From experimental zoology to big data: Observation and integration in the study of animal development.

The founding of the Journal of Experimental Zoology in 1904 was inspired by a widespread turn toward experimental biology in the 19th century. The founding editors sought to promote experimental, laboratory-based approaches, particularly in developmental biology. This agenda raised key practical and epistemological questions about how and where to study development: Does the environment matter? How do we know that a cell or embryo isolated to facilitate observation reveals normal developmental processes? How can we integrate descriptive and experimental data? R.G. Harrison, the journal's first editor, grappled with these questions in justifying his use of cell culture to study neural patterning. Others confronted them in different contexts: for example, F.B. Sumner insisted on the primacy of fieldwork in his studies on adaptation, but also performed breeding experiments using wild-collected animals. The work of Harrison, Sumner, and other early contributors exemplified both the power of new techniques, and the meticulous explanation of practice and epistemology that was marshaled to promote experimental approaches. A century later, experimentation is widely viewed as the standard way to study development; yet at the same time, cutting-edge "big data" projects are essentially descriptive, closer to natural history than to the approaches championed by Harrison et al. Thus, the original questions about how and where we can best learn about development are still with us. Examining their history can inform current efforts to incorporate data from experiment and description, lab and field, and a broad range of organisms and disciplines, into an integrated understanding of animal development.