The spread of Japanese knotweed in Scania: Invasion suitability prediction using species distribution modelling
(2025) In Master Thesis in Geographic Information Science GISM01 20252Dept of Physical Geography and Ecosystem Science
- Abstract
- Invasive species can cause serious damage to ecosystems and built-up structures. “Japanese knotweed” is one of the most problematic invasive plant species in Scania. It has spread in a fast and uncontrollable manner throughout the years, and major efforts are being made to decrease its extension and further spread. Once established, it's difficult to exterminate without causing environmental harm. To locate and map where Japanese knotweed has potential to grow would be of enormous help in the struggle to decrease its impact. Species distribution models (SDM) are useful tools in identifying and illustrating areas vulnerable to invasions of alien species.
The aim of this study was to assess the capability of using an SDM to locate areas... (More) - Invasive species can cause serious damage to ecosystems and built-up structures. “Japanese knotweed” is one of the most problematic invasive plant species in Scania. It has spread in a fast and uncontrollable manner throughout the years, and major efforts are being made to decrease its extension and further spread. Once established, it's difficult to exterminate without causing environmental harm. To locate and map where Japanese knotweed has potential to grow would be of enormous help in the struggle to decrease its impact. Species distribution models (SDM) are useful tools in identifying and illustrating areas vulnerable to invasions of alien species.
The aim of this study was to assess the capability of using an SDM to locate areas where Japanese knotweeds potentially would invade, and if possible, to estimate how areas classified as having high ecological values are or could be threatened by the plant species. Two common and popular models were tested and compared, namely MaxEnt and Random forest.
The model outcomes were based on a sample size of 122 points, where 75% were used for modelling and 25% for evaluation. Both models performed well, MaxEnt (0,87) and Random forest (0,85), based on AUCroc-score. The most contributing variables to the MaxEnt predictions were distance to roads, land cover and elevation, where distance to roads contributed the most. 10% of Scania has a invasion suitability over 50% based on MaxEnt output, while Random forest predicts an extent of 22%. Based on binary conversion of MaxEnt output, 16.6% of the total land area of Scania, and 17% of areas classified with high ecological values is predicted to be at risk of being invaded by Japanese knotweed. (Less) - Popular Abstract
- Invasive species pose a threat to our ecosystems and built-up structures. One of the most problematic invasive species in Scania, Sweden Is Japanese knotweed. It has spread in a fast and uncontrollable manner throughout the years, and major efforts are being made to decrease its extension and further spread.
Locating and mapping areas that are at risk of being invaded by Japanese knotweed would be of enormous help in combating its spread and establishment. Species distribution models (SDMs) are complex mathematical algorithms that can do just that. They often require information on where the studied species have been observed within the study area, along with the environmental conditions that affect the species ability to establish and... (More) - Invasive species pose a threat to our ecosystems and built-up structures. One of the most problematic invasive species in Scania, Sweden Is Japanese knotweed. It has spread in a fast and uncontrollable manner throughout the years, and major efforts are being made to decrease its extension and further spread.
Locating and mapping areas that are at risk of being invaded by Japanese knotweed would be of enormous help in combating its spread and establishment. Species distribution models (SDMs) are complex mathematical algorithms that can do just that. They often require information on where the studied species have been observed within the study area, along with the environmental conditions that affect the species ability to establish and spread.
The aim of this thesis was to evaluate how vulnerable Scania is to being invaded by Japanese knotweed by using two of the most common SDMs (MaxEnt and Random forest) to assess whether areas that are highly ecologically valued are at risk of being invaded and to evaluate the performance of the SDMs.
The coordinates of 122 observations of Japanese knotweed in Scania and 9 environmental variables that, according to the literature, affect the distribution of the species were used when producing the SDMs. 25% of the 122 observations were set aside for the models to test their predictive capabilities.
Both MaxEnt and Random forest performed well in predicting potential invasion sites in Scania. The environmental factor that influences the invasion risk of an area the most is roads, where the shorter the distance to a road, the higher the invasion risk. Land cover was the second most important variable, where urban areas and beaches are predicted to be at particular risk of being invaded. Third was elevation, where the lower the elevation, the higher the invasion risk. Based on MaxENT predictions, over 16% of Scania is suitable to be invaded by Japanese knotweed, as well as 17% of the extent of areas with high ecological value. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9214737
- author
- Järemo Lawin, Agaton LU
- supervisor
- organization
- course
- GISM01 20252
- year
- 2025
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- Geography, GIS, Invasion suitability, Species distribution models, Environmental variables
- publication/series
- Master Thesis in Geographic Information Science
- report number
- 198
- language
- English
- id
- 9214737
- date added to LUP
- 2025-11-03 10:22:55
- date last changed
- 2025-11-03 10:22:55
@misc{9214737,
abstract = {{Invasive species can cause serious damage to ecosystems and built-up structures. “Japanese knotweed” is one of the most problematic invasive plant species in Scania. It has spread in a fast and uncontrollable manner throughout the years, and major efforts are being made to decrease its extension and further spread. Once established, it's difficult to exterminate without causing environmental harm. To locate and map where Japanese knotweed has potential to grow would be of enormous help in the struggle to decrease its impact. Species distribution models (SDM) are useful tools in identifying and illustrating areas vulnerable to invasions of alien species.
The aim of this study was to assess the capability of using an SDM to locate areas where Japanese knotweeds potentially would invade, and if possible, to estimate how areas classified as having high ecological values are or could be threatened by the plant species. Two common and popular models were tested and compared, namely MaxEnt and Random forest.
The model outcomes were based on a sample size of 122 points, where 75% were used for modelling and 25% for evaluation. Both models performed well, MaxEnt (0,87) and Random forest (0,85), based on AUCroc-score. The most contributing variables to the MaxEnt predictions were distance to roads, land cover and elevation, where distance to roads contributed the most. 10% of Scania has a invasion suitability over 50% based on MaxEnt output, while Random forest predicts an extent of 22%. Based on binary conversion of MaxEnt output, 16.6% of the total land area of Scania, and 17% of areas classified with high ecological values is predicted to be at risk of being invaded by Japanese knotweed.}},
author = {{Järemo Lawin, Agaton}},
language = {{eng}},
note = {{Student Paper}},
series = {{Master Thesis in Geographic Information Science}},
title = {{The spread of Japanese knotweed in Scania: Invasion suitability prediction using species distribution modelling}},
year = {{2025}},
}