Appendix A
Laboratory Methods
The profile samples were analysed in the Institut für Geoökologie, Universität Potsdam, from whom full details of the methods are available. Some granulometric analyses were done at the University of Mainz.
Two general points about laboratory methodology affect the characterisation and correlation of BCI soils:
Firstly, the granulometric analysis of fine textured kaolinitic and oxidic soils is problematic. The oxidic minerals have strong bonding effects, and much of the clay in such soils is aggregated into pseudo-silt or pseudo-sand particles. These give the soil high macroporosity and free drainage typical of medium or coarse textured soils. However, the aggregates are not dense and relatively inert primary minerals, as in true sands and silts. They have high internal microporosity and the voids are lined with clay surfaces that have low or moderate electrochemical activities. The result is that these soils behave like sands or silts with respect to drainage, aeration and moisture release at low moisture tensions, but like clays with respect to cation exchange and other chemical sorption/desorption attributes, and also for moisture retention at high tensions.
The micro-aggregation also affects the determination of clay content. If the soils are treated with a mild dispersant, as normally used in soils with non-oxidic mineralogies, the aggregates remain intact and the soils analyse as high in silt and/or sand. If dispersed fiercely, the aggregates break up and the soils analyse as clays. Reliance on only one method of dispersion gives an incomplete picture of these soils, with mild dispersion missing the clay characteristics and fierce dispersion missing the aggregation and silt- and sand-like features. We have therefore estimated the clay contents of some samples with both sodium pyrophosphate (mild) and sodium dithionite (strong) dispersants and indirectly from moisture retention at tensions above the conventional permanent wilting point (1500kPa) (Soil Survey Staff, 1999). The indirect moisture retention estimates of clay for the same soils are in the range 60 – 90%. The dithionite dispersion gives intermediate estimates, with clay contents in the range 40 – 60%. These are similar to the hand textures in the field. All of the laboratory granulometric methods make heavy demands on resources, and the number of determinations is limited. In contrast, we have several thousands of field textures, many of them by multiple observers. Although only qualitative, hand texturing is in some ways superior to laboratory methods, because it reveal both aspects of micro-aggregated 1:1 clays, with initial impressions of high silt diminishing after continued kneading to give a more clay-rich texture. Such clays are designated as light, in contrast to heavy clays, which have substantial proportions of 2:1 minerals and do not initially feel silty. Our soil characterisations and correlations are based on the field textures
The second point concerns the determination of cation exchange capacity (CEC). The CEC’s of kaolinitic and oxidic clay minerals are variable and are usually low in their natural proton-rich acid environments. The conventional laboratory determination of CEC is done at pH 7. Deprotonation of cation exchange sites occurs at the artificially high pH, and conventional CEC values are considerably higher than the actual working CEC in field conditions. This problem has long been recognised (e.g. Hesse, 1971; Pleysier et al. 1979). There are other problems with CEC, and they are sometimes considerably lower than the sum of the exchangeable bases and appear to be substantially underestimated. This is not a problem that is unique to BCI and has been noted in somewhat similar soils in Mexico
These problems are partially circumvented by the use of the effective CEC (ECEC). We here use the term in the ST sense, i.e. ECEC = the sum of the exchangeable bases (TEB) + extractable Al (Soil Survey Staff, 1999), rather than the WRB definition of ECEC = TEB + exchangeable acidity (FAO, 1998), although the two are probably similar in BCI soils. TEB is relatively unaffected by pH, and labile Al is extracted at close to field pH, so ECEC is a less variable and more realistic characterisation of the exchange complex in acid soils in field conditions.
ST and WRB differentiate some of their taxa on base saturation (BS = TEB/CEC). The methodological problems with CEC lead to potential anomalies in BS. In order to better characterise base status, we estimate Effective BS (EBS = TEB/ECEC). EBS is the complement of Al saturation, and indicates the balance on the exchange complex between nutrient bases and toxic Al.