New Zealand Journal of Ecology (1996) 20(1): 37- 43

A small predator removal experiment to protect North Island weka (Gallirallus australis greyi) and the case for single- subject approaches in determining agents of decline

Research Article
Gary N. Bramley 1,2
  1. Department of Ecology, Massey University, Private Bag 11-222, Palmerston North, New Zealand
  2. Present address: Department of Biological Sciences, University of Waikato, Private Bag 3105, Hamilton, New Zealand

The hypothesis that predation on eggs and chicks by ferrets (Mustela furo) and cats (Felis catus) was limiting the productivity of North Island weka (Gallirallus australis greyi), was tested by removing predators from the home ranges of four breeding pairs of weka. Reproduction by four other breeding pairs was monitored to provide a control. I was not able to follow the breeding success of some weka because they died or removed their radio transmitters. Two of the pairs breeding in predator-removal areas reared five chicks to independence, while two control pairs reared no chicks to independence after three breeding attempts. It was not possible to draw solid conclusions from group comparison data gained from so few individuals. Because it is necessary to identify the factors preventing weka recovery now, I suggest an alternative way to gather reliable data about rare species like weka. Single subject experimental designs like those developed in the social sciences offer an alternative route for investigating agents of decline in rare species; such a design would have been preferable to the one used in the present study. A common experimental procedure in single subject studies is the A-B-A series or some variant of it where A and B refer to experimental conditions with A being the control and B the treatment. Treatments are switched on and off for individuals (e.g., animals, or areas) which are considered representative of the study population. Well designed, replicated single subject studies might allow data to be used even after unplanned alterations to experimental conditions. The results of such studies would be cumulative, interpretable and more readily publishable. For some species all that may be required is a reanalysis of existing data.