Thesis (M. Sc.)--University of Canterbury, 2007. / Typescript (photocopy). Includes bibliographical references (leaves 65-75). Also available via the World Wide Web.

Thesis (Ph. D.)--University of Canterbury, 2009. / Typescript (photocopy). Includes bibliographical references (leaves 126-141). Also available via the World Wide Web.

The impact of invasive predators on native prey may depend on the availability and distribution of invader-free refugia across landscapes, if predators create demographic ‘sink’ populations in invaded patches, giving rise to source-sink dynamics in prey populations. Propagule pressure of immigrants dispersing from refugia (or sources) may consequently drive persistence in sink habitat, affecting predator-prey co-existence across the landscape. I studied whether introduced brown (Salmo trutta) and rainbow trout (Oncorhynchus mykiss) create source-sink structuring in two native galaxiid fish species (Galaxias vulgaris and G. paucispondylus) in the rivers of the central South Island, New Zealand, and whether such dynamics affected the distributions of either species across river networks or “riverscapes”. Young-of-the-year (YOY) G. vulgaris recruitment was rare in trout-invaded streams and consistently high in trout-free streams. Thus, trout-free reaches appeared to act as sources in a river network, while the majority of the trout-invaded riverscape was a demographic sink (i.e., no local recruitment occurred). Surveys of YOY G. paucispondylus did not reveal trout-induced source-sink dynamics, although mesocosm predation experiments suggested both species were highly vulnerable to predation by large trout. Galaxias paucispondylus recruitment was highest in intermittently flowing streams that were marginal habitats for trout, suggesting indirect interactions between trout and habitat affect G. paucispondylus distribution. Network configuration of trout-free source populations affected the distribution of G. vulgaris, as galaxiids were excluded from small streams with high bed stability that were far from sources. The interaction between propagule pressure and habitat gradients in mediating effects of trout on G. vulgaris distributions indicates habitat characteristics affect predator-prey interactions in a spatially explicit manner. Furthermore, the outcome of predator-prey interactions should be able to be modelled using habitat data alone if habitat consistently mediates predator impacts. I developed a GIS-based spatial model to predict where trout would exclude G. vulgaris in river networks, based on stream size and distance to galaxiid source populations. The model was tested in three different riverscapes using fish occurrence patterns obtained from electrofishing surveys, and successfully predicted G. vulgaris exclusion by trout. This further demonstrates the importance of habitat configuration in driving interspecific interactions at the landscape scale. These findings suggest removing trout from small, stable tributaries to create new demographic sources could improve overall persistence of G. vulgaris across trout-invaded riverscapes. The galaxiid exclusion model should also be used to detect undiscovered trout-free source populations, and to aid in selecting streams for restoration of galaxiid populations through trout eradication.

The impact of invasive predators on native prey may depend on the availability and distribution of invader-free refugia across landscapes, if predators create demographic ‘sink’ populations in invaded patches, giving rise to source-sink dynamics in prey populations. Propagule pressure of immigrants dispersing from refugia (or sources) may consequently drive persistence in sink habitat, affecting predator-prey co-existence across the landscape. I studied whether introduced brown (Salmo trutta) and rainbow trout (Oncorhynchus mykiss) create source-sink structuring in two native galaxiid fish species (Galaxias vulgaris and G. paucispondylus) in the rivers of the central South Island, New Zealand, and whether such dynamics affected the distributions of either species across river networks or “riverscapes”. Young-of-the-year (YOY) G. vulgaris recruitment was rare in trout-invaded streams and consistently high in trout-free streams. Thus, trout-free reaches appeared to act as sources in a river network, while the majority of the trout-invaded riverscape was a demographic sink (i.e., no local recruitment occurred). Surveys of YOY G. paucispondylus did not reveal trout-induced source-sink dynamics, although mesocosm predation experiments suggested both species were highly vulnerable to predation by large trout. Galaxias paucispondylus recruitment was highest in intermittently flowing streams that were marginal habitats for trout, suggesting indirect interactions between trout and habitat affect G. paucispondylus distribution. Network configuration of trout-free source populations affected the distribution of G. vulgaris, as galaxiids were excluded from small streams with high bed stability that were far from sources. The interaction between propagule pressure and habitat gradients in mediating effects of trout on G. vulgaris distributions indicates habitat characteristics affect predator-prey interactions in a spatially explicit manner. Furthermore, the outcome of predator-prey interactions should be able to be modelled using habitat data alone if habitat consistently mediates predator impacts. I developed a GIS-based spatial model to predict where trout would exclude G. vulgaris in river networks, based on stream size and distance to galaxiid source populations. The model was tested in three different riverscapes using fish occurrence patterns obtained from electrofishing surveys, and successfully predicted G. vulgaris exclusion by trout. This further demonstrates the importance of habitat configuration in driving interspecific interactions at the landscape scale. These findings suggest removing trout from small, stable tributaries to create new demographic sources could improve overall persistence of G. vulgaris across trout-invaded riverscapes. The galaxiid exclusion model should also be used to detect undiscovered trout-free source populations, and to aid in selecting streams for restoration of galaxiid populations through trout eradication.