A global analysis of changes in invertebrate species richness with area: deriving global species-area relationships from ecoregional species richness using occurrence records from the GBIF database

Abstract

Although invertebrate species are a major component of global biodiversity, they remain much less studied when compared to vertebrate species and plants. This research aims to increase the knowledge on patterns in the global spatial distribution of terrestrial invertebrate species by deriving species-area relationships (SARs; which denote the relationship between species richness and area) on a global scale. Terrestrial Ecoregions of the World (TEOW) is used for geospatial data on the distribution and classification of global terrestrial ecosystems. Georeferenced species occurrence records are collected from the Global Biodiversity Information Facility (GBIF) database and are used to calculate the invertebrate species richness per ecoregion. SARs and corresponding coefficients are derived by using invertebrate species richness and area per ecoregion as input variables; this was done for the global SAR as well as for the individual biomes from TEOW. Different SAR models are tested with the dataset and compared using Information Criterion (IC) weights, with the log-transformed power model showing the best fit. The slope (z-values) and intercept (c-values) of the SARs were calculated using linear regression in log-log space. The same analysis is subsequently performed for subsets of the dataset: separate analyses were performed for only insects and for all records excluding insects, in order to compare the differences between taxonomic groups; and for subsets containing only island-based and only continental mainland records, in order to compare differences between geographic features. For the global SAR, the value for z was calculated at 0.281 and the value for log(c) at 1.107. In general, the SARs for biomes presented slopes that were steeper than for the global SAR. The biome with the highest ecoregional mean species richness was Temperate Broadleaf and Mixed Forests. The biome with the highest statistically significant observed z-value (slope) was Flooded Grasslands and Savannas, although this biome also displayed the lowest ecoregional mean species richness. The biome with the lowest statistically significant z-value was Tropical and Subtropical Dry Broadleaf Forests. Observed z-values were higher for exclusively insects and for island-based records, but lower for invertebrates other than insects and for continental mainland based records when compared with the SAR for the complete dataset. The obtained SARs can be used to predict changes in invertebrate species richness with changes in surface area of natural habitat and thus as a baseline in ecological modelling and conservation management.

Date of Award	25 Jan 2023
Original language	English
Supervisor	Dennis Uit de Weerd (Examiner), Hadassa Cruz Moreira (External assessor), Koen Kuipers (External assessor), Aafke M. Schipper (Co-assessor) & Jikke van Wijnen (Co-assessor)