Table 3.

Spatial terms, estimates of autocorrelations, and variance components for spatially dependent (${\mathit{\sigma }}_{\mathit{\xi }}^2$) and independent residuals (${\mathit{\sigma }}_e^2$) from the best standard spatial (BSS) models fitted to grain yield data in two example trials

Trial	BSS modela	${\mathit{\rho }}_r$	${\mathit{\rho }}_c$	${\mathit{\sigma }}_{\mathit{\xi }}^2$	${\mathit{\sigma }}_e^2$
DYS05	R + Spl(r) + AR1xAR1 + n	0.87	0.67	0.103	0.064
DAB08	R + AR1xAR1 + n	0.24	0.96	0.201	0.611

^aR: random row effects; Spl(r): cubic smoothing spline indexed by row positions; AR1xAR1: correlated residual modelled as two-dimensional first-order autoregressive process; n: spatially independent residual (or nugget variance)