Some biological systems can be described in terms of changes that take places over the course of “evolutionary time” – thousands or even millions of years. Others take place over a much shorter period. Interventions conducted by humans could be seen as among the most rapid ecological processes. The success or failure of a given intervention could turn on whether a handful of rats finds enough food at the bottom of the embankment on one side of a road, or if they have to cross the road and end up consuming poisoned bait on the other side. Gene drive, a new intervention technique enabled by transgenic constructs with super-Mendelian inheritance ratios, promises to be the new record holder for rapidity of an evolutionary process. In order to simulate such rapid processes, abstractions that have long been relied upon in computational eco-evolutionary models must be abandoned in favor of models that adopt a higher resolution view of populations targeted for intervention, along with predators and competitors thereto. However, as the resolution of such models improves, computational performance can sharply decline. This work enables these more detailed models where previously impractical by introducing techniques to substantially improve their performance. The first chapter of this work explores the use of machine learning combined with sensitivity analysis methods by which computationally burdensome models, such as the rat gene drive model that was developed for this study, can be more efficiently explored in order to develop a deeper understanding of the modelled system. The second chapter of this work introduces a new foundational technique for continuous-space individual-based population models in which the resources exploited by the modeled individuals are directly simulated using an architecture that results in runtime speeds that are much faster than comparable models in many contexts, enabling the simulation of populations that were hitherto impractically large. The third chapter of this work utilizes this resource-explicit framework to addresses just such a population. The model introduced in this third chapter simulates the release of a gene drive into rats within a multispecies model that also includes mice and mustelids, and which plays out over a large portion of the South Island of New Zealand, with a combined census size of over 50 million animals present in the model. Thanks to our resource-explicit technique and an extensive focus on runtime optimization, this model maintains good performance, requiring only a few minutes per time-step.