Choosing Spatial Thinning Distance

[Abigail Minnekine]

Hello everyone, I am researching the effect of climate change on invasive terrestrial plants and animals in the Northeast United States. I am wondering what other users are using for the Thinning Distance (km) in the Process Occurance Data component for plants. I was previously using the 10-km rule for both plants and animals, but have since learned that it is quite arbitrary. Any helpful literature, support of the 10-km rule, or an explanation of what you chose for a distance would be greatly appreciated. Thanks in advance!

[Wallace Ecological Modeling App]

Hello,

Spatial thinning of occurrence data is done to reduce the effects of sampling bias across geographic space, whereas the other two options for processing the occurrence data included in Wallace (Select Occurrences on Map & Remove Occurrences by ID) are more for removing erroneous points. The distance chosen for spatial thinning is species-specific and may relate to the environment of sampling. 

Some reading material on this topic is, of course, the Module Guidance Text, found within Wallace for the Spatial thinning module.
You can also check out:
https://doi.org/10.1111/ecog.01132 (Aiello-Lammens et al. 2015) This is the paper on spThin, the package that is within Wallace used to thin
https://doi.org/10.1016/j.ecolmodel.2013.12.012 (Boria et al. 2014) A paper on how spatial thinning can improve model performance
https://doi.org/10.1111/j.1365-2699.2009.02174.x (Veloz 2009) A paper on how spatially autocorrelated sampling affects measures of model accuracy