Papua New Guinea has major ex situ field collections of plant genetic material in its major staple food crops (aibika, banana, cassava, sago, sweet potato, taro, yams). The PNG Government has become concerned at the cost of maintaining these collections. With limited germplasm conservation resources available, difficult choices must be made as to which plants to maintain. What resources should be devoted to maintaining plant genetic diversity in the wild or in collections. How should these resources be allocated among various plant kinds, especially when some of them are currently important in agricultural production, whereas others may only be of potential future importance. How should resources be allocated across the various methods of conserving plant germplasm? Should genetic collections only be maintained where the material has been collected or should it also be stored in other countries. A dynamic optimisation model is developed of the crop plant improvement process, including selection from the wild and farmers' fields, conventional plant breeding, the use of advanced plant breeding techniques, and the contribution of plant germplasm collections in the efficiency of this process. The objective of the study is to provide a better basis for evaluating the efficient allocation of resources to plant germplasm conservation in food staples in PNG.