two codes of nomenclature governing plants.“Informal”classification refers to
organizational and naming systems that do not conform to one of the codes.
A number of theorists of plant classification have espoused the view that clas-
sification of crop-wild complexes, in which there is at least some interbreeding, is
preferably carried out informally. There are endless definitions of“species,”no
universally accepted criterion or criteria for this fundamental grouping, and con-
siderable heterogeneity in the nature of groups that are called species. Nevertheless,
the ability to interbreed and the actual degree to which interbreeding occurs are
critical considerations in recognizing species of plants, because gene exchange
among populations tends to eliminate the differences that are employed to define
species. The so-called“biological species concept”defines species on the basis of
actual or potential breeding separateness (and clearly on this basis there is only one
species ofCannabis). Above the biological species level, evolution is largely
bifurcating (although there is debate about the degree to which hybridization among
groups at the genus level and above has occurred), a pattern which is compatible
with the hierarchical structure of conventional plant taxonomy. However, some
systematists (e.g. Minelli 1993 ; Pickersgill et al. 2003 ) have concluded that variants
below the biological species level (often classified as subspecies and varieties) are
usually not generated in a hierarchical fashion, either in nature or in cultivation, and
so using more than one infraspecific rank for crop-weed complexes, as has been
commonly done in an attempt to reflect evolutionary patterns, is usually unjustified.
Harlan and de Wet ( 1971 ), frustrated with the inconsistent treatment of crops and
their closely related wild relatives, proposed a non-formal system of classification,
which is in fact an elaboration of the biological species concept (Spooner et al.
2003 ). Their so-called“gene pool classification”recognizes: (a) a“primary gene-
pool,”based on the crop and wild populations (whether or not recognized as
different species) that interbreed readily with it (Harlan and de Wet characterized
their primary gene pool as equivalent to the traditional biological species concept);
(b) a“secondary genepool,”made up of populations that can interbreed with the
crop but only with some difficulty; and (c) a“tertiary genepool,”made up of
populations that can interbreed with the crop but only with considerable difficulty.
Harlan and de Wet further proposed a scheme of hierarchical subpartitioning using
non-formal categories (i.e. independent of the codes of nomenclature). No one has
succeeded in hybridizingC. sativawith any other species in the Cannabaceae, and
all plants ofCannabisinterbreed freely, so classification ofCannabisaccording to
Harlan and de Wet’s concept is simple: all plants belong to the primary genepool of
the one biological species,C. sativa.
Jeffrey ( 1968 ), consistent with his view that“cultivated plants differ from one
another so greatly in their variation patterns that a formal system applicable to all is
not only impossible but undesirable,”recommended a non-formal system of clas-
sification with a maximum of two hierarchical categories to classify cultivars, and
proposed a new term (“subspecioid”) to separate the domesticated from the related
wild-growing plants. Other schemes have been advanced to treat crop classification
1 Classification ofCannabis sativaL. in Relation... 49