Modelling the germination of Melia Volkensii using an A-, D-, T- Optimal Four Factor Rotatable Central Composite Design

Abstract

Melia volkensii is a tree species endemic to the arid and semi-arid areas of Eastern Africa. Its natural range falls within areas which are characterized by dry bush land and wooded grassland. It is a fast growing species with the growth tremendously faster on farm than in the wild, suggesting remarkable potential gains through domestication. Despite the potential of the species, the tree is yet to be massively propagated for plantation establishment especially by farmers due to difficulties experienced in seed extraction, germination and propagation through cuttings when compared to other species. The objectives of the study were to construct an A-, D-, T- optimal four factor rotatable central composite design, to develop models for the germination of Melia volkensii and to determine optimal conditions for the germination of Melia volkensii. The experiments were conducted at Kenya Forestry Research Institute (KEFRI) laboratories in Muguga, Kiambu County. An A-, D-, T- optimal four factor rotatable central composite design was constructed from the general central composite design by determining the optimal weights satisfying the A-, D- and T- optimality criteria. Response surface methodology techniques were used to develop second order models for the germination of Melia volkensii as well as to analyze the associated response surfaces. The variables under investigation were soil pH, temperature, chemical concentration and length of time of seed pre-treatment. Comparisons were made on the use of four different chemicals for seed pre-treatment. These were Potassium Nitrate (KNO 3 ), Hydrogen Peroxide (H 2 O 2 ), Gibberellic Acid (GA 3 ) and Sulphuric Acid (H 2 SO 4 ). The experiment was performed by soaking 20 seeds of Melia volkensii in a chemical solution for a specified period of time. The seeds were then placed in a petri-dish containing soil of a particular pH. They were then placed in germination chambers of a defined temperature. The outcome was the number of seeds that germinated in a particular petri-dish. We established that in general, germination rates of Melia volkensii seeds were low. For the four chemicals used in the experiment the germination rates were found to be 31.67% for KNO 3 , 39.08% for H 2 O 2 , 42.00% for GA 3 and 28.25% for H 2 SO 4 . The overall germination rate was found to be 35%. However when the conditions were favorable and set correctly germination rate was optimized at between 57% and 76%. Temperature and soil pH were found to be the most significant factors across the models. The optimum temperature ranged between 26.77 o C and 31.13 o C while the optimum soil pH was found to be between 3.95 and 5.52. To maximize germination rates, we recommend the soaking of Melia volkensii seeds for 8 hours in a 0.03% solution of GA 3 before planting them in soil of pH 5.5 at a constant temperature of 27 o C. The results of the study will profoundly contribute towards large scale adoption and availability of seedlings for Melia volkensii thereby transforming the landscape in the arid and semi-arid lands and in the long term changing their climate.