Statistical Considerations for Monitoring and Sampling [Version 1]. [SUPERSEDED by DOI: http://dx.doi.org/10.25607/OBP-918]

Abstract

This chapter will not follow the usual presentation for statistical design in ecology. Rather, we will focus on what we believe to be most important aspects from a practical (and management) viewpoint. We do not intend it to be like a ‘text-book’ and explicitly do not include formulae or descriptions of tangential details. Readers will want to look elsewhere for such detail (Urquhart and Kincaid, 1999; Gitzen et al., 2012; Thompson, 2012, are a good start, although there are many). We hope to only introduce the relevant concepts and stress that these are the things that should be thought about by all researchers involved with survey planning. In particular, we discuss: (i) randomisation, (ii) efficiency of design, (iii) uncertainty reduction, (iv) sampling in space and time, and (v) specifics for different gear types. This all leads to an illustrative example design, using the MBHdesign R-package. (available from CRAN, https://cran.r-project.org/package=MBHdesign). For those readers interested in acronyms: MBHdesignstands for Marine Biodiversity Hub design. The goals and techniques implemented in MBHdesign are outlined throughout this chapter.