. (African Rice Center, WARDA (Cotonou)) | Oikeh, S. (Centre National de Recherche Agronomique, Gagnoa (Cote d'Ivoire)) | Yoro, G. | Diatta, S.
Soil texture is an important property for evaluating its inherent fertility especially by using pedo-transfers functions requiring particle size data. However, there is no existing quantitative method for in situ estimation of soil particle size, delaying judgement of soil chemical properties in the field. For this purpose, laboratory particle size analyses of 1028 samples from 281 Ferralsol profiles, located between latitudes 7 degree N and 10 degree N in Cote d'Ivoire and their respective colour notation by Munsell chart were used to generate prediction models. Multiple Linear Regression Analysis by Group was processed to identify clay, sand and silt contents in the soil based on color hue (2.5YR, 5YR, 7.5YR, and 10YR) and Chroma (1, 2, 3, 4, 5, 6, 7, 8). The evaluation was conducted for each horizon coded as HI (0-20 cm), H2 (20-60 cm), H3 (60-80 cm) and H4 (80-150 em) and used as grouping variables. Highly significant (P less than 0.001) m odels were identified for clay and sand. These models were used to estimate successfully clay and sand contents for other Ferralsol samples by comparing calculated and measured mean using the null hypothesis of difference and Turkey's tests. They were accurate for at all depths, except 80-150 cm, for sand in 10YR soils. The method was deemed appropriate for in situ estimation of soil particle size contents in Ferralsol environment for improving reconnaissance agricultural soil surveys.
显示更多 [+] 显示较少 [-]