An interesting report [pdf] from the University of Minnesota Libraries looks at the behaviors of researchers in the sciences. It extends the earlier work done by the Libraries on researchers in the humanities and social sciences.
Not unsurprisingly there is a major focus on having resources available online and when online reducing the number of clicks required to use something. There is also considerable discussion about research data issues, covering the need for better ways of organizing and managing data outputs.
Scientists make heavy and regular use of library resources availably electronically, but regard the physical library buildings as a place of last resort -- where you go when you have no other way to find something. Library buildings are places of “disclosure” rather than “discovery,” inasmuch as researchers go to libraries to retrieve what they have already identified. At the same time, many scientists speak nostalgically about the lost art of browsing and serendipitous discovery in libraries and depend on technology to provide browsing proxies. [Understanding Research Behaviors, Information Resources, and Service Needs of Scientists and Graduate Students: A Study by the University of Minnesota Libraries]
The findings appear consistent with the RIN report on researchers' use of libraries [pdf] which was released a short while ago, although it ranges over wider ground.