Conservation of select South African Disa Berg. species (Orchidaceae) through in vitro seed germination.
Disa comprises 163 species, 131 of which occur in South Africa (SA). The genus is distributed across winter- and summer-rainfall areas, but few species transverse both climatic regions. Species are therefore regarded as winter-rainfall or summer-rainfall endemics - yet release their seeds in autumn, irrespective of provenance. Disa contributes 40 % of threatened Orchidaceae in SA, with half of the local species requiring conservation initiatives. In vitro seed germination is a potential conservation tool for producing large numbers of genetically diverse plants in relatively short periods. However, only 11 winter-rainfall Disa species are easily germinated ex situ. Studies were therefore undertaken on summer-rainfall taxa, which are ungerminated in vitro, in an effort to define their germination parameters. This thesis describes mechanisms that control germination in Disa and establishes practical propagation methods for seed culture. Two seed types occur in Disa; i) comparatively large, pale and pyriform seeds in members of the D. uniflora sub-c1ade, which populate streamside habitats under conditions of winter-rainfall maxima, and ii) smaller, variously brown and fusiform seeds in the remainder of the genus. Seed morphometrics distinguished seed types, although embryo dimensions were similar. Testa continuity, which is disrupted in the large seeds, also supported separation. Typically, small seeds are ungerminated in vitro, whilst large seeds germinated readily. Increased seed size did not necessarily impart increased germ inability, as several germinable, small-seeded species exist - being winter-rainfall species Attempts to establish in vitro germinability revealed that increased water availability and charcoal supplementation promoted germination in intractable species. The control of germination was therefore proposed as a trade-off between water availability and the presence of phyto-inhibitors - two features typical of seeds exhibiting water-impermeable dormancy. Three germinability categories were recognized; i) easily germinable species, ii) poorly germinable species through media manipulation, and iii) ungerminated species. Germination of immature seed in the absence of media modification was comparable to mature seed germination under modified conditions, providing evidence of the role of an impermeable seed testa in regulating germination. Testa impermeability in mature, small-seeded species was demonstrated using aqueous EVANS' blue dye and was linked to i) testa integrity and ii) increased levels of leachable phenolics (LPC) - which are hydrophobic and phytotoxic. In addition, this research revealed an impervious and elaborate embryo carapace in small seeds. Large-seeded species were highly permeable at dehiscence, with perforated testae and negligible LPC. Germinability was ultimately defined by a significant regression with LPC. Phenolic deposition increased exponentially with increasing seed maturity and reflected decreased permeability and the development of testa colouration. The testa precludes the use of viability stains such as nc and FDA, unless rendered permeable through scarification. This was achieved using NaOCI. Viability and germinability percentages did not correlate well for the small-seeded Disa species, indicating that i) the methods used to break dormancy are inadequate, ii) additional factors may be acting in concert with the testa to regulate germination and iii) that the determination of mature Disa seed viability is ineffective. As an alternative, the germination potential of immature seed was estimated as the ratio between the proportion of embryos stained with TTC and the proportion of seeds permeable to EVANS' blue. Attempts to relieve water-impermeable dormancy in Disa resulted in the formulation of a dual-phase protocol - with the specific aim of increasing water availability to the embryo. Dual-phase cultures comprised a solid, charcoal-rich medium overlaid with a reduced strength, liquid medium fraction of the same type. The solid fraction negated the influence of leached phenols and allowed protocorms to establish polarity, whilst the fluid fraction increased water availability. The dual-phase protocol allowed germination of nine summer-rainfall Disa species, usually in percentages that approximated their estimated germination potential. For the remaining species, germination is controlled by more complex factors. Large seeds are atypical in containing starch, the hydrolysis of which facilitated their rapid, autonomous germination. Small-seeded Disa species stored lipids and proteins and germinable species accumulated starch post-germination. The embryo protoplasts of all species contained appreciable amounts of soluble sugars, irrespective of germinability. However, decreased sucrose and increased fructose correlated significantly with decreased seed germinability. This study provides evidence of the nutritional value of mycotrophy, with endophytes liberating soluble carbohydrate and non-carbohydrate compounds upon lysis. However, few species were germinated symbiotically, suggesting that endophytes isolated from adult roots do not necessarily support germination in the same species. Similar endophytic fungi occur in Australian and Holarctic orchids.