Advanced biological intelligence learns efficiently from an in-formation-rich stream of stimulus information, even when feed-back on behaviour quality is sparse or absent. Such learning ex-ploits implicit assumptions about task domains. We refer to such learning as Domain-Adapted Learning (DAL). In contrast, AI learning algorithms rely on explicit externally provided measures of behaviour quality to acquire fit behaviour. This im-poses an information bottleneck that precludes learning from di-verse non-reward stimulus information, limiting learning effi-ciency. We consider the question of how biological evolution circumvents this bottleneck to produce DAL. We propose that species first evolve the ability to learn from reward signals, providing inefficient (bottlenecked) but broad adaptivity. From there, integration of non-reward information into the learning process can proceed via gradual accumulation of biases induced by such information on specific task domains. This scenario pro-vides a biologically plausible pathway towards bottleneck-free, domain-adapted learning. Focusing on the second phase of this scenario, we set up a population of NNs with reward-driven learning modelled as Reinforcement Learning (A2C), and allow evolution to improve learning efficiency by integrating non-re-ward information into the learning process using a neuromodu-latory update mechanism. On a navigation task in continuous 2D space, evolved DAL agents show a 300-fold increase in learning speed compared to pure RL agents. Evolution is found to elimi-nate reliance on reward information altogether, allowing DAL agents to learn from non-reward information exclusively, using local neuromodulation-based connection weight updates only. Code available at github.com/aislab/dal.