The Action has provided grants for the following completed STSMs:
Dr Ana Sofia Vaz
STSM title: Improving the surveillance of alien plants through social participation: a comparison of distinct citizen science-generated data
Grantee name: Ana Sofia Vaz, PhD
Home institution: IISTA-CEAMA - Andalusian Inter-University Institute for Earth System Research, University of Granada, Spain
Host institution: CIBIO/InBIO - Research Centre in Biodiversity and Genetic Resources, Research Network in Biodiversity and Evolutionary Biology, University of Porto, Portugal
Citizen science represents a great opportunity for collecting data on the occurrence of alien and invasive species under great time- and cost-efficiency. These data are needed to feed modelling approaches that support the surveillance and monitoring of alien species distributions. However, evaluating the usefulness of different citizen science approaches to support the monitoring of alien species through modelling exercises is far from being explored. The goal of this STSM was to evaluate the usefulness, strengths and limitations of data collected by different user groups and citizen science approaches for predicting the potential distribution of alien plant species and for guiding future surveillance efforts. For that, we first compiled species occurrence data from:
Stakeholders’ participatory mapping – involving the organization of a workshop with different decision-makers who where invited to identify (on a “printed” map of Portugal), the locations of different invasive alien plants;
Biodiversity-oriented mobile apps - collecting pictures of invasive species observed and respective coordinates from popular apps on nature observations in general (iNaturalist) and on Portuguese invasive species (invasoras: http://www.invasoras.pt);
Public social media data - considering georeferenced pictures from social media platforms (Flickr: https://www.flickr.com) and identifying those pictures where invasive species occur.
After compiling the previous data, we developed habitat suitability models (HSM), which aimed to identify the social-ecological drivers and the predicted spatial distribution of alien invasive plants in mainland Portugal. For the stakeholders’ participatory mapping, results showed the lack of awareness on the location of invasive plants from the participating stakeholders, emphasising the need for training and information outreach strategies. For data based on mobile apps we were able to identify those areas more prone to the invasion by Cortaderia selloana, and therefor to identify the priority areas for the species management (Figure 1). For the social media data and due to the high number of available data (i.e. 583369 Flickr pictures), we decided to implement an automated detection approach of C. selloana in social media data, using Faster RCNN-ResNet machine learning models in Google Colab (Figure 2). The model has been trained and is now ready to be tested in the social media networks. Results are expected to be published soon.
This STSM has been achieved with the collaboration of João Honrado, Joana R. Vicente (University of Porto, Portugal), Hélia Marchante and Elizabete Marchante (University of Coimbra, Portugal). Results from this STSM contribute to the Working Group WG2 by testing the value of new and emerging technologies for citizen science in the context of alien and invasion species, as well as to WG4, by evaluating the strengths and limitations of methods used for managing different kinds of citizen science data to provide relevant information for decision-making (WG4).
Figure 1 - Occurrences of Cortaderia selloana in the citizen-science mobile app invasoras app (a) and respective results of the habitat suitability models (b).
Figure 2 - Distribution of publicly available Flickr pictures for Portugal (a). Example of an annotated picture from invasoras app (b), currently being used for model refinement (c).