Taking the escalator to extinction: Understanding the role of rapid forest expansion in driving hotspots of extinction risk for endemic species in Taiwan


Despite an abundance of data quantifying and predicting the response of forest distribution to ongoing climatic changes in temperate and boreal regions, there is a near absence of data, and hence little understanding, of how tropical mountain systems will respond to climate change(1). While we understand that forests are often expanding rapidly at high elevation as temperatures warm, little attention has been given to the implications of such changes for plants above the forest limit (2,3). Data frequently demonstrate upward migration of such alpine species also, however, with highly limited habitat availability at high altitude, such upward migration is dubbed the ‘elevator to extinction’ since once at the highest elevations, such alpine species have no-where left to go (4). Data on high elevation extinction risk are, however, vanishingly rare. This significant knowledge gap has major implications for our ability to predict future impacts of environmental change on biodiversity and ecosystem services (5).

Previous work by the supervisory team in the region has identified rapid shifts in the altitudinal distribution of alpine plant species driven directly by rising temperature and through displacement due to elevation of the mountain treeline (4). However, expansion of forest at the treeline is highly heterogeneous; expanding rapidly in some areas but remaining static in others (6). While elevational shifts of forest risk alpine plant extinction, spatial heterogeneity in advance allows a mechanism for continued coexistence (7).

This project will combine existing knowledge on the pattern and limitations of forest advance at high elevations in Taiwan with data on the distribution of endemic alpine plants. By integrating remote sensing of habitat type and distribution with modelling of the spatial and temporal patterns of forest and alpine plant migration, we’ll identify hotspots of extinction risk for high elevation endemic species to inform conservation prioritisation across the region.

The project will be guided by the following objectives
1) To characterise and quantify the spatial extent of the principal high elevation habitat types in Taiwan’s Central Mountain Range using high-resolution satellite imagery.
2) To use existing survey data to assess the current distribution of alpine endemic plant species across the region and identify their topographical and habitat associations.
3) To combine survey data, climate predictions, topography and habitat availability to model future distributions of high elevation trees and alpine plants.
4) To estimate future changes in forest distribution, carbon sequestration potential and grassland area loss in the Central Mountain Region.

Click on an image to expand

Image Captions

Fig1_Jump2019.jpg: Barberry (Berberis sp. growing amongst prostrate Yushan cane (Yushania niitakayamensis) in Taiwan’s Central Mountains (With permission from Erin Stoll).
Fig2_Jump2019.jpg: Anaphalis nepalensis (With permission from Erin Stoll).
Fig3_Jump2019.jpg: Gentiana scarbrida in the high elevation grasslands in Taiwan (With permission from Erin Stoll).


High resolution remote sensing (RS) imagery would be used to derive land use cover classifications to enable estimation of habitat extent and distribution. RS classifications would be ground-truthed via field visits with established research partners (Dr Jan-Chang Chen, National Pingtung University of Science and Technology) in the region. We would then integrate scenario-based modelled estimates of forest change built using climate predictions, topographical information and land-use cover to predict changes in alpine plant habitat extent. Existing and modelled future distributions of endemic alpine species built using distributional data held by Taiwan Endemic Species Research Institute will then be derived and compared with habitat availability using GIS tools to assess vulnerability and forecast hotspots of extinction risk. The research would involve two periods of fieldwork surveying habitat type and distribution in the Central Mountain Range and working with ESRI on distribution and risk forecasts for regional conservation planning. All local work in Taiwan is conducted through longstanding agreements with collaborators at national Pingtung University of Science and Technology, Taiwan.

Project Timeline

Year 1

Year 1: Project preparation, literature research, RS training and land use classification, preliminary forest change modelling.

Year 2

Year 2: Ground-truthing fieldwork in Taiwan and visit to ESRI, finalise forest change modelling, preliminary climate-distribution modelling of alpine plant species. Present work at national conference.

Year 3

Year 3: Field and planning work with ESRI and additional habitat surveys. Finalise species distribution modelling and build composite extinction risk maps. Present work at international conference.

Year 3.5

Years 3 and 3.5, finalise writing of papers and thesis.

& Skills

This studentship will equip the graduate with a range of highly desirable skills from technical skills in remote sensing and geographical information systems through field survey to statistical modelling and spatial analysis skills, the majority of which are in high demand in academic and industry employment opportunities. Development of expert knowledge in the R language will be encouraged and facilitated by the supervisory team. The successful candidate will be able to access postgraduate training opportunities at the University of Stirling and Durham University and will be part of two active research groups with highly active research in this and related areas. The student will gain from the opportunity to spend a significant amount of time working at Durham University for individual training in remote sensing techniques and analysis.

References & further reading

1) Moser et al 2011 Global Change Biology 17, 221
2) Greenwood and Jump 2014. Arctic, Antarctic, and Alpine Research, 46, 829-840
3) Morley et al 2019 Remote Sensing of Environment, 223, 291-306
4) Jump et al 2012 Ecography 35, 204-210
5) Bonan 2008 Science 320, 1444
6) Huang 2002 International Journal of Remote Sensing 23, 2051-2069
7) Greenwood et al 2015 Journal of Vegetation, Science 26, 711-721

Further information on the Stirling research group and projects related to this one can be found here: https://biogeo.org/wp/

Further Information

Apply Now