4. Baseline reference levels

Baseline reference levels are derived from time periods or locations free from human pressures. We use the term baseline inclusive of the structure and function of an ecosystem (1) prior to substantial human impact (i.e., during some ‘baseline’ time period (Karr 1992, Pauly 1995), (2) inside of areas protected from human impacts (Castilla 1999, Dayton et al. 2000), and (3) in remote geographic locations subject to minimal human pressures (Knowlton and Jackson 2008). Recognition of these types of reference levels is crucial for avoiding the shifting baselines syndrome—failing to identify the state of nature absent human impacts so that it is impossible to determine the extent of degradation (Pauly 1995). As such, there is value in reconstructing time series of both desired and undesired changes in indicators, such as shifts in the abundances of iconic and nuisance species. It can also be quite useful to make comparisons across spatial locations that vary in the extent to which they have been altered by human activities (Pitcher 2001). Even where detailed information is not available, the qualitative difference between present and historic, or disturbed and undisturbed, values of ecosystem indicators can provide a reasonable starting point for determining target reference directions (Figure 12; Jackson et al. 2001).

Figure 12. Comparison of a simplified historical and present-day Puget Sound marine food web. Larger, bold font indicates great erabundance/biomass. This figure is intended to be a conceptual schematic, and is not based on historical data. Historical and present-day could be replaced with unexploited and exploited areas or remote and metropolitan locations.

Figure 12

Historical information can be gleaned from a variety of sources, including paleo-ecological records (Aronson et al. 2005) archaeological findings (Butler and Campbell 2004), historical documents (Jackson 1997, Levin and Dufault 2010), and long-term ecological data (Lichatowich 2001, Nichols 2003). Additionally, interviews with people who have experience with an ecosystem during different eras of human impact can provide valuable insights into changes in ecosystem indicators over time (Lozano-Montes et al. 2008, Sàenz-Arroyo et al. 2005). Indeed, subjective impressions of how indicators have varied through time can be standardized with known values and used to establish reference levels (e.g., unfished biomass of currently harvested species; Ainsworth, C.H., T.J. Pitcher, and C. Rotinsulu. 2008. Evidence of fishery depletions and shifting cognitive baselines in Eastern Indonesia. Biological Conservation 141(3):848-859.). One concern with using historical baselines, however, is that ecosystem dynamics are not necessarily stationary. Climatic shifts and other sources of variation can render historic states unattainable (Jennings and Dulvy 2005). Such fundamental changes must be appreciated before making the decision to associate an indicator with a target reference level derived from a historical baseline.

Marine protected areas (MPAs) and areas with low human impact provide useful experiments for evaluating the natural biophysical state of an ecosystem absent major, direct anthropogenic influences (Rice and Rochet 2005). Such spatial baseline ecosystems make particularly useful reference levels because they represent one extreme in a spectrum of management possibilities in the contemporary time period. Admittedly, problems exist with these approaches. For instance, geographic variability among reference and impacted sites and anthropogenic activities that manifest effects on regional and even global scales (e.g., climate change) can confound comparisons. Nonetheless, differences between indicators inside and outside of MPAs (Lester et al. 2009, Stewart et al. 2009) and near to and far from locations with high human population densities (Dulvy et al. 2004, Friedlander and DeMartini 2002, Sandin et al. 2008, Stallings 2009) can provide a useful basis for calibrating expectations regarding the healthy state of an ecosystem (Knowlton and Jackson 2008, Collins and Sheikh 2002, McClanahan et al. 2007).

In Puget Sound, many untapped sources of baseline information exist. For example, archival papers document changes in the abundances of harvested species dating back to at least the 19th century (Hammond 1886). According to these accounts, species declines appear to have occurred long ago, and quite rapidly: “[f]rom 1869 to 1877 it was not an uncommon occurrence for us to catch from 200 to 300 barrels of herring in a night, but since 1877... the largest night’s work is about 20 barrels” (Hammond 1886). Similarly, historical habitats have been altered drastically: <20% of tidal marshes present in the mid-19th century exist today (Collins and Sheikh 2005). Even shorter intervals reveal surprisingly large changes in ecosystem status: current concentrations of polybrominated diphenyl ethers (PBDEs) in southern resident killer whales dwarf the levels detected 10 years earlier (Shaw et al. 2009). In modern times, spatial differences in the ecological communities within and outside of marine reserves near Edmonds, Hood Canal, and the San Juan Islands suggest the direct negative impacts of fisheries on rockfishes and lingcod (Palsson et al. 2004, Tuya et al. 2000). Similarly, comparison of the most populated areas of Puget Sound to more rural areas reveals dramatic differences in the abundance of kelp (Collins et al. 2002, Thom and Hallum 1991).

In terms of actually setting target and benchmark reference levels using information about baselines, the ultimate decision lies in the hands of policymakers (Campbell et al. 2009). Following on the example of the change analysis conducted for Puget Sound’s tidal marshes, the question remains as to what target reference level is most appropriate given that >80% of the historic habitat has been destroyed since 1850. There is no single and absolutely correct answer to this question. It is up for negotiation among stakeholders, but the knowledge of what existed historically and/or what is currently observed in remote or protected locations provides an idea of what is possible.