History of soil geography in the context of scale

• Soil maps produced manually reflect the soil geographer's (mapper's) mental model.
• Advancement of soil science has been focused by the scale of the maps.
• The scale hierarchy of soil formation factors is observable in historical soil maps.
• The debate over dominant factors is moot in the context of scale.
• Better maps of land surface topography are the next opportunity for improving soil maps.

We review historical soil maps from a geographical perspective, in contrast to the more traditional temporal–historical perspective. Our geographical approach examines and compares soil maps based on their scale and classification system. To analyze the connection between scale in historical soil maps and their associated classification systems, we place soil maps into three categories of cartographic scale. We then examine how categories of cartographic scale correspond to the selection of environmental soil predictors used to initially create the maps, as reflected by the maps' legend. Previous analyses of soil mapping from the temporal perspective have concluded that soil classification systems have co-evolved with gains in soil knowledge. We conclude that paradigm shifts in soil mapping and classification can be better explained by not only their correlation to historical improvements in scientific understanding, but also by differences in purpose for mapping, and due to advancements in geographic technology. We observe that, throughout history, small cartographic scale maps have tended to emphasize climate–vegetation zonation. Medium cartographic scale maps have put more emphasis on parent material as a variable to explain soil distributions. And finally, soil maps at large cartographic scales have relied more on topography as a predictive factor. Importantly, a key characteristic of modern soil classification systems is their multi-scale approach, which incorporates these phenomena scales within their classification hierarchies. Although most modern soil classification systems are based on soil properties, the soil map remains a model, the purpose of which is to predict the spatial distributions of those properties. Hence, multi-scale classification systems still tend to be organized, at least in part, by this observed spatial hierarchy. Although the hierarchy observed in this study is generally known in pedology today, it also represents a new view on the evolution of soil science. Increased recognition of this hierarchy may also help to more holistically combine soil formation factors with soil geography and pattern, particularly in the context of digital soil mapping.