Sample design pertains to the choice of where and when to sample. In contrast, experimental design pertains to the selection and arrangement of treatments (Section 4.7). Sample design choices are made to describe the ecosystem with levels of accuracy and precision so that knowledge may be gained. Hypotheses may be developed and tested regarding patterns or responses to specific impacts (Underwood 1991, Osenberg and Schmitt 1996). Alternatively, confidence intervals may be established to estimate effects and guide decision-making (Stewart-Oaten 1996). Pilot studies are employed to inform and guide sampling design decisions.