320 likes | 333 Views
This case study explores the factors that determine species distribution patterns at a continental scale, focusing on niche theory and neutrality theory. It highlights the value of data cleaning in analyzing ecological theories.
E N D
Using GBIF data to test niche vs. neutrality theories at a continental scale, and the value of data cleaning Tomer Gueta, AviBar-Massada and Yohay Carmel Faculty of Civil and Environmental Engineering Technion– Israel Institute of Technology
This case study What determine species’ distribution pattern at a continental scale?
The Niche theory (Grinnell 1924) Environment! Niche theory Topography Land-use Niche Veg. cover
The Neutral theory (Hubbell 2001) Stochastic processes and/or dispersal limitation predominant (Neutral theory)
Niche vs. Neutrality The continuum hypothesis (Gravel et al 2006) Niche and neutral theories are located at the two ends of a continuum Niche Neutral
The continuum hypothesis The continuum hypothesis (Gravel et al 2006) Niche and neutral theories are located at the two ends of a continuum Niche Neutral Modeling studies suggested that species richness is a main determinant • Species-rich communities are driven more strongly by neutral processes • Species-poor communities are driven more strongly by the niche Species-poor communities Species-rich communities species richness gradient
Prediction Species richness gradient Species richness Low High
Prediction- the missing link High effect Niche theory The effect of environmental factors on species distribution Neutral theory Low effect Species-poor communities Species-rich communities species richness gradient
Prediction 1: Continuum hypothesis Modeling studies suggested that species richness is a main determinant • Species-rich communities are driven more strongly by neutral processes • Species-poor communities are driven more strongly by the niche (Niche) High effect A clear negative correlation= continuum (-) Environment effect (Neutral) Low effect Species-poor communities species richness gradient Species-rich communities
Prediction 2: Niche Species-rich communities Species-poor communities Occurrence prob. Occurrence prob. Environmental gradient Environmental gradient (Niche) High effect Environment effect A clear positive correlation= niche (+) (Neutral) Low effect Species-poor communities species richness gradient Species-rich communities
Where? Australia
What? (IUCN, 2008)
Quantifying environmental effect High effect Niche theory The effect of environmental factors on species distribution Neutral theory Low effect
Species Distribution Model (SDM) Niche characterization + Species “X” Occurrencedata Difference environmental factors Predicted species distribution map Extracting datafor validation Model validation SDM performance
MaxEnt(Phillips et al. 2006) Niche characterization + Species “X” Occurrencedata Difference environmental factors Predicted species distribution map Extracting datafor validation Model validation MaxEnt ‘gain’
Methods Species-richness gradient for each species
Prediction 1: Continuum hypothesis Modeling studies suggested that species richness is a main determinant • Species-rich communities are driven more strongly by neutral processes • Species-poor communities are driven more strongly by the niche (Niche) High effect A clear negative correlation= continuum (-) MaxEnt ‘gain’ (Neutral) Low effect Species-poor communities species richness gradient Species-rich communities
Prediction 2: Niche Occurrence prob. Occurrence prob. Environmental gradient Environmental gradient (Niche) High effect MaxEnt ‘gain’ A clear positive correlation= niche (+) (Neutral) Low effect Species-poor communities species richness gradient Species-rich communities
Species richness • Organism perspective • Guilds • Biological characteristics: taxon, trophic level and body weight All Mammals Herbivores Carnivore Bats 1g-100g 100g-5000g
The data Global Biodiversity Information Facility
Study design Conclusions Conclusions Raw data Data cleaning
Data cleaning Conventional data check and filtering Geospatial
Data cleaning Taxonomic Temporal From 1,041,867 records to 515,479
Results- 100km grid Spearman rank correlation test (rho)a non-parametric correlation test Total
Results- 200km grid Total
Results- 300km grid Total
Conclusions • The effect of the Environment decreases with species richness • We exapmlefy the crucial role of data cleaning!
Thank you for listening Citations (Data resource)