Mechanisms underlying the failure of an attempt to eradicate the invasive Asian musk shrew Suncus murinus from an island nature reserve

The Asian musk shrew Suncus murinus is an invasive predator that has had a considerable impact on biodiversity. An eradication was attempted from Ile aux Aigrettes, a 25-ha island reserve off the coast of Mauritius as part of a wider restoration program. A total of 759 shrews were removed using up to 1112 live-traps, in six trapping sweeps of the island. After the first four sweeps, in which 705 shrews were caught, complete eradication appeared to have been achieved, with only four shrews caught in the next three months. However, thereafter shrew captures increased and after six months shrews were again being captured in large numbers. We examine three hypotheses for the failure of the eradication program: (i) low efficiency of all traps ('low efficiency' hypothesis); (ii) low trap efficiency in certain locations ('refuge' hypothesis); and (iii) presence of trap-shy individuals ('trap-shy' hypothesis). We constructed an individual-based simulation of the first four trapping sweeps to predict expected patterns of shrew capture for different shrew densities and capture probabilities and in the presence of trapping 'refugia'. The observed patterns of capture were not consistent with the expected patterns of capture under the 'low efficiency' and 'refuge' hypotheses, but were intuitively consistent with the 'trap-shy' hypothesis. We use the results from this study to help plan future eradication programs for S. murinus. © 2005 Elsevier Ltd. All rights reserved.