Ecological Indicators 16 (2012) 76–83 Contents lists available at ScienceDirect Ecological Indicators jo ur n al homep ag e: www.elsevier.com/locate/ecolind Uncertainty and natural variability in the ecological footprint of fisheries: A case study of reduction fisheries for meal and oil Robert W.R. Parker ∗ , Peter H. Tyedmers School for Resource and Environmental Studies, Dalhousie University, 6100 University Avenue, Suite 5010, Halifax, Nova Scotia, Canada B3H 3J5 a r t i c l e i n f o Keywords: Ecological footprint Uncertainty Fisheries Aquaculture Reduction fisheries a b s t r a c t It is well understood that measurements of ecological footprint and many other ecological indicators are associated with varying degrees of uncertainty, yet imprecision in ecological footprint results is rarely assessed or communicated. We calculated the marine portion of the ecological footprint of prod- ucts derived from five reduction fisheries: Peruvian anchovy (Engraulis ringens), Atlantic herring (Clupea harengus), Gulf menhaden (Brevoortia patronus), blue whiting (Micromesistius poutassou) and Antarctic krill (Euphausia superba). Monte Carlo analysis was used to measure the imprecision in marine footprint measurements resulting from multiple sources of uncertainty and natural variability in input parame- ters, and to determine the degree to which imprecision affects our ability to draw meaningful conclusions when comparing products sourced from different fisheries on the basis of ecological footprint. Gulf men- haden and Antarctic krill were found to have the smallest marine footprints, while blue whiting was found to have the largest. Results show that there is much uncertainty associated with marine foot- print calculations and that the most significant drivers of this imprecision are uncertainty and natural variability regarding measurements of trophic level and trophic interactions. Marine footprint is highly correlated with trophic level, and clear differences can be seen when comparing species of very different trophic levels. However, comparisons of products derived from species’ with similar trophic levels are less likely to provide conclusive results. The choice of mass, protein or energy content as the basis of comparison was also considered and was found to influence the results, particularly when comparing species with similar trophic levels. While it is likely that imprecision of marine footprint measurements of fishery-derived products will remain high, technological improvements and a better understanding of marine ecosystem dynamics may make future studies more precise. © 2011 Elsevier Ltd. All rights reserved. 1. Introduction In the context of modern environmental concerns, there is grow- ing interest in the ability to understand and successfully measure the degree to which human beings are placing demands upon the resources and services of the ecosphere (Millennium Ecosystem Assessment, 2005; Pollard et al., 2010; Butchart et al., 2010). Measuring environmental burden and improving the environmen- tal performance of human activities and products requires close inspection of the tools with which we quantify and communicate environmental impact and guide decision makers. The ecological footprint (EF) (Rees, 1992; Rees and Wackernagel, 1994) is a rep- resentation of the land and sea area required to sustain human populations and human activities and the degree to which demand on ecological resources and services fits within, or overshoots, the capacity of the earth to provide them. It has been widely applied ∗ Corresponding author. Tel.: +1 902 476 7317. E-mail address: rob.parker@dal.ca (R.W.R. Parker). to inform individuals, governments, businesses and others of the pressure their activities place on the capacity of natural systems to provide resources and assimilate wastes (Wackernagel and Rees, 1996; Wackernagel et al., 1999; Talberth et al., 2006; Huijbregts et al., 2008; Ewing et al., 2010; Pollard et al., 2010). Measurements of EF are typically communicated in absolute val- ues of land and sea area, and are obtained using calculations which incorporate absolute values of input parameters (Wackernagel et al., 1999; Talberth et al., 2006; Huijbregts et al., 2008; Ewing et al., 2010). The importance of natural variability and uncertainty in influencing the actual value of input variables, as well as the accuracy with which they are translated into spatial reflections of environmental burden, is recognized (GFN, 2009), though not commonly assessed and communicated in EF studies. To date, the sources and influence of uncertainty and natural variability for many elements of the EF have not been formally addressed. Uncertainty can be broadly defined as “any departure from the unachievable ideal of complete determinism” (Walker et al., 2003, p. 8). This definition infers that sources of uncertainty include any forces which inhibit our ability to produce single, precise and 1470-160X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.ecolind.2011.06.015