Notes

The species are difficult Some of the reasons for this are:—Environment-induced morphological variability. Individuals exposed to fire and capable of resprouting from a woody base often differ markedly from shrubby plants of the same taxon sheltered from fire.—The occurrence of characters which are constant in some but variable in other taxa. Notably the arrangement of the leaves—e.g., decussate or in whorls of 3 or 4 in A. ternnatum (variable even within populations), or either strictly decussate or strictly in whorls of 3 (e.g., A. welwitschii vs. A. whyteanum). For this reason (and also in view of the variable sex distributions within a taxon: see below) taxa should be studied with care in the field and representative samples should be collected.—The occasional occurrence of flowers ‘transitional’ between hermaphrodite and ‘pure’ female. Such ‘transitional’ flowers have corollas intermediate in size between hermaphrodite and female and are characterized by the presence of small but clearly discernible, (±) pollenless anther rudiments; ‘pure’ female with small corollas have no rudimentary anthers. In some essentially dioecious taxa, male flowers do occasionally revert to hermaphrodite; in some it is only odd hermaphrodite flowers on a single male plant of a population, in others a considerable number of individuals may show this phenomenon. Also the reverse (transitions from hermaphrodite to male) is possible (Note: the flowers are always protandrous; hermaphrodite flowers at anthesis—i.e., with stigmas which have not yet fully elongated—should, therefore, not be confused with such transitional flowers). In both transitions male? hermaphrodite and hermaphrodite ?male there are, however, no marked corolla size and shape differences.—Sex dimorphism in dioecious taxa. Differences between the sexes may go beyond corolla size and shape differences; dimorphic inflorescences are rather common (contracted and often quite conspicuous in female); occasionally male and female differ somewhat in leaf size and shape. It is recommended that both male and femlae are collected whenever possible.Hybridization represents an additional problem. Experiments carried out in the greenhouse proved that (not even so closely allied) species of Anthospermum can be crossed and form viable hybrids. Thus, virtually anywhere where two species happen to grow sympatrically and flower at the same time, hybrids can be expected to occur. Especially in the Eastern Highlands of Zimbabwe (and adjacent parts of Mozambique), where no less than 7 taxa (out of a total of 10 in Flora Zambesiaca area) may occur together, the frequency of putative hybrids appears to be quite high. Because of the variability of numerous taxa and the presence of only few ‘good’ and useable distinguishing characters between taxa, definite proof for the hybrid nature of an individual or an entire population is not easily obtained. One can, thus, not always be certain whether an ‘atypical’ specimen still belongs to one taxon (i.e. represents a geno- and/or phenotypic ‘extreme’) or is of hybrid origin.