Uni and multivariate analyses of soil physical quality indicators of a Cambisol from Apodi Plateau ― CE, Brazil

• Published in: Soil & tillage research Vol. 140; pp. 66 - 73
• Main Authors: ANUNCIATO MOTA, Jaedson Cláudio, OLIVEIRA ALVES, Carlos Vítor, FREIRE, Alcione Guimarães, DE ASSIS, Raimundo Nonato
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier 01.07.2014
• Subjects: Agronomy. Soil science and plant productions; Apodi; Biological and medical sciences; Fundamental and applied biological sciences. Psychology; Soil science; South America; Brazil; America; ASW, Brazil; Irrigation; Cambisols; Soil quality; Multivariate analysis; Physical properties; Statistics; Indicator; Soils; Resistance to penetration; Irrigated agriculture; Agriculture; Soil science; Plateaus
• ISSN: 0167-1987; 1879-3444
• DOI: 10.1016/j.still.2014.02.004

Considering that the conventional tillage changes the soil physical properties in a negative way, and that the use of multivariate statistics provides better perception and interpretation of the interactions between the variables involved in the changes than the univariate does, this study aimed to evaluate soil physical quality indicators through statistical techniques and then infer about the physical quality of the soils. The studied area has been cultivated with pineapple for eight years. Water is supplied through a drip irrigation system and the soil is classified as Cambisol. In order to evaluate soil physical quality, disturbed and undisturbed soil samples were collected in these layers for physical analyses: granulometry, dispersed clay, soil bulk and particle densities, organic carbon, aggregate stability, mean weight diameter and resistance to penetration. A completely randomized design was considered with split-plot treatments 2 3 5 (two soil scenarios: under cultivation and secondary forest; three soil layers: 0.0-0.1, 0.1-0.2 and 0.2-0.3 m; five replicates). Statistical tests were used: Kolmogorov-Smirnov to verify data normality, F for variance analysis and Tukey for comparing treatment means, all at 5% of probability. Cluster analysis, principal component analysis and factor analysis were used to better analyze the dataset of the studied variables. As a conclusion, it was refuted the hypothesis of worse physical quality for the soil under conventional pineapple cultivation comparing it to the soil under secondary forest. The multivariate statistical techniques were essential to better understand the set of studied variables and their interaction with soil uses, thus being possible to determine objectively which attributes discriminated each soil use situation and the respective layers. It was also possible to find results not shown by univariate analysis. Therefore, multivariate analysis is recommended as an auxiliary tool in studies to evaluate soil physical quality.