TicoTyler
Active member
I wanted to share a bit of a project as well as provide some tips on how to replicate it for your area. This year I'm trying to focus more on audio recording and specifically where I live (Costa Rica) there are many underrepresented species. Programs like Merlin Sound ID require 100 - 150 good-quality recordings to train their machine-learning model. 385 out of 910 species are under the 150 threshold. Some species are completely non-vocal (Yellow-billed and Snowy Cotingas) and others are unlikely to be identified by voice (most of the pelagic species and a lot of hummingbirds). A few on the official list are also either vagrants or ghost birds (Colombian Crake). So even winnowing down the list to species that are not good candidates it still has 191 under 100 recordings and 279 under 150 recordings. I want to focus on these species, my question was: without checking each species individually, how do I generate a target list?
In the past, I manually made a list like this with the data on all of the 107 endemic species, but it was a lot of work. Macaulay Library/eBird puts out a document with target species but at a lower threshold (10) than I was looking for. However, there is a page on the document that contains all the species and their media totals. If there is an Excel or Google Sheet version of the bird list from your target area, you can combine the ML data with that list and a simple formula to create a sortable/filterable list of targets. (If not there are semi-simple ways to generate a list but that's for another post). This would work for finding audio, photo, and/or video needs.
My Costa Rica Target List
Here's what I did:
1. Take the Google Sheet from Macaulay Library and make a copy to my drive.
2. Delete all the tabs except "ML Media Totals".
3. Create a new tab and copy/paste the species from your target area. I went with both the common and scientific names for ease of reference, but used the scientific name as the data point for connecting the two.
4. Use the Xlookup formula in Google Sheets to pull data from the ML Media Totals onto your local targets tab. If you look at my document, there are three tabs. The ML Media Totals is where the data comes from. The Birds of Costa Rica tab is the whole list from eBird of species reported in the country and I used the Xlookup to populate the media totals. The third tab is "Curated List" where I whittled down the ones I'm not likely to record. In the case of either tab you can use the "Sort sheet by column" option to show the number of media items from least to most.
Sample formula:
=IFERROR(XLOOKUP(A2, 'ML Media Totals'!C:C, 'ML Media Totals'!D
, "Not Found"), "")
In the past, I manually made a list like this with the data on all of the 107 endemic species, but it was a lot of work. Macaulay Library/eBird puts out a document with target species but at a lower threshold (10) than I was looking for. However, there is a page on the document that contains all the species and their media totals. If there is an Excel or Google Sheet version of the bird list from your target area, you can combine the ML data with that list and a simple formula to create a sortable/filterable list of targets. (If not there are semi-simple ways to generate a list but that's for another post). This would work for finding audio, photo, and/or video needs.
My Costa Rica Target List
Here's what I did:
1. Take the Google Sheet from Macaulay Library and make a copy to my drive.
2. Delete all the tabs except "ML Media Totals".
3. Create a new tab and copy/paste the species from your target area. I went with both the common and scientific names for ease of reference, but used the scientific name as the data point for connecting the two.
4. Use the Xlookup formula in Google Sheets to pull data from the ML Media Totals onto your local targets tab. If you look at my document, there are three tabs. The ML Media Totals is where the data comes from. The Birds of Costa Rica tab is the whole list from eBird of species reported in the country and I used the Xlookup to populate the media totals. The third tab is "Curated List" where I whittled down the ones I'm not likely to record. In the case of either tab you can use the "Sort sheet by column" option to show the number of media items from least to most.
Sample formula:
=IFERROR(XLOOKUP(A2, 'ML Media Totals'!C:C, 'ML Media Totals'!D
, "Not Found"), "") 
