Geoscience Reference
In-Depth Information
mean square (MS) error at the FICS was calculated and compared with MS error derived from
blocking in the plot-scale experiment. Experimental errors were similar, indicating that field-scale
EC
a
-classiied variability effectively estimated soil heterogeneity partitioned by plot-scale blocking.
These findings were corroborated at a second and disparate site in central California, an irrigated
system with saline soils (Johnson et al., 2005).
Comparisons were also made between EC
a
-delineated estimates of experimental error and
experimental error derived from replication using multiple soil properties measured at the FICS.
Again, within-field variance was an effective measure of experimental error for most of the nineteen
parameters evaluated. These findings indicate that, for some experimental objectives, within-field
variance may serve as a surrogate for replication.
In locations where soil factors contributing to EC
a
are also yield limiting, EC
a
productivity
zones can be used to design and place plot-scale experiments. They can also be used to design and
evaluate field-scale experiments, functioning as an alternative to replication and blocking. This is
appropriate because EC
a
productivity zones are related to outcome (crop yield) differences expected
in the absence of treatments, the rationale for blocking.
18.4 ConClUSIonS
Although soil heterogeneity complicates the identification and implementation of sustainable
management practices, the ability to delineate this variability offers unique opportunities for both
production agriculture and research. In this dryland experiment, EC
a
effectively mapped soil het-
erogeneity, depicting a gradient of within-field productivity useful for delineating zones of similar
production potential. Practical applications for EC
a
-based productivity zones include zone sampling
and site-specific nutrient management. Zone sampling is superior to traditional random sampling
because it reduces standard error and decreases the number of samples required to evaluate a field.
These advantages are particularly significant in semiarid regions where intensive grid sampling is
cost prohibitive in the predominately large, low-input, dryland farms (McCann et al., 1996).
In high-rainfall regions where both drought and excessive precipitation are yield limiting,
inconsistent relationships have been found between EC
a
and yield across years (Jaynes et al., 1993;
Kitchen et al., 1999). Yet, in semiarid and arid regions yield reductions from excessive precipitation
are rare, which may make EC
a
a more reliable predictor of yield. Consistency of spatial patterns
(Sudduth et al., 2000; Veris Technologies, 2001) and yield relationships (Johnson et al., 2003b;
Corwin et al., 2003b) across years reinforce EC
a
productivity zones as a useful basis for site-specific
management in these regions.
Continued field-scale research is required for EC
a
productivity-zone-based site-specific nutrient
management to determine optimal fertilizer rates for each zone, evaluate soil and crop ecological
response, and compare economic return to that of uniform fertilizer management. It may be pos-
sible to “fine-tune” productivity zones by using different classification methods or depths of EC
a
measurement, or by integrating yield, topographical, or soil color maps. Other factors affecting farm
economics and crop production that are potentially delineated by EC
a
productivity zones should also
be investigated, including grain quality and pest populations (weed, insect, and disease agents).
Beyond practical utility for production agriculture, EC
a
productivity zones have important
applications in agronomic research. As farmers increasingly move toward management of the soil
resource in space and time, new options are needed for experimental design and evaluation if agro-
nomic research is to address relevant issues at relevant levels of scale. Research at the FICS indi-
cates that EC
a
productivity zones offer such an option.
First, EC
a
-productivity zones serve as a pivotal point of reference, a means to integrate and
compare data collected at different biological, spatial, and temporal scales. The framework pro-
vided by these zones can be used to monitor management-induced trends in soil quality at the field
scale (Corwin et al., 2005). It allows the linkage of microbial-scale findings to farm-scale economic
and ecological outcomes in intact agroecosystems. This may advance farm management as a tool
