|Scudiero, Elia -|
|Teatini, Pietro -|
|Deiana, Rita -|
|Berti, Antonio -|
|Morari, Francesco -|
Submitted to: Geoderma
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: August 30, 2013
Publication Date: N/A
Interpretive Summary: Crop yield varies within fields due to nonuniformity of a number of factors including climate, pests, disease, management, topography, and soil. Conventional farming manages a field uniformly; as a result, conventional farming tends to wastes resources and money, and tends to detrimentally impact the environment. One way of handling crop yield variability in a cost and resource effective manner is to divide a field into management units based on the observed yield and soil variability so that each unit can be treated similarly in order to optimize yield, resource utilization, and profitability, and minimize detrimental environmental impacts. These site-specific management units or SSMUs are a key component of precision agriculture. It was the objective of this study to build on previous work for delineating SSMUs with a single sensor by using a combination of sensors that provides complementary information. Two sensors were used to define SSMUs: remote imagery (normalized difference vegetation index; NDVI) and apparent soil electrical conductivity (ECa). This study provided strong evidence for the need to use multiple-sensor platforms to characterize more extensively the spatial variability of soil properties influencing yield as a means of making maps of SSMUs. Maps of the SSMUs provide producers, extension specialist, agricultural consultants, and NRCS staff with information for variable-rate technology (e.g., site-specific fertilizer and irrigation water application).
Technical Abstract: Site-specific management units (SSMUs) are spatial domains within a field that enable producers to apply inputs when, where, and in the amount needed as a means of optimizing productivity, resource utilization, and profitability, and minimizing detrimental environmental impacts. It is the objective of this study to delineate SSMUs using remote imagery and geospatial apparent soil electrical conductivity measurements (ECa). The study site was a 21-ha field on the southern margin of the Venice Lagoon, Italy, which is known to have considerable spatial variability of soil properties influencing crop yield (i.e., salinity, texture, available water, organic carbon). Maize (Zea mais L.) yield maps from 2010 and 2011 showed high spatial heterogeneity primarily due to variation in soil-related factors. Spatial variation in maize yield was successfully modeled according to the variability of four soil properties: soil salinity, texture, organic carbon content, and bulk density. The spatial variability of these soil properties was characterized by the combined use of intensive geospatial ECa measurements and bare-soil normalized difference vegetation index (NDVI) survey data. On the basis of the relationships with these soil properties, ECa and NDVI were used to divide the field into SSMUs using fuzzy c-means clustering. The delineation procedure identified five SSMUs: one homogeneous with optimal maize yield, one unit affected by high soil salinity, one characterized by very coarse texture (i.e., sandy paleochannels), and two zones with both soil salinity and high organic carbon content. Findings revealed that even in cases where ECa and bare-soil NDVI are not directly correlated to maize yield, their combined use can help determine those soil properties influencing maize production using a limited number of soil samples. This study expanded on previous work delineating SSMUs with a single ECa sensor and provides strong support for the use of multiple-sensor platforms in precision agriculture. Producers, agriculture consultants, extension specialists, and Natural Resource Conservation Service field staff are the beneficiaries of SSMUs.