This article explores ways in which virtual environments can be used for crowdsourcing and behavior mining for filling gaps within the information space of topical research. Behavior mining in this article refers to the act of harvesting the latent or instinctive behavior of participants, usually a crowd, and injecting the population behavior into a preset context, such as within a virtual environment so that the subjective behaviors and the contexts are merged. The experimental approach combines various modalities centered upon virtual environments so as to induce presence in order to bring participants into the context. This approach is new and not well studied; however, it has real potential in research dealing with behaviors and culture in reconstructed virtual environments. Two virtual environments case studies at the 2012 and 2015 Royal Society Summer Science Exhibition are presented, which demonstrate that the unique crowdsourcing activity is able to fill gaps within the information space so that answers to research questions can be more complete. Thus, by reconstructing and replicating a lost landscape, and by injecting harvested human behavior into the context of the landscape, we may be able to gather much more information than conventional methods will allow.