Advanced

Time to establishment success for introduced signal crayfish in Sweden - a statistical evaluation when success is partially known

Sahlin, Ullrika LU ; Smith, Henrik LU ; Edsman, Lennart and Bengtsson, Göran LU (2010) In Journal of Applied Ecology 47(5). p.1044-1052
Abstract
The signal crayfish Pacifastacus leniusculus is an invasive species in Sweden, threatening the red-listed nobel crayfish Astacus astacus through spreading the crayfish plague. Time-to-event models can handle censored data on such introduced populations for which the state (successful or not) is only partially known at the last observation, but even though data on introduced populations most often are censored, this type of model is usually not used for likelihood-based inference and predictions of the dynamics of establishing populations. 2. We specified and fitted a probabilistic time-to-event model to be used to predict the time to successful establishment of signal crayfish populations introduced into Sweden. Important covariates of... (More)
The signal crayfish Pacifastacus leniusculus is an invasive species in Sweden, threatening the red-listed nobel crayfish Astacus astacus through spreading the crayfish plague. Time-to-event models can handle censored data on such introduced populations for which the state (successful or not) is only partially known at the last observation, but even though data on introduced populations most often are censored, this type of model is usually not used for likelihood-based inference and predictions of the dynamics of establishing populations. 2. We specified and fitted a probabilistic time-to-event model to be used to predict the time to successful establishment of signal crayfish populations introduced into Sweden. Important covariates of establishment success were found by the methods of 'model averaging' and 'hierarchical partitioning', considering model uncertainty and multi-colinearity, respectively. 3. The hazard function that received the highest evidence based on the empirical data showed that the chances of establishment were highest in the time periods immediately following the first introduction. The model predicts establishment success to be < 50% within 5 years after first introduction over the current distributional range of signal crayfish in Sweden today. 4. Among covariates related to temperature, fish species and physical properties of the habitat, the length of the growing season was the most important and consistent covariate of establishment success. We found that establishment success of signal crayfish is expected to increase with the number of days when growth is possible, and decrease with the number of days with extremely high temperatures, which can be seen to approximate conditions of stress. 5. Synthesis and applications. The results demonstrate lower establishment success of signal crayfish further north in Sweden, which may decrease the incentives of additional illegal introductions that may threaten the red-listed noble crayfish Astacus astacus. We provide a fully probabilistic statistical evaluation that quantifies uncertainty in the duration of the establishment stage that is useful for management decisions of invasive species. The combination of model averaging and hierarchical partitioning provides a comprehensive method to address multi-colinearity common to retrospective data on establishment success of invasive species. (Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
risk analysis, probabilistic, model averaging, likelihood-based method, partitioning, hierarchical, hazard function, censored data, failure-time model, quantitative assessment
in
Journal of Applied Ecology
volume
47
issue
5
pages
1044 - 1052
publisher
Wiley-Blackwell
external identifiers
  • wos:000281286800010
  • scopus:77956105941
ISSN
1365-2664
DOI
10.1111/j.1365-2664.2010.01849.x
project
BECC
language
English
LU publication?
yes
id
01f299ca-ee4f-4b7c-9455-f6ce52b70917 (old id 1673126)
date added to LUP
2010-09-23 12:19:32
date last changed
2018-05-29 09:40:05
@article{01f299ca-ee4f-4b7c-9455-f6ce52b70917,
  abstract     = {The signal crayfish Pacifastacus leniusculus is an invasive species in Sweden, threatening the red-listed nobel crayfish Astacus astacus through spreading the crayfish plague. Time-to-event models can handle censored data on such introduced populations for which the state (successful or not) is only partially known at the last observation, but even though data on introduced populations most often are censored, this type of model is usually not used for likelihood-based inference and predictions of the dynamics of establishing populations. 2. We specified and fitted a probabilistic time-to-event model to be used to predict the time to successful establishment of signal crayfish populations introduced into Sweden. Important covariates of establishment success were found by the methods of 'model averaging' and 'hierarchical partitioning', considering model uncertainty and multi-colinearity, respectively. 3. The hazard function that received the highest evidence based on the empirical data showed that the chances of establishment were highest in the time periods immediately following the first introduction. The model predicts establishment success to be &lt; 50% within 5 years after first introduction over the current distributional range of signal crayfish in Sweden today. 4. Among covariates related to temperature, fish species and physical properties of the habitat, the length of the growing season was the most important and consistent covariate of establishment success. We found that establishment success of signal crayfish is expected to increase with the number of days when growth is possible, and decrease with the number of days with extremely high temperatures, which can be seen to approximate conditions of stress. 5. Synthesis and applications. The results demonstrate lower establishment success of signal crayfish further north in Sweden, which may decrease the incentives of additional illegal introductions that may threaten the red-listed noble crayfish Astacus astacus. We provide a fully probabilistic statistical evaluation that quantifies uncertainty in the duration of the establishment stage that is useful for management decisions of invasive species. The combination of model averaging and hierarchical partitioning provides a comprehensive method to address multi-colinearity common to retrospective data on establishment success of invasive species.},
  author       = {Sahlin, Ullrika and Smith, Henrik and Edsman, Lennart and Bengtsson, Göran},
  issn         = {1365-2664},
  keyword      = {risk analysis,probabilistic,model averaging,likelihood-based method,partitioning,hierarchical,hazard function,censored data,failure-time model,quantitative assessment},
  language     = {eng},
  number       = {5},
  pages        = {1044--1052},
  publisher    = {Wiley-Blackwell},
  series       = {Journal of Applied Ecology},
  title        = {Time to establishment success for introduced signal crayfish in Sweden - a statistical evaluation when success is partially known},
  url          = {http://dx.doi.org/10.1111/j.1365-2664.2010.01849.x},
  volume       = {47},
  year         = {2010},
}