Statistical Modeling of the Variability of Selected Colo Soil Properties

Forty-seven soils representing the Colo soil series (fine-silty, mixed, mesic Cumulic Haplaquolls) as mapped on 0 to 2% slopes in the Upper Midwest were described and sampled to determine selected physical, chemical, and morphological properties to use in statistical modeling. The soils were sampled in Iowa (35 pedons), Nebraska (4), Missouri (4), Illinois (2), and Minnesota (2), and studied in the following groups: all Colo pedons, by master horizon, and by the presence or absence of postcultural sediments (PCS). The groups were statistically analyzed by using means, ranges, standard deviations, coefficients of variation, correlation coefficients, and multiple regression equations. The objectives of this study were to (i) characterize the Colo soil throughout its geographic range; (ii) stratify the data to study the relationship between selected soil physical, chemical, and morphological properties; and (iii) develop statistical regression models using several soil chemical and physical properties of one soil series. Ranges in soil properties such as particle size, hydrogen-ion activity, and total carbon (TC) content are restricted by the limits of the official series used as guides in mapping. Therefore, these properties did not have as wide a range as other soil properties studied-available, total (TP), organic (OP), and inorganic phosphorus (IP) or organic carbon/organic phosphorus (OC/OP) and organic phosphorus/total phosphorus (OC/TP) ratios which are not defined for the series. The Colo samples were stratified according to master horizon designations. TC, OP/TP, OP, and OC/OP means were highest in the A horizon and decreased with increasing depth in the profile. Soil properties used to define the series had a narrower range (e.g. particle size, TC amount) than other soil properties especially in the A horizon. Forty-five percent of the pedons sampled had some evidence of PCS. The sand, silt, clay, and TC means were similar for those soils with or without PCS as were IP, OC/OP, and OP/TP. The ranges differed but were not as great as anticipated. If less than 50 cm thick, PCS should not be considered part of the mollic epipedon. Still all the soils studied had a mollic epipedon greater than or equal to 91 cm thick. Morphologically, the majority of the Colo soils described had an A-B-C horizon sequence. A B horizon was described to indicate “structural” alteration of original material. Alluvial-derived soils are reported to be in an early stage of development and, therefore, should have little alteration of the original material. As indicated by some of the Colo soil properties, this assumption is not necessarily correct. As a result of this study, the genesis of some soils derived from alluvial sediments in the Upper Midwest is better understood.