Science is a social endeavor that progresses through the concerted effort of many individuals. With the surge in the complexity of the most pressing open problems in science and society, the role of collaboration networks among researchers is becoming more and more relevant.
Despite the importance of the structure of such networks, we still do not have a comprehensive picture of the most relevant factors affecting it. Here, we suggest that the research environment (for example, institutional arrangements or culture) may leave measurable structural fingerprints in collaboration networks. We assess this hypothesis by comparing the structure of collaboration networks between prominent researchers in North America and Europe, and we find systematic differences that are robust across research fields. We find that prominent researchers in North America and Europe fulfill different structural roles in field-specific collaboration networks, with Europeans establishing much denser collaboration networks with each other, and North Americans establishing more decentralized networks. We also observe collaboration-dependent differences in the impact of the publications of each community. Specifically, prominent researchers in North America are significantly more impactful than those in Europe when they do not collaborate with other prominent researchers. When prominent researchers collaborate with each other, which Europeans do more often, impact differences between North America and Europe often disappear. Our results suggest that funding schemes and other science-related policies may have long-term, systemic impacts on how researchers establish collaborations, and therefore on how they develop new ideas and share them with others.