How to distinguish increased observations of a species from overall increased observations

I would like to use observations of an alien species in natural areas over time to determine whether the abundance of that species is increasing. Just as an example, if I could look at sawtooth oak (native somewhere in Asia) observations going back to the first year it was observed in iNat to the most recent year. I would need a way to compensate for the overall increase in iNat observations each year. Has anyone else explored this topic?

You can check this topic https://forum.inaturalist.org/t/biases-in-inat-data/23943/84 and I remember others close to what you ask. Also it may be helpful checking topics with published papers that used iNat data, there may be something similar to your idea (here and more recent ones)

1 Like

The simplest way would be to divide observations per week/month/year by active users in that area. However, one very active user can wildly skew data when they visit a new region or ramp up activity.

3 Likes

A common method would be to find a comparison species (or more of those) for which you are reasonably sure that the population is indeed constant and then just track the ratios of observations of your target species to this.

Right - for example, I recently learned a species of Solanum that wasn’t in the CV was present in my city. I went out of my way to make dozens of observations of that plant in order to help it make it into the next model revision. In the process I became the number one observer (simply because there were previously so few) and made my city a hotspot for this species. Was there an actual spike in the proliferation of this species in my area? No, but the data might mislead someone to believe that.

1 Like

Thank you everyone, I am reading your comments with interest and following up on the links!

I heard about the “List Length method”
WHich is about the Red List and estimation of progress per Land by reacalculating it to each year and and each square km2 and the amount of days with observations.

And occupancy moddeling.

https://jamesepaterson.github.io/jamespatersonblog/2020-11-09_occupancy_part2.html
As a reminder: occupancy models are useful for estimating species’ distributions and relationships between species occurrence and habitat or landscape variables (among many other things!).

A basic occupancy model estimates two parameters: