Knowledge of soil conditions and the ability to predict soil response to changes in climate and management are essential for strategic resource planning and assessment at local, regional and global scales. Soil and its dynamic interactions as an ecosystem component, including anthropogenic impacts, must be characterized in order to portray causes and effects necessary to make useful predictions of future soil responses. There is sufficient expertise to achieve this goal, although it is fragmented and exists in a variety of disciplines including, among others, geomorphology, terrestrial, community and landscape ecology, ecophysiology, climatology, pedology, agronomy, range sciences, and forestry. Key processes involved in soil change are the transfer of energy and matter among communities within a landscape; feedbacks between the biotic and abiotic components; spatial and temporal patterning that reflects self-reorganization; and the threshold response within these processes that drives change in the system. The goal of this paper is to present necessary elements for integrated approaches to characterize soil change as one step toward multi-use enhancements of soil survey databases. The elements include the human time scale, attributes of soil change, disturbance impacts on ecological, pedogenic, and geomorphic processes, function-based soil resistance and resilience and non-equilibrium models.

The human time scale should be defined on the basis of past human impacts on the land within the current climatic regime and likely periods of recovery, with respect to processes. Most generally stated, the human time scale covers decades to centuries. The current climate governs the geomorphic processes that are shaping or preserving the landscape today. Consequently, time zero of the human time scale should be no earlier than the beginning of the most recent episode of climatically-controlled landscape development, roughly 300-3000 years ago, even though human impacts may have occurred prior to that time. Because of the emphasis on geomorphic processes, this period differs from the Anthropocene. Predictions of soil response to management over the human time scale also require a time limit in the future. Some have proposed a management time period of decades based on common agricultural enterprise. However, the time frame of recovery must be included and it may be periods of centuries in arid and semi-arid systems where recovery of nutrients is slow and landscape aggradation and degradation processes are active.

Data collection to characterize soil and its dynamic interactions within agricultural and non-agricultural ecosystems should be based on concepts that reflect change, metrics that can be measured or modeled, and conceptual models that represent biotic-abiotic interactions. Soil change is temporal variation in soil properties at a given location. Metapedogenesis provides a schematic representation of human impacts on pedogenesis and is designed to help organize research and long-term studies. Processes and concepts essential for characterizing soil change, which this author identifies as attributes of soil change, include state variables (dynamic soil properties), time changeability, reversibility, trends, rates, drivers, thresholds, and pathways of change. These attributes need to be quantified and used to calculate other function-based metrics, including resistance and resilience, which are vital parameters for sound resource management. The state and transition model as applied to rangeland ecosystems is capable of capturing non-equilibrium and non-linear dynamics of the effects of disturbances, such as management or climate change on plant-soil interactions. The state and transition model can provide a tool to organize soil-plant-disturbance information for both improved decision making at a wide variety of scales and generating research. It can also be used to illustrate and help define metrics for soil change attributes and function-based soil interpretations.