Fauna in decline: First do no harm

See allHide authors and affiliations

Science  22 Aug 2014:
Vol. 345, Issue 6199, pp. 884
DOI: 10.1126/science.345.6199.884-b

In their Review “Reversing defaunation: Restoring species in a changing world” (special section on Vanishing Fauna, 25 July, p. 406), P. J. Seddon et al. warn that loss of animal species can disrupt ecological communities, cause cascading effects, and alter ecosystem functions. Introduced nonnative animals can have similar consequences.

Burgeoning evidence implicates nonnative species as driving biodiversity loss (13) and a host of other ecological disruptions (4). Whereas some can have positive effects on ecosystem services, others have disproportionately large negative effects. Risk assessment of these outcomes is undermined by context-dependence and time lags (4, 5). An introduced species that has negligible effects in some areas, or whose population is threatened in its native range, can have strong impacts when translocated elsewhere (6, 7). Such species may appear innocuous for decades—well beyond the attention span of monitoring programs—before suddenly becoming problematic (8). Moreover, their impacts may be subtle, but nonetheless great, and remain unrecognized until damage is incurred and containment is impossible (9). Even carefully planned introductions for conservation purposes can have devastating consequences (10, 11).

These considerations are largely ignored by Seddon et al. in their discussion of assisted colonization and ecological replacements—deliberate introductions of species beyond their native range. Although Seddon et al. reassuringly cite new approaches (quantitative risk analysis, active adaptive management, and structured decision-making) for managing what could go wrong, none of the cited references offer reliable methods for predicting impacts of nonnative animal releases. Despite making considerable progress in understanding impact (5), invasion science has not developed a predictive capacity sufficient to engage in frequent introductions without harming biodiversity and ecosystems (7). Thus, risks of unintended effects cannot be evaluated and weighed against expected benefits.

At best, assisted colonization is analogous to other human interventions (such as geoengineering) that are prone to unpredictable consequences and do not address root causes of the problems they are supposed to mitigate (7, 12). Ironically, in an earlier article on using nonnative species for conservation purposes, Seddon et al. (13) rightly conclude that “the concern is not the failure to establish the intended ecological interactions, but rather the risk of creating new and unwanted interactions.” Perhaps what is needed is a Hippocratic oath (“Do no harm”) applicable to conservation biologists.


Navigate This Article