Hello everyone,
I am a PhD student working on the species richness of birds in Morocco, using GBIF data. My research is not only about describing the biases in biodiversity data (spatial, temporal, taxonomic, sampling effort), but more importantly about how to handle these biases in order to obtain reliable and publishable results.
Specifically, I am:
-
Comparing two periods (2015–2016 vs. 2024–2025) to test whether time has a significant effect on species richness.
-
Identifying regions with an increase or decrease in richness.
-
Integrating explanatory variables such as urbanization, population density, land use, and road density to understand the drivers of biodiversity change.
I am aware that GBIF data comes with important biases (e.g. empty municipalities, spatial clustering of records, uneven sampling effort). My goal is to apply the appropriate corrections (e.g. spatial subsampling, rarefaction, coverage-based completeness estimates) before testing temporal and regional effects.
My questions to the community:
-
What are the most robust approaches you recommend to deal with communes/municipalities without any records in GBIF?
-
Do you have experience with coverage-based rarefaction or similar methods for standardizing richness across administrative units?
-
Are there R packages, workflows, or references that you would particularly recommend for this type of analysis (birds, regional scale, temporal comparisons)?
Any advice, resources, or methodological suggestions would be very valuable to strengthen my work.
Thank you in advance for your time and support 