Who Sings Tonight? Decoding Drivers of Species-Specific Insect Calling Activity - with Arpit Omprakash
IEN Research Spotlights are community events hosted by the India Ecoacoustics Network (IEN), where we highlight works of researchers who employ acoustics to explore and understand their environments. These sessions are free and open to join for anyone with an interest in ecoacoustics and environmental research.
We invite you to the 21st IEN Research Spotlight featuring Arpit Omprakash.
Arpit is broadly interested in applying and building Machine Learning tools to analyse large bioacoustic/ecoacoustic datasets and gain insights into community level patterns of acoustic behaviour and processes dictating them, particularly orthopteran and avian communities. He started his journey in acoustics by studying bat feeding aggregations using acoustics, and eventually moved on to recording bird communities, soundscapes from different habitats, and finally insect communities in the Western Ghats. His current research focuses on developing methods to monitor habitat quality and biodiversity in a wide range of paleotropical habitats including evergreen forests, deciduous forests, scrublands, grasslands, and land uses including plantations, primary protected forests, reserve forests, and naturally occurring monodominant forests in Southern India. Some of the methods he developed to predict land use and habitat quality led to valuable insights in terms of the role insect sounds play in predicting habitat quality across different habitats; since then his focus has been on studying understudied insect communities. Currently, he is curating an insect call library for the Western Ghats (particularly for Kudremukh National Park and surrounding regions), which he aims to use to build a sonotype level insect classifier and study the drivers of orthopteran acoustic community calling patterns in and around Kudremukh National Park.
Apart from working on acoustics, he loves to listen to music (probably this might be the real reason for him to start listening to animals in the wild) and he plays a little bit of guitar.
Abstract of the talk:
There have been significant developments in the field of Passive Acoustic Monitoring including the development of machine learning based models for estimating faunal activity using acoustics. However, most of the research focuses on taxa such as birds, bats, and frogs, ignoring the most abundant source of biophony in most habitats, insects. A small but growing body of literature on insect passive acoustic studies tries to study insect phenology, temporal patterns in calling behavior, and drivers of insect calling activity. Most such studies report patterns of insect calling activity measured using proxies such as ecoacoustic indices, which often fail to generalize in species-dense habitats such as tropical rainforests and are not useful for examining activity patterns at the species level. Factors such as temperature, moonlight, and habitat structure drive insect calling behavior. It has recently been shown in literature that some of these factors, e.g., moonlight, affect calling behavior in a species-specific manner based on life history traits and ecology of the species in question. These factors often interact to produce dynamic species-specific calling patterns, however, the interaction between these factors and how they affect insect calling has rarely been studied in the field. In this study, we use passive acoustic monitoring to study drivers of calling patterns in two species of orthopterans, namely Phyllomimus sp. and Pirmeda rosetta in Kudremukh National Park. We use machine learning based classifiers to analyze nocturnal recordings from multiple habitats and study how temperature, moonlight, and habitat variables interact and drive calling behavior of these species in a paleotropical ecosystem
Join us to learn more! To receive the the Zoom link and event updates, kindly RSVP via this Luma page.