An allostratigraphical unit is a body of sedimentary rock that is defined and identified only by its bounding discontinuities.  Although the formal units allomember, alloformation and allogroup were explained in the North American Stratigraphic Code, in practice, allostratigraphy has beem less used than other systems, although it has been used in the USA (e.g. Morrison 1985). Allostratigraphic units may be defined to distinguish between superposed deposits (a succession of lithologically similar alluvial and lacustrine deposits, separated by soil horizons), contiguous discontinuity-bounded deposits (onlapping lobes of soliflucted material from different cold-climate periods or glaciations), and geographically separated bodies (river terrace deposits), all of similar lithology; or to distinguish as single units discontinuity-bounded bodies characterised by lithological heterogeneity (cf. also Morphostratigraphy).  The principles of naming these units are the same as for lithostratigraphical units.  The bounding discontinuities are described in the North American Stratigraphic Code as unconformities, disconformities or the present-day land surface.  Fault boundaries are excluded.

While sequence stratigraphy was also developed to classify unconformity-bounded units, allostratigraphy covers the special case of rock bodies of similar lithology separated by discontinuities.  The International Stratigraphic Code, however, made no such distinction and recommended an hierarchy of formal terms based on the fundamental unit synthem for all unconformity-bounded units – i.e. it did not use alloformation, etc.

The most easily recognised allostratigraphical units are river terrace deposits.  They rest unconformably on bedrock; their upper surfaces are the current land surface, and the deposits of the different terraces may be lithologically indistinguishable from each other.  That they are temporally separate can normally be established by their spatial relationships and altitude.

On geological maps terrace deposits are frequently classified in terms of their form and origin. Individual terrace surfaces are identified by numbers, the first terrace being the lowest and youngest.  In well studied areas, numbering may be replaced by named deposits and surfaces, following international practice.  There is a full succession of named terrace surfaces and underlying deposits along the River Thames, England, for example.  These names are not always the same; thus, the Harefield Terrace is developed on the Gerrards’ Cross Gravel deposit. There is disagreement among various authors about the rank of the Thames terrace deposits, but all have been defined as lithostratigraphical units; allostratigraphical classification and nomenclature have not been adopted (cf. Gibbard 1985).

Alluvial and lacustrine deposits in a rift environment, in which stacked sediment units of similar lithology are separated by disconformities, are allostratigraphical units.  Note that these are the same as depositional sequences in sequence stratigraphy, or subsynthems in the classification of unconformity-bounded units given in the International Stratigraphic Code.

It is a requirement in lithostratigraphy adopted in both the North American and International stratigraphical codes for a succession of sediments of uniform composition to constitute a single lithostratigraphical unit, even if there is an unconformity in the middle. That approach is questioned here. Unconformities and disconformities are legitimately used to define bounding surfaces of lithostratigraphical units.  Thus there may only be a need for an allostratigraphical scheme when it is impossible to distinguish the deposits above and below a discontinuity by any method.  However, Quaternary researchers have no difficulty ascribing an unique identity to compositionally similar terrace gravel deposits using altitude and consider that this is sufficient justification to give individual names (or numbers) to deposits of the same composition. Disconformity-bounded units, separated by palaeosols would also be regarded, strictly, as allostratigraphical units, but have been classified successfully as lithostratigraphical ones.

There appears to be a consensus that if superposed or contiguous deposits of similar composition can be distinguished from each other by any method, it is justifiable to classify them as separately named lithostratigraphical units, a practice which is endorsed in this guide.

*This guide is based on that produced by Rawson et al. (2002) for the Stratigraphy Commission of the Geological Society of London.