Application of Predictive Distribution Modelling to Invertebrates: Odonata in South Africa

The application of distributional modelling techniques to invertebrates has seldom been explored, primarily due to a lack in adequate distributional data for these taxa. Here, we have selected a simple modelling approach for the generation of distribution maps from a limited dataset, as a first step...

Full description

Saved in:
Bibliographic Details
Published in:Biodiversity and conservation 2006-12, Vol.15 (13), p.4239-4251
Main Authors: Finch, Jemma M, Samways, Michael J, Hill, Trevor R, Piper, Steven E, Taylor, Stuart
Format: Article
Language:English
Subjects:
Bioclimatology
Distribution patterns
Invertebrata
Invertebrates
Odonata
Sensitivity analysis
Species richness
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The application of distributional modelling techniques to invertebrates has seldom been explored, primarily due to a lack in adequate distributional data for these taxa. Here, we have selected a simple modelling approach for the generation of distribution maps from a limited dataset, as a first step to the atlassing of Odonata in South Africa. The BIOCLIM-type approach was selected for this purpose, as it requires minimal data for model building and validation procedures. BIOCLIM partitions an area climatically prior to survey, and predicts species distributions on a bioclimatic basis. Conservative deterministic models were developed using point presence/absence data for each of the regions' 160 described species. These models were validated by cross-validation, and the Jaccard coefficient of similarity was used as an index of model performance. A sensitivity analysis investigated the influence of extreme values and errors in the data on predictive ability. Models identified disjunct distribution patterns and accurately predicted the restricted ranges of habitat-specialist species. However, models overstated the distribution of habitat generalists and species with distinct outlier records. For accurate predictions of broad-ranging species, it is suggested that a probabilistic approach be adopted. Nevertheless, basic distribution patterns generated through this conservative approach can be further applied to the investigation of species richness and issues relating to conservation, such as reserve design. The BIOCLIM-type approach provided a means of predicting species distributions, allowing for broad-scale atlassing and thereby providing the first step towards Odonata conservation in South Africa.[PUBLICATION ABSTRACT]
ISSN:0960-3115
1572-9710
DOI:10.1007/s10531-005-3577-z