Although 55 years have gone by since the Apollo 11 Moonwalk in 1969, mankind still faces that head-scratching question: with whom and what exactly do we share life on Earth?

Of the estimated up to 11 million species on our planet, only around 2.16 million have been formally described [1]. For insects, the number of species is frequently estimated at 5 million, of which only 1 million have been formally described [2]. That’s a big knowledge gap to fill! Insects are particularly challenging to discover and monitor. Whilst some insects are surprisingly large (a Phryganistria “chinensis” stick insect measured in at 64 cm in 2016) [3], others are so small that they can hardly be seen with the naked eye (the fairyfly wasp Dicopomorpha echmepterygis measures just 0.139 mm in length). Of the 4 million insect species yet to be discovered and described, it’s the tiny ones that make up most of this number. 

It is crucial to know what species exist, where they live, and what they do – in other words, to monitor and describe species. Research suggests that undescribed and newly described species are more likely to be threatened by extinction than earlier described ones [5]. They are often found in areas at risk or undergoing habitat loss and fragmentation. For example, the Neotropical and Indo-Malayan forests likely hold many undescribed species but are also hotspots for habitat destruction [5]. The reality of nature conservation is that only by knowing what exists can we make informed decisions, provide evidence, model scenarios, make predictions and assess conservation intervention efforts and usefulness to protect non-human, and consequently human, life on Earth [6].

Tasks of monitoring and identifying species are often in the hands of taxa specific experts – dedicated enthusiasts (both professionals and volunteers) who sort and sift and sieve their way to new species after potential suspects have been found in the field. Antenna are lined up straight, legs are delicately unfolded, and the insect specimen is cleaned and polished for its subsequent inspection under the microscope. Classification involves morphological or molecular methods. The former are based on physical characteristics (in 2017, a newly described moth species was named Neopalpa donaldtrumpi after the white-yellowish scales on its head [7].) The latter use DNA sequencing to compare the genetic material of species.

Hand sorting for monitoring (and describing) species requires a lot of labor, and there aren’t all that many people with the taxonomic expertise to allow for large-scale and high-resolution (species-level) assessments. Technologies to monitor what is around us may provide the giant leap we need to find, monitor, and protect the life around us. By harnessing the mind-blowing pace of technological advancement and concerting our efforts towards systems for automated monitoring, large-scale, automatic biodiversity assessment may be only a step away, even if it’s a big one. 

Almost 300 years ago, Carl Linnaeus formalized binomial nomenclature – the modern system for naming, classifying, and organizing organisms. But humans have been observing, naming and interacting with other species for a long time. The “discovery” of species cannot be separated from the Victorian history of exploitation, colonization and taking of natural resources without consent or fair remuneration, which although controlled nowadays, continues in forms of biopiracy. To monitor insects, a diverse set of classical methods have been traditionally implemented. These tend to be invasive and time-consuming, but also effective and accurate for high-resolution monitoring of species. 

So how do entomologists go about monitoring the insects of this world? Depends on what they want to find! Examples of classical methods include: malaise traps (big screens that catch insects flying by); sweep nets (for butterflies and other flying insects); pan traps with attractive colors (again for flying insects); beating sheets (used by hitting plants with a stick whilst holding canvas sheets underneath); traps with pheromones (for example for attracting moths); pitfall traps (for ground-dwellers); and active visual surveys (for larger insects that can be easily visually identified) [8].

Fig.1: Through the looking glass: exploring the world of insects through a microscope (J.Lüdtke. Bolzano, Sep. 2024)

Novel technologies can help to reduce labor-times and increase the spatial, temporal and taxonomic scales of monitoring. They can be grouped into four categories: computer vision; acoustic monitoring; radar; and molecular methods [9]. Cameras, often used with traps (e.g. light traps bait insects to a lit screen and take a photo of them when they land), are combined with image analysis for identifying species using artificial intelligence. Citizen scientists also collect huge amounts of data when they upload photos of interesting species to web portals like www.iNaturalist.org. Acoustic monitoring may be limited to insects that emit sounds. An AudioMoth or Raspberry Pi connected to an AI algorithm can allow for recording bird and bat calls and similar may be applied to insects [10]. Specialized recording systems do the same at higher quality. 

Radar monitoring uses radio waves to detect insects in the air [9]. It can detect both large swarms and give information on the size, shape, speed, trajectory and wing-beat frequency of individual insects [9]. Lidar technology uses the effect of the backscatter of a laser beam from flying insects to assess their abundance, diurnal activity patterns, and even identify insects to genus or species level [9]. DNA-barcoding and metabarcoding (identifying many taxa at the same time from one sample) can be performed directly on insects or on soil or water samples (DNA is accumulated in the surroundings as a species passes through it – this is known as environmental DNA). 

Other cool designs have been developed: for example, a sorting robot for insects from malaise traps [6]. This first takes photos of insects laid out on a tray for species identification and then single species are pipetted into wells for barcoding. Automated Multisensor stations for Monitoring of species Diversity (AMMODs) have been developed in Germany and combine several autonomous samplers to collect biodiversity data: audio recorders for vocalizing animals, sensors for volatile organic compounds emitted by plants, camera traps for mammals and small invertebrates. The data from these stations are then sent to receiver stations for storage and analysis [6]. They are like a climate weather station but for biodiversity!

With the pacey development of novel technologies, automated biodiversity monitoring is taking on the monitoring world by storm, yet some remain unconvinced by its current efficacy. There are several challenges that need to be overcome before a widescale implementation of automated biodiversity monitoring can be realized. To truly automate species identification, previously established databases are required. These include databases of DNA barcodes, animal sounds and images for training algorithms that aim to automate species identification [6]. A huge effort is required for curating and filling these databases. 

Questions remain: Can we provide enough high-quality data for pattern recognition to identify even the tiniest of insects? Can we create large enough reference and training data for all but the most unknown species? How can we adapt and integrate existing hardware and sensors for biodiversity monitoring? How can we employ and synchronize a large network of automated monitoring stations? Despite the many challenges and questions, automated biodiversity monitoring offers an exciting path forward. With a combination of coordinated efforts for technological advancement in this field and collective action to build databases and train algorithms, automated biodiversity monitoring may help to assemble the pieces of the puzzle and come closer to completing the picture of the life that exists on Earth. 

References