Biodiversity loss ranks among the top four risks to humanity, as stated in the 2023 World Economic Forum Global Risks Report. Understanding biodiversity's basic building blocks is essential for monitoring changes, identifying threats, and implementing policy changes. However, the available data are biased with most of the terrestrial animal diversity remaining unknown because they are in "dark taxa." These taxa are poorly represented in databases like the Global Biodiversity Information portal (GBIF) which has nine times more information on birds than arthropods, despite birds only accounting for 0.2% of biodiversity. Overcoming this taxon bias is an enormous challenge, because most of the undescribed diversity is in insects. Yet, describing and characterising these species is crucial for understanding and managing ecosystems since their combined biomass and biodiversity far exceed those of all wild vertebrates, making them indispensable for survival. But where are our main knowledge gaps?
Fig 1. Top ten insect families in Malaise trap samples. NI represents Neglect Index which is the ratio of number of molecular Operational Taxonomic Units (estimated species richness) and total number of described species in the family. The higher the value, the more neglected the insect family.
We employed Malaise traps to determine the global taxonomic composition of flying insects. These standardised traps are widely used in global biomonitoring programs, but the samples contain so many specimens that they have to be analysed with modern methods. We here use new DNA sequencing technologies to estimate species diversity by obtaining DNA barcodes of thousands of specimens rapidly and cost-effectively.
We started by sequencing all insects in 24 samples from forests and marshes in Singapore. We then combined our data with existing information and eventually analysed 225,261 specimens to 25,000 species to 458 families. The most striking observation was that 10-20 families dominate flying insect communities worldwide. It is remarkable considering that the samples were collected from a variety of climates and habitats such as tropical rainforests, montane forests, cedar savannas, scrub forests, thorn fields, mangroves, and swamps, with only Australia and Antarctica not being sampled.
Fig 2. Species richness for different families in malaise traps, classified as top 10, 11-20th and other families. Black sections of the outer ring represent the dominance of Cecidomyiidae. Pie charts are scaled to total species richness.
But we also noticed that the most important insect families were often not the ones for which the largest number of species have been described. We thus quantified taxonomic research and found that many species-rich insect families have been neglected in the past and are still neglected today. Relative to their species diversity, few species have been described and there are few authors dedicated to changing this for the most species-rich insect families.
Describing insects is an immense challenge in comprehending life on Earth, with over 80% of them still undescribed. It is critical to note that a substantial portion of terrestrial animal biodiversity remains unknown to science, and this trend will persist unless "dark taxa" become a top priority target in biodiversity research. With modern-day species discovery efforts, we can expedite the process by leveraging the use of robotics, artificial intelligence, and new sequencing technologies.
Despite the increasing recognition of the importance of invertebrates, such as insects, in the biodiversity of ecosystems, they continue to be overlooked in research and conservation efforts. The neglect of species-rich insect families only adds to the challenge of preserving biodiversity. Hence, it is crucial to address this issue by providing incentives and support for the research and discovery of new species. By doing so, we can make significant progress towards a comprehensive understanding of biodiversity, and thus, help establish effective conservation strategies.
A video with highlights of our work is below: