Choosing sowing strategies for containerized seedling operations

Abstract

Choosing a container sowing strategy was formulated as a linear programming problem. The independent variables are fractions of containers seeded to one, two, and three seeds. An optimal sowing strategy is achieved by choosing, subject to certain constraints, the three fractions to minimize an economic penalty function, which penalizes a sowing strategy if it does not deal effectively with problems caused by blank cells. Two options were considered for producing a required number of seedlings with a given number of containers. With option 1 strategies, the three fractions of containers must be chosen so that the predicted number of plantable seedlings cannot be less than the required number, but no provision is made for replanting blank cells. Option 2 strategies have the same constraint that the predicted number of blank cells cannot exceed the predicted number of excess seedlings. Numerical analyses were conducted with hypothetical container sowing problems using estimated costs for procuring and sowing seeds, for carrying or replanting blank cells, and for thinning excess seedlings. Linear programming methods were used to choose a sowing strategy that minimized a penalty function. Mixed sowing strategies, as opposed to the standard strategy of sowing two seeds per cell, were generally optimal for a wide range of values for the germination and survival rate of seeds. A user-oriented, interactive computer program was developed for the nursery manager who wishes to use these methods to choose a sowing strategy.