Abstract
The rapid loss of the world’s tropical rainforest has serious consequences for mankind and global biodiversity. The conservation and sustainable management of tropical rainforests is helped greatly by the longterm storage and predictable availability of germplasm. Cryopreservation is one of the main methods of conservation. However, protocols must be developed quickly since the seeds of many tropical rainforest trees germinate rapidly and are subject to fungal contamination. Their availability is also unpredictable and hence, often, only small samples are available.This thesis investigates the novel applicability of statistically designed experiments that use less raw materials and can be performed quicker than traditional designs. Taguchi approach was explored for the first time, for the optimisation of cryopreservation protocols. These approaches can save much valuable time and scarce resources without losing important information. This thesis investigates, for the first time, if these techniques are suitable when cryopreserving tropical rainforest germplasm.
A variety of single and mixed-level traditional 'full factorial’ experiments are performed to cryopreserve tropical rainforest seed germplasm from orthodox, intermediate and recalcitrant species. Novel statistical analyses including stepwise and binary logistic regression techniques are used in addition to traditional techniques such as analysis of variance and multiple regression are applied to the experimental data. The results of these experiments are compared with carefully chosen subsets (to simulate fractionally replicated experiments) and those from the Taguchi approach.
The cryopreservation experiments in this thesis were applied to four tropical tree species (Cassia siamea, Koompassia malaccensis, Sterculia cordata and Parkia speciosa). The whole seed, embryos (at different developmental stages) and shoot-tips were used with techniques including a desiccation protocol for orthodox and desiccation and vitrification for intermediate seeds. An encapsulation-vitrification method including novel use of trehalose as a cryoprotectant was developed for a recalcitrant seed. Differential scanning calorimetry was used to elucidate the critical points at which vitrification occurred.
The conclusions of the cryopreservation experiments suggest that performing a fraction of the full factorial experiments saved valuable raw material and time. When the conclusions of the Taguchi and ANOVA approaches differed, the former resulted in a more robust optimal protocol. Hence the Taguchi approach to optimising cryopreservation experiments is recommended.
This approach can be extended to many areas of conservation of flora or fauna when optimal conditions of a large number of factors must be found when time and/or resources are scarce. However, these techniques work best when the experimenter has sufficient prior knowledge to identify those two-way interactions to be investigated.
| Date of Award | Sept 2005 |
|---|---|
| Original language | English |