首页 | 本学科首页   官方微博 | 高级检索  
   检索      


Predictions of establishment risk highlight biosurveillance priorities for invasive fish in New Zealand lakes
Authors:John R Leathwick  Kevin J Collier  Brendan J Hicks  Nicholas Ling  Glen Stichbury  Mary de Winton
Institution:1. Independent researcher, Hamilton, New Zealand;2. Environmental Research Institute, Faculty of Science and Engineering, School of Science, The University of Waikato, Hamilton, New Zealand;3. National Institute of Water and Atmospheric Research Ltd, Hamilton, New Zealand
Abstract:
  1. The ability to predict invasive species spread is essential for effective biosecurity management and the allocation of scarce monitoring resources. Prevention of invasive fish incursions poses a significant challenge because of the wide physiological tolerances of many species, their mobility and the role that human vectors play in their spread. In New Zealand, seven introduced fish species are distributed to varying extents in lakes across the two main islands. We used field survey data from 470 New Zealand lakes to fit statistical models of the current geographic distributions of seven introduced species; the resulting models were then used to predict risks of future establishment of each species in 3595 New Zealand lakes >1 ha.
  2. Initial models fitted using lake‐ and catchment‐scale environmental predictors identified summer temperature among the top two most influential variables, with lake density and size also important for some species. Distribution models for Eurasian perch (Perca fluviatilis), rudd (Scardinius erythrophthalmus) and tench (Tinca tinca) were substantially improved by the addition of variables describing human population densities and lake accessibility. All seven species occurred most frequently in lakes close to human population centres suggesting that human‐mediated dispersal has played at least some role in determining current distributions.
  3. Addition of a spatial variable, representing the presence or absence of the modelled species within the broader catchment within which each lake is located, improved the predictive performance of models for the brown bullhead catfish (Ameiurus nebulosus), perch and rudd. This finding indicates that the current distributions of these species include clusters of lakes within ‘occupied’ catchments, resulting in geographic patchiness that is independent of the available environmental and human population predictors. This distribution has most likely resulted from spread into accessible and suitable lakes from one or more initial liberation points, either by natural dispersal along waterways or through human‐assisted movement.
  4. Predictions to all mapped lakes throughout New Zealand indicate (i) that the potential for future spread is greatest for catfish, perch and rudd and (ii) the high vulnerability to invasion for lakes along the east coast of both islands and in inland montane regions of the South Island. Our results allow for improved identification of lakes likely to be suitable for invasive fish species and which should therefore be accorded priority for surveillance; they highlight in particular the potential for perch and catfish to establish in higher‐elevation lakes distant from human population centres.
Keywords:biosecurity  fish distribution  invasion  pest species  regression model
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号