Various working memory (WM) models based on a diverse set of biologically plausible synaptic and neural plasticity mechanisms have been proposed. We show that these proposed short-term plasticity mechanisms may not necessarily be competing explanations, but instead yield interesting functional interactions on a wide set of WM tasks and enhance the biological plausibility of spiking neural network models of WM. While reductionist models (WM explained by one particular mechanism) are theoretically appealing and have increased our understanding of specific mechanisms, they are insufficient. WM models need to become more capable, robust and flexible to account for new experimental evidence of bursty and activity-silent multi-item maintenance in more challenging WM tasks. We believe that spiking models should address known electrophysiological constraints from recordings and generally solve more than one task. We evaluate the interactions between three commonly used classes of plasticity, namely intrinsic excitability, synaptic facilitation/augmentation, and Hebbian plasticity. Combinations of these are systematically tested in a spiking neural network model on a broad suite of tasks deemed principally important for WM function, such as one-shot encoding, free and cued recall, active delay maintenance after fast task item selection and and updating of that task set. We compare the performance and biological plausibility of a robust, integrated model against other model variants with fewer plasticity mechanisms. Our results indicate synergies between commonly proposed plasticity mechanisms for WM function and show that more reductionist models fail to achieve comparable performance in some tasks due to the principle nature of different types of plasticity. When parameter tuning can improve their performance, this typically comes at the price of reduced biological plausibility or task generality.