ζwahlenbergia Schrader
ζWahlenbergia Schrader ζ has 100-150 spp., mainly in the S. Hemisphere. Allan (1961) described 10 indigenous spp. occurring in many open inland habitats. One of the commonest and most widespread of them is said to be W. gracilis (Forster f.) Schrader but there appear to be 2 taxa amongst the many specimens labelled as this sp. despite the great variation which has led some to consider that only a single sp. is involved. This variation appears to be paralleled in Australia. In N.Z., plants from cliffs, rocky ledges and outcrops, seem to represent true W. gracilis, whereas those from grasslands, especially modified lowland pastures along the E. sides of both islands, as well as roadsides and waste places, represent another sp. which may be introduced. Comparable Australian specimens of the latter taxon have been called W. marginata (Thunb.) A. DC., a sp. occurring from Japan to New Caledonia and Australia. However, some authorities consider that W. marginata, which has a Japanese type, does not occur in Australasia. Also, in Australia similar plants are sometimes referred to W. communis Carolin, W. gracilis or even W. stricta Sweet. All these spp. have a confused taxonomic history [Carolin, R. C., Proc. Linn. Soc. New S. Wales 89: 235-240 (1965)]. The situation is further complicated in that the taxon described as W. trichogyna Stearn in Allan (1961) and confined to a coastal area in Marlborough is really W. stricta subsp. stricta (P. Smith, pers. comm.). Because it is widespread in Australia, but only found in a single small area of modified grassland in N.Z., it may also have been introduced; these Marlborough plants have mostly opposite lvs, as well as much larger fls than W. marginata or W. gracilis. In the absence of definitive studies of the taprooted wahlenbergias in N.Z. and the uncertainty over their status, entities which might have been introduced have not been treated fully here. Hatch, E. D., Trans. Roy. Soc. N.Z. 79: 369 (1952), suggested varietal status as W. marginata var. australis for the component of W. gracilis thought to be introduced in N.Z. but raised no doubts about its indigenous status. This is strongly disputed by A. P. Druce (pers. comm.). From W. gracilis sens. strict. it is distinguished by the narrower, linear or linear-lanceolate lvs and usually larger fls which are usually blue but sometimes white or nearly so, as well as its different habitat. A few specimens which are here referred to W. marginata are even more distinct from W. gracilis, e.g., AK 173524, Whakatane County, Motuhora (Whale) Id, Wright, 1.1.1986. The sheet contains a plant with prominent taproot, erect stems (including peduncles) to 40 cm high, and with a branched infl. of many blue fls. Such plants occur in widely separated areas and possibly represent a third taxon.From all other N.Z. spp. of the genus, W. marginata sens. lat. may be distinguished by the following combination of characters: plant taprooted and not rhizomatous; stems slender and usually branched, 10-50 cm long, often with a number of generally deep blue fls; lvs mostly cauline, alternate and narrow, not forming rosettes although plants can appear tufted.
