Spatial Transferability of Vegetation Types in Distribution Models Based on Sample Surveys from an Alpine Region

Aune-Lundberg, Linda; Bryn, Anders

doi:10.4236/jgis.2018.101005

Cited by 3 publications

(9 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our results show that the prevalence of the modelled target is an important predictor of model performance; in particular, our results demonstrate that models for rare VTs with restricted distribution obtain higher AUC values than models for common and widespread VTs (Figure a). Similar results are obtained in previous studies of distribution models for species (Evangelista et al., ; Franklin, Wejnert, Hathaway, Rochester, & Fisher, ; Wollan, Bakkestuen, Kauserud, Gulden, & Halvorsen, ) as well as for VTs (Aune‐Lundberg & Bryn, ; Pottier et al., ; Ullerud et al., ). It should be noted that our study is based on presence and absence data from a representative area frame survey.…”

Section: Discussionsupporting

confidence: 89%

“…() applied the MaxEnt method to a smaller spatial extent and evaluated their models with respect to local transferability to an ecologically similar neighbouring area. Ecological similarity between the areas in which training and evaluation data are collected is important for the outcome of DM evaluation, for species (Elith, Kearney, & Phillips, ) as well as for VTs (Aune‐Lundberg & Bryn, ). We argue that use of unbiased and area‐representative datasets, both for training and evaluation, which provides maximum spatial dispersion of presence/absence data points across the total extent of the study (Strand, ), is likely to provide more realistic evaluation statistics than most alternative approaches to evaluation (i.e., data‐splitting, data re‐substitution; Halvorsen, ).…”

Section: Discussionmentioning

confidence: 99%

“…From the full classification scheme of AR18X18 (see Appendix ), a total of 31 VTs were selected for DM as follows: Firstly, all VTs represented by less than 40 polygons in the set of 1081 plots were excluded (van Proosdij, Sosef, Wieringa, & Raes, ; : Appendix ). Thereafter, ecologically very closely related VTs, for example “ Moss snowbed ” and “ Sedge and grass snowbed ”, were merged into one composite VT based on the assumption that these basic VTs were inseparable by means of available environmental data (see Appendix ) (Aune‐Lundberg & Bryn, ; Hemsing & Bryn, ; Ullerud et al., ).…”

Section: Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Distribution modelling of vegetation types based on area frame survey data

Horvath

Halvorsen

Stordal

et al. 2019

Applied Vegetation Science

Self Cite

View full text Add to dashboard Cite

Aim Many countries lack informative, high‐resolution, wall‐to‐wall vegetation or land cover maps. Such maps are useful for land use and nature management, and for input to regional climate and hydrological models. Land cover maps based on remote sensing data typically lack the required ecological information, whereas traditional field‐based mapping is too expensive to be carried out over large areas. In this study, we therefore explore the extent to which distribution modelling (DM) methods are useful for predicting the current distribution of vegetation types (VT) on a national scale. Location Mainland Norway, covering ca. 324,000 km2. Methods We used presence/absence data for 31 different VTs, mapped wall‐to‐wall in an area frame survey with 1081 rectangular plots of 0.9 km2. Distribution models for each VT were obtained by logistic generalised linear modelling, using stepwise forward selection with an F‐ratio test. A total of 116 explanatory variables, recorded in 100 m × 100 m grid cells, were used. The 31 models were evaluated by applying the AUC criterion to an independent evaluation dataset. Results Twenty‐one of the 31 models had AUC values higher than 0.8. The highest AUC value (0.989) was obtained for Poor/rich broadleaf deciduous forest, whereas the lowest AUC (0.671) was obtained for Lichen and heather spruce forest. Overall, we found that rare VTs are predicted better than common ones, and coastal VTs are predicted better than inland ones. Conclusions Our study establishes DM as a viable tool for spatial prediction of aggregated species‐based entities such as VTs on a regional scale and at a fine (100 m) spatial resolution, provided relevant predictor variables are available. We discuss the potential uses of distribution models in utilizing large‐scale international vegetation surveys. We also argue that predictions from such models may improve parameterisation of vegetation distribution in earth system models.

show abstract

Section: Discussionsupporting

confidence: 89%

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Distribution modelling of vegetation types based on area frame survey data

