Illegal trade of birds using Facebook in Argentina
Data files
Feb 14, 2025 version files 57.31 KB
-
Lopez_et_al_2025_DataSet.xlsx
56.25 KB
-
README.md
1.06 KB
Abstract
The illegal trade of wild birds is a large threat to biodiversity and sellers often use social media to traffic wildlife. By monitoring and characterizing the trade of wild and domestic birds on Facebook, we show that this illegal e-commerce remains widespread throughout Argentina, despite a decrease in activity in 2023 compared to 2017, highlighting the persistence and magnitude of this issue. In 2017, we found 96,354 users in 127 buying/selling groups in 18 provinces, of which 61% were public. In 2023, we found 58,647 users in 76 groups in 14 provinces, of which 66% were public. Analyzing one group in 2017 and two groups in 2023, we found 67 bird species were offered for sale. All birds for sale in these groups had illegal origins (i.e., without authorized bands or permits), including two Endangered and one Near Threatened species. Passerines (mainly Thraupidae, Fringilidae, and Cardinalidae species) were the most offered group, but parrots, raptors, and other bird groups were also available. Saltator aurantiirostris (Golden-billed saltator) was most frequently offered for sale and had the highest interest from users in the two largest analyzed groups. The Endangered Gubernatrix cristata (Yellow cardinal) reached the highest price (up to 328 USD), it was also frequently offered for sale (5th position in 2017) and had high user interest (2nd in 2017). We found that Facebook was frequently used for illegal bird trade in Argentina, including species of conservation concern, which can threaten a number of species. We call for coordinated and interdisciplinary actions that engage authorities, social media companies, and local communities, which will be needed to enact top-down and bottom-up measures to continue reducing and eventually halting wildlife-related illegal activities.
https://doi.org/10.5061/dryad.rn8pk0pnr
Description of the data and file structure
File: Lopez_et_al_2025_DataSet.xlsx
Description: Facebook groups we found in 2017 and 2023. Bird species we found in one group in 2017 (La Pampa) and two groups in 2023 (La Pampa and National).
Variables
- FacebookGroups: GroupID, Year, Province, Group_Type, Users, Posts. "NA" is information that is not available because the number of posts was not reported in some groups.
- BirdSpecies: Species, English_Name, Spanish_Name, Family, Order, Days, the last 20 USD prices, and the last 20 numbers of Comments_and_likes as subrogates of interest from users, recorded in each analyzed group (2017 La Pampa, 2023 La Pampa, and 2023 National). So, each variable consists of the
[comments and likes or price]_[1-20]_[FB group]
. "NA" is information not applicable because there were fewer than 20 posts for some offered species.
Code/software
Excel
Data collection
We created a Facebook user account and anonymously collected all data without interacting directly with other users (i.e., passive collection). No personal, private, or identifiable information was obtained, ensuring compliance with ethical guidelines for research involving humans. We conducted the same search twice, in March 2017 and in April 2023, using four terms related to buying and selling birds in Spanish (aves, pájaros, venta, and compra) and the names of 23 Argentinean provinces and the federal capital district through the inbuilt ‘search’ function. From each identified group, we collected: 1) the province where the group was based, 2) the type of group (i.e.; public for open groups and closed for groups where authorization from the group administrator is required), 3) the number of users, and 4) the average number of posts per year.
We conducted a detailed analysis of three buy/sell groups. In 2017, we analyzed one of the largest groups in the country located in La Pampa province (hereafter, ‘2017 La Pampa’). Since most of the authors were located in this province, we had a deeper knowledge of the local bird species, management authorities, and administration process of natural resources. Six years later, we analyzed a different group in La Pampa province (‘2023 La Pampa’) as well as a group covering the whole country (‘2023 national’).
We identified bird species for sale by focusing on the ‘Recent media’ section of each group, which contained user-generated photos and videos. We excluded text posts since sellers do not use scientific names and the same common names could refer to various species. Once we identified a species in a photo or a video, we searched the group for posts related to this species. For each species, we included up to the first 20 posts that were identified by the search engine, but only if they related to the sale of the species. Next, we recorded the range of prices for each species in USD based on official pesos prices in 2017 and 2023 exchange rates, respectively, for each year. We also recorded the origin of the species (i.e., native to La Pampa province, native to Argentina but does not naturally occur in La Pampa province, exotic species, or domestic species that have been genetically modified from their wild ancestors through artificial selection to serve human purposes), as well as the conservation status and population trend of the species according to the IUCN Red List of Threatened Species (IUCN 2023). Finally, we specified if the species was on the list of seized birds in Argentina in 2020 (Grisolia and Grosskopf 2021).
Statistical analysis
We tested the change in the number of groups and total users per province between 2017 and 2023 using a Wilcoxon test and the relation between the number of users per province and the human population for the same province (INDEC 2012) using Spearman’s correlation at a significance level of 0.05. We report the mean ± S.E. (Range).
For the selected groups, we estimated the ‘offer’ of each species and the ‘interest’ that users had in each species. The ‘offer’ was defined as the relative percentage of posts in which birds of a given species were offered for sale, while the ‘interest’ represented the relative percentage of user interaction with each species. To calculate these values, we first determined the number of days over which the 20 most recent posts offering a particular species for sale had been made, which is the ‘observed time’ for this species, under the assumption that group activity remained constant over time. For the relative percentage of offers, we used the number of posts per species (i.e., 1-20). For the relative interest metric, we used the number of comments and ‘likes’ of each post (ensuring that each user interaction was counted only once per post, even if the same person liked and also commented on the same post) as a surrogate of the interests of users. Then, we obtained these values based on the ‘total time group has existed’ (2017 La Pampa: 941 days, 2023 La Pampa: 3,170 days, 2023 national: 1,014 days) using the following equations, where i equals each species and k denotes the total number of species offered for sale in each analyzed group. (See the equations in the published paper by López et al. 2025 Ornithological Applications).