Behavioural responses of brushtail possums to live trapping and implications for trap-catch correction

Abstract

Context The behavioural response of animals to repeated trapping has implications for correction of population and monitoring indices that use catch per unit effort. Failure to account for sprung traps introduces biases into estimates of relative abundance. The time when animals get caught in live traps is often ignored, but it can provide important information about temporal movement patterns relevant to this issue. Aims We assessed changes in the behaviour of brushtail possums (Trichosurus vulpecula), a nocturnal marsupial, in response to repeated trapping and evaluated the potential benefit of correcting a commonly used index of abundance by using time-of-capture information. Methods Possums were live-trapped for three nights each month over a 20-month period in baited cage traps in a 6-ha area of native lowland forest in the southern North Island, New Zealand. Trapped possums were individually identified on first capture. Timing devices were attached to the traps to measure how long after sunset traps were sprung and how that time related to the duration of the trap-night (sunset to sunrise). Key results Possums were trapped, on average, ~1.25 h after sunset. Traps triggered other than by possums were sprung on average 1–2 h later. Possums caught on the first night of a trapping session were caught significantly earlier than those caught on subsequent nights. Previous capture influenced the time of subsequent capture in a trapping session in complex ways, and recapture times were generally earlier than times of first capture. Possums were captured, on average, after 11% of the duration of a trap-night and traps were triggered by animals other than possums, on average, after 22% of the duration of a trap-night. Conclusions The data on time of capture of possums and triggering of sprung traps suggested a need to alter the commonly used correction factor for population indices for possums, because, on average, traps were sprung for significantly more of each trapping interval (i.e. trap-night) than the half a trap-night assumed in the correction factor. Implications Better understanding of possum foraging behaviour is a key to more effective control using traps. In that context, more research is needed to understand the reasons for individual differences in trappability. Although it is theoretically desirable to account for sprung traps when trapping is used to index populations, to reduce biases in estimates of relative abundance, correction of the standardised residual trap-catch index for possums is probably unimportant in practical terms, because most possum control reduces numbers to levels (2–5% trap catch) at which correction of the index is unimportant. The principal exception to this is likely to be when there is a high level of non-target interference from rodents.