Horvath

Halvorsen

Stordal

et al. 2019

Applied Vegetation Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…After flowing more than 200 km through the Taklimakan Desert, the Keriya River, originating in the south of the Kunlun Mountains, forms the natural Daliyaboyi Oasis. The core area of the oasis is 324 km 2 [25], the annual average precipitation is less than 10 mm [26], the weather is dominated by dust storms and floating dust, the vegetation composition is dominated by Populus euphratica and Tamarix chinensis [23]. Due to the lack of industrial and agricultural activities, the oasis is very rarely disturbed by human activities [27].…”

Section: Study Areamentioning

confidence: 99%

“…Random sampling survey is one of the most important methods for conducting ecological research; it is widely used in studies of plant species diversity, species association analysis, vegetation spatial pattern research, biomass estimation, and other areas [1][2][3]. In field vegetation surveys, especially in large-scale comprehensive surveys, the use of a single quadrat size is both convenient and effective; this approach is thus widely used in field vegetation surveys [4,5].…”

Section: Introductionmentioning

confidence: 99%

Methodology for optimizing quadrat size in sparse vegetation surveys: A desert case study from the Tarim Basin

et al. 2020

View full text Add to dashboard Cite

Random sampling is an important approach to field vegetation surveys. However, sampling surveys in desert areas are difficult because determining an appropriate quadrat size that represent the sparse and unevenly distributed vegetation is challenging. In this study, we present a methodology for quadrat size optimization based on low-altitude high-precision unmanned aerial vehicle (UAV) images. Using the Daliyaboyi Oasis as our study area, we simulated random sampling and analyzed the frequency distribution and variation in the fractional vegetation cover (FVC) index of the samples. Our results show that quadrats of 50 m × 50 m size are the most representative for sampling surveys in this location. The method exploits UAV technology to rapidly acquire vegetation information and overcomes the shortcomings of traditional methods that rely on labor-intensive fieldwork to collect species-area relationship (SAR) data. Our method presents two major advantages: (1) speed and efficiency stemming from the application of UAV, which also effectively overcomes the difficulties posed in vegetation surveys by the challenging desert climate and terrain; (2) the large sample size enabled by the use of a sampling simulation. Our methodology is thus highly suitable for selecting the optimal quadrat size and making accurate estimates, and can improve the efficiency and accuracy of field vegetation sampling surveys.

show abstract

Choice of predictors and complexity for ecosystem distribution models: effects on performance and transferability

Naas,

Keetz,

Halvorsen

et al. 2024

Ecography

Self Cite

View full text Add to dashboard Cite

There is an increasing need for ecosystem‐level distribution models (EDMs) and a better understanding of which factors affect their quality. We investigated how the performance and transferability of EDMs are influenced by 1) the choice of predictors and 2) model complexity. We modelled the distribution of 15 pre‐classified ecosystem types in Norway using 252 predictors gridded to 100 × 100 m resolution. The ecosystem types are major types in the ‘Nature in Norway' system mainly defined by rule‐based criteria such as whether soil or specific functional groups (e.g. trees) are present. The predictors were categorised into four groups, of which three represented proxies for natural, anthropogenic, or terrain processes (‘ecological predictors') and one represented spectral and structural characteristics of the surface observable from above (‘surface predictors'). Models were generated for five levels of model complexity. Model performance and transferability were evaluated with data collected independently of the training data. We found that 1) models trained with surface predictors only performed considerably better and were more transferable than models trained with ecological predictors, and 2) model performance increased with model complexity, levelling off from approximately 10 parameters and reaching a peak at approximately 20 parameters, while model transferability decreased with model complexity. Our findings suggest that surface predictors enhance EDM performance and transferability, most likely because they represent discernible surface characteristics of the ecosystem types. A poor match between the rule‐based criteria that define the ecosystem types and the ecological predictors, which represent ecological processes, is a plausible explanation for why surface predictors better predict the distribution of ecosystem types. Our results indicate that, in most cases, the same models are not well suited for contrasting purposes, such as predicting where ecosystems are and explaining why they are there.

show abstract

Spatial Transferability of Vegetation Types in Distribution Models Based on Sample Surveys from an Alpine Region

Cited by 3 publications

References 53 publications

Distribution modelling of vegetation types based on area frame survey data

Distribution modelling of vegetation types based on area frame survey data

Methodology for optimizing quadrat size in sparse vegetation surveys: A desert case study from the Tarim Basin

Choice of predictors and complexity for ecosystem distribution models: effects on performance and transferability

Contact Info

Product

Resources

About