As a challenging high-dimensional multimodal optimization problem, path planning for unmanned combat aerial vehicles (UCAVs) has evolved into a hard optimization problem with multiple objectives and types of constraints in complex operational environments. The traditional approaches lack good search capability in complex multimodal search spaces, resulting in difficulties in providing satisfactory flight paths under multiple constraints. Intelligent optimization algorithms have become the first choice for solving path planning problems because of their excellent global exploration capabilities. Artificial ecosystem optimizer (AEO) is an efficient and intelligent optimization algorithm that has remarkable effectiveness in handling optimization problems. However, it also has the disadvantage of slow convergence. In this work, a modified version of the AEO, named MAEO, is proposed to overcome its shortcomings and apply it to the UCAV path planning problem. In the MAEO algorithm, a new production model that focuses on the current optimal search area is proposed to improve the quality of the producer, thereby effectively guiding consumers to search for the optimal space. Then, an enhanced updating consumption mechanism inspired by animal predation behavior is designed, including two predation operations: one is the introduction of the dynamic elite individual in the original update method to simulate that consumers are also predated by their natural enemies during predation to enhance the convergence speed of the algorithm. The second is a unique spiral predation strategy adopted by the consumers to round up the prey. This mechanism not only effectively avoids the lack of population diversity but also improves the global exploration ability of the algorithm. Furthermore, an adaptive Cauchy mutation strategy is successfully hybridized with the AEO algorithm to boost the ability of the algorithm to escape from the local optimum. Simulation experiments of path planning in a series of complex three-dimensional environments show that the algorithm can plan a path satisfying the constraints stably and efficiently, which proves the superiority of the algorithm.