OptOrch: a modular optimization toolkit for forest tree seed orchard deployment

Abstract

OptOrch is a transparent and modular optimization toolkit for forest tree seed orchard (SO) deployment. It implements SO-specific optimal contribution models in AMPL (A Mathematical Programing Language), separating biological assumptions, input data, and numerical solvers. Unlike existing optimal-contribution or mate-allocation software, OptOrch exposes the objective function and constraints directly, allowing breeders to modify SO-specific requirements, including status-number thresholds, proportional contribution bounds, graft availability, pairwise coancestry restrictions, female and male gametic contributions, and external pollen flow. When the declared formulation is supported by the selected solver, solutions can be evaluated using solver-reported feasibility status, objective bounds, and optimality gaps. The algebraic framework also allows alternative formulations to be tested when biological constraints create mixed-integer, nonlinear, or nonconvex instances. We demonstrate OptOrch with simulation-based scenarios involving alternative status-number thresholds and pollen contamination. The framework quantifies gain–diversity trade-offs, evaluates external pollen effects, and compares feasible deployment strategies under realistic biological and operational constraints.