Automated species identification tools as biodiversity teachers

Automated image classification is transforming ecological research, in particular in the domain of species identification. Moreover, modern algorithms are increasingly capable of mapping the distributions of thousands of species with remarkable precision. These innovations are deepening our ecological understanding, and when integrated thoughtfully, they can substantially boost biodiversity monitoring efforts, both by speeding up expert identification and by providing automated quality control. Beyond their scientific applications, these technologies hold an underappreciated yet powerful potential to function as readily accessible field guides, helping curious users become familiar with the species that live in their immediate surroundings.

Traditional tools for species identification, such as dichotomous keys, demand substantial investments of time and effort before their usage becomes efficient. Users must learn specialized terminology and acquire an overview of major taxonomic groups before they can make successful identifications. Automated identification services, on the other hand, deliver results instantly. In regions such as central Europe, several plant identification applications achieve high success rates, particularly for common and conspicuous species, which are the organisms that tend to attract beginners. Even when these tools do not provide a definitive answer, almost always they narrow the possibilities to a small number of relevant candidates, making subsequent confirmation with keys or field guides manageable even for non-experts.

With SpeciesID, we are building a suite of openly accessible species identification tools for Switzerland, covering several taxonomic groups and using information on local habitat conditions to refine image-based interpretations. I will illustrate how these developments can support learning and exploration, both for well-described groups such as vascular plants and butterflies, and for groups whose fascinating diversity so far has only been accessible to a handful of experts, such as moths. When built upon sufficient, high-quality training data, automated identification tools can empower anyone willing to embark on the journey to biodiversity literacy.