Digital Soil Mapping in the Absence of Field Training Data: A Case Study Using Terrain Attributes and Semiautomated Soil Signature Derivation to Distinguish Ecological Potential

Table 4

Field attributes associated with mapped soil classes that include percent of study area for each soil class, distribution of soil taxonomic orders and suborders (where possible) within each soil class, and average soil class properties. Soil classes were mapped by supervised classification of Landsat ETM+ imagery and topographic indices. Data were derived from field surveys of 24 sites (𝑛=8 for each soil class) where soils were augered to depth of 1 m.

Soil	Area^†	Argid^‡	Cambid^‡	Petro-calcid^‡	Soil Depth^§	Surface^¶			Subsurface^¶
						Sand	Silt	Clay	Sand	Silt	Clay
	%				cm	%			%
1	41.2	31.5	68.5	0.0	>100.0	87.3 (1.7)a^#	9.0 (1.2)a	3.8 (0.5)a	79.1 (1.7)a	11.7 (0.7)a	9.2 (1.0)b
2	50.1	0.0	0.0	100.0	51.1 (19.2)	83.3 (1.8)a	11.5 (1.0)a	5.1 (0.8)a	71.0 (2.4)b	15.5 (0.7)b	13.5 (2.1)a
6	7.1	36.4	0.0	63.6	36.3 (27.1)	82.6 (1.7)a	11.8 (1.0)a	5.6 (0.7)a	71.0 (2.1)b	17.5 (1.5)b	11.5 (0.9)ab

^†Percent of study area corresponding to three mapped soil classes that were evaluated in the field.
^‡Percent soil class.
^§Average in soil class where petrocalcic occurred within 100 cm of the soil surface (Petrocalcids).
^¶Average in soil class; surface is 0 to 5 cm; subsurface is horizon with maximum clay in the top meter.
^#Standard deviation in parenthesis, numbers in columns followed by the same letter are not significantly different from Fisher’s least significant difference (𝑃>.05).