Co-authorship networks can provide key insights into the production of scientific knowledge. This is particularly interesting in Antarctica, where most human activity relates to scientific research. Bibliometric studies of Antarctic science have provided a useful understanding of international and interdisciplinary collaboration, yet most research has focused on broad-scale analyses over recent time periods. Here, we take advantage of a ‘Goldilocks’ opportunity in the McMurdo Dry Valleys, an internationally important region of Antarctica and the largest ice-free region on the continent. The McMurdo Dry Valleys have attracted continuous and diverse scientific activity since 1958. It is a geographically confined region with limited access, making it possible to evaluate the influence of specific events and individuals. We trace the history of environmental science in this region using bibliometrics and social network analysis. Our results show a marked shift in focus from the geosciences to the biosciences, which mirrors wider trends in the history of science. Collaboration among individuals and academic disciplines increased through time, and the most productive scientists in the network are also the most interdisciplinary. Patterns of collaboration among disciplines resemble the biogeochemical relationships among respective landscape features, raising interesting questions about the role of the material environment in the development of scientific networks in the region, and the dynamic interaction with socio-cultural and political factors. Our focused, historical approach adds nuance to broad-scale bibliometric studies and could be applied to understand the dynamics of scientific research in other regions of Antarctica and elsewhere.