Dealing with overdispersed count data in applied ecology

ing ecological complexity by identifying key biological processes, and the factors that affect them, is important when solving applied ecological problems (e.g. Hilborn & Mangel 1997; McPherson & DeStefano 2003; Clark 2007), as the effectiveness of a management action is likely to depend upon the processes that dominate a system (Walters 1986). Hence, it is important that ecologists should be able to compare their data reliably with multiple, process-derived, ecological models. It is common for ecologists to choose the factors and processes that best explain their data using stepwise multiple regression approaches; however, biased parameter estimates and inconsistencies among model selection algorithms (Whittingham et al . 2006) have led many ecologists to adopt Correspondence: Shane A. Richards, Department of Biological and Biomedical Sciences, University of Durham, South Road, Durham DH1 3LE, UK (tel: + 44 (0) 191 3341308; fax: + 44 (0) 191 3341201; e-mail: [email protected]). *Correspondence author. E-mail: [email protected]