Ecological research on drought has progressed sporadically, hampered by the many challenges of working on what are often unpredictable events. Past research has tended to be opportunistic, often beset by confounding gradients or lacking adequate controls or pre-impact data.
Controlled manipulative experiments are required to overcome these confounding influences and to identify the mechanistic basis of cause-and-effect relationships that cannot be determined using field surveys. At the other extreme, laboratory-based flume or microcosm studies inevitably sacrifice realism for control and replication, undermining the validity of extrapolating to relevant spatiotemporal scales and levels of biological complexity needed to understand system-level responses.
We are using mesocosm experiments to circumvent the shortcomings of correlational studies whilst maintaining greater realism than would be possible in smaller scale (e.g. laboratory flumes or microcosms) manipulations.