Modelling spatial exclusion of a vulnerable native fish by introduced trout in rivers using landscape features: a new tool for conservation management

摘要: Habitat variability can affect the ability of invasive predators to exclude native prey species from parts a landscape. If habitat consistently mediates predator impacts, outcome predator–prey interactions could be modelled using data alone. In New Zealand streams, predatory trout create source–sink dynamics in populations fish Galaxias vulgaris. Within trout-invaded sink habitat, G. vulgaris are excluded small, stable streams that occur far galaxiid demographic sources (streams above barriers trout). A GIS-based spatial model stream size and distance-to-source was developed predict where would river network, detect undiscovered trout-free source populations. The maximum distance at which co-occur with two riverscapes quantile limit function GIS. predictive tested training datasets independent catchments occurrence patterns were quantified electrofishing surveys. The correctly predicted absence first test catchment, but failed second dataset because an initial lack knowledge insufficient survey data. This demonstrated exclusion by model, provided it is constructed validated iterative fashion. The current provides useful tool both for conserving critical populations, as well selecting rehabilitation through eradication. Since interaction models based on landscape likely generally guiding conservation management, researchers should investigate whether other biotic exist nature. Copyright © 2011 John Wiley & Sons, Ltd.