Using regional geochemical data for agriculturally relevant applications

Geochemical data collected for regional mapping have considerable, and largely untapped, potential to benefit the agricultural sector. However, variables measured in geochemical surveys do not necessarily correspond directly... [ view full abstract ]

Geochemical data collected for regional mapping have considerable, and largely untapped, potential to benefit the agricultural sector. However, variables measured in geochemical surveys do not necessarily correspond directly to those which are used for agronomic guidance. It is also necessary to manage the considerable variability exhibited at a range of scales, which introduces uncertainty into the interpretation of regional-scale data at the scale of the field or farm. This uncertainty must be quantified and communicated effectively to the data-user.

Here we present examples from Irish and British surveys, relevant to the grazing livestock sector. Data on pH of soils under pasture are compared with recommended soil pH thresholds (6.0 and 6.5 for grass and grass–clover systems respectively) across the north of Ireland. These demonstrate that over much of the area liming is required to achieve the chosen target pH values (Lark et al., in press). The variation in soil cobalt and manganese is assessed against Teagasc guidance, demonstrating the importance of the joint variation in the variable of primary interest to livestock health (cobalt) and of a key soil complexing element (manganese) (Lark et al., 2014, Geoderma 226:64). Both studies quantify the uncertainty in predictions at un-sampled locations arising from spatial variability, particularly at short scales. The uncertainty is expressed in probabilistic terms on a verbal scale and visualized, accounting for recent psychological research on the efficacy and limitations of such scales. Other elements, such as iodine and magnesium, require more work to understand transfer from soil to herbage.