
Test CEC (1M NH4OAc, pH 7) per kg clay <= threshold
Source:R/utils-diagnostic-tests.R
test_cec_per_clay.RdDefault threshold is 16 cmol_c/kg clay (WRB 2022 ferralic horizon).
v0.9.69 ECEC fallback (opt-in)
Brazilian / SOTERLAC / BDsolos profiles often record the exchange
complex as separate Ca, Mg, K, Na, Al cmol values without an
explicit "Valor T" CEC column, so cec_cmol is NA for
the entire profile. With
options(soilKey.ferralic_ecec_fallback = TRUE) the test
falls back to the ECEC sum
(ca_cmol + mg_cmol + k_cmol + na_cmol + al_cmol) on layers
where cec_cmol is missing but the components are present.
Default is FALSE (canonical WRB behaviour preserved).
Note: ECEC is typically smaller than CEC at acidic pH because it omits H+; using ECEC against the same threshold is therefore conservative (MORE permissive) – it should not produce false positives, only recover Latossolos that lacked Valor T.
v0.9.86 engine="aqp" auto-enables the ECEC fallback
soilKey.diagnostic_engine = "aqp" now auto-enables the
v0.9.69 ECEC fallback (the user can still suppress it explicitly
by setting soilKey.ferralic_ecec_fallback = FALSE). The
rationale: the aqp engine is the "data-quality-aware" mode,
designed for field-described datasets like BDsolos / Redape
where Valor T is rarely recorded. Bundling these two opt-ins
lifts BDsolos RJ Latossolo recall from 14.9\
28.1\