Agriculture Reference
In-Depth Information
second, the question of whether the diagnostic properties used in creating gen-
eral-purpose classes are relevant to wine grapes.
The relationship between map scale and map accuracy is discussed in box 8.1.
Because viticulture is an intensive operation, typically with small, individual vine-
yards (5-50 ha), soil data should be collected at a high density of observations.
With the aid of a Global Positioning System (GPS), the observation sites can be
accurately spatially referenced, which facilitates entry of data into the layers of a
GIS. In this case, general-purpose soil maps can be replaced with a database of
key soil properties, which can be visually displayed through the mapping func-
tion of the GIS software. The GIS facilitates the extraction and graphic display
of soil properties specifically relevant to viticulture. These data can be used to cre-
ate a
special-purpose classification
of soils, such as the Australian Viticultural Soil
Key developed by Maschmedt et al. (2002). As far as possible, this key uses di-
agnostic soil properties that are easily recognizable in the field and are described
in nontechnical language. The focus is on identifying soil layers that restrict root
growth, such as the depth to waterlogging features (mottling and gleying), depth
to rock that is soft (rippable) or hard (nonrippable), stoniness, soil structure and
consistence, presence and hardness of CaCO
3
layers at different depths, soil tex-
ture, cracking, and the change in texture with depth.
Although a special-purpose classification may bear little relation to a general-
purpose classification such as the Australian Soil Classification (ASC) (Isbell 1996)
or the U.S. Soil Taxonomy (Soil Survey Staff 1996), the classes can be approxi-
mately allocated to higher level classes of such general-purpose classifications at
the regional level. For example, soil types identified in the Australian Viticultural
Soil Key have been allocated to Great Soil Group (Stace et al. 1968), ASC, Soil
Taxonomy and World Soil Reference Base (FAO 1998) classes. Similarly, in the
Box 8.1
Map Scale and Accuracy
The accuracy of a soil map is checked by determining whether a chosen soil
type, identified at a point on the map, is actually found at that point on the
ground. If the density of sampling is too low and soil variability too high, a soil
map is unlikely to be accurate.
Normally, soil variability within blocks
10 m) cannot be
mapped at an acceptable scale, so this restriction sets an upper limit to the density
of sampling. The practical limit is therefore a scale of 1:1,000, for which 1 cm on
the map represents 10 m on the ground. This is a
large scale
. Where proximal
sensors of soil properties are used (section 8.2.2.1) and digitized data recorded in a
GIS, maps of even larger scale can be produced. The scale of a map decreases as
distance on the map decreases relative to distance on the ground, so 1:1,000,000 is
a
small scale
.
Vineyard soil maps at a large scale (
100 m
2
(10
1:10,000) are needed to show local
variations as accurately as possible. Many soil maps based on general-purpose
classifications are small scale (1:1,000,000 or smaller). Sometimes, to identify soils
suitable for vineyards, small-scale maps are expanded to larger scales, for example,
by photocopier enlargement or by digitizing the map and entering the data into a
GIS. This is a bad practice because the density of the original soil sampling is
insufficient to justify showing soil boundaries at the larger scale.
