Acoustic traits of bat-pollinated flowers compared to flowers of other pollination syndromes and their echo-based classification using convolutional neural networks

Ralph Simon, Yossi Yovel (Editor), Karol Bakunowski, Angel Eduardo Reyes-Vasques, Marco Tschapka, Mirjam Knörnschild, Jan Steckel, Dan Stowell

Research output: Contribution to journalArticleScientificpeer-review


Bats orient and forage by night relying on their biosonar system. Consequently, some plant species that are pollinated by bats evolved echo acoustic signals to attract their nocturnal pollinators. Only few of the more than 400 bat-pollinated neotropical plants have been studied in this regard and it remained unclear if bat-pollinated flowers share general echo acoustic adaptations and differ acoustically from flowers that are pollinated by other animals (e.g. birds, insects). In our study we directly compare the acoustics and morphology of flowers from plants belonging to different pollination syndromes. We compared six species pairs from different families and for each species we collected 14 individual flowers. With an automated ensonification setup we measured the echoes from 202 positions around the opening of each flower. We also took morphological measurements of each flower. We found that the acoustic reflectivity of the flowers, measured as target strength, was higher for bat-pollinated flowers irrespective of their size. To understand if flowers have species-specific acoustic signatures we trained a convolutional neural network to classify flowers by their echoes. We show that a high classification performance can be achieved already with a very limited number of echoes and that bat-pollinated flowers tend to be easier to classify. Our study not only provides new insights into the acoustics of floral signals but also is relevant from a signal processing point of view as we demonstrate how complex objects can be characterized with a very limited amount of acoustic data.

Bat-pollinated flowers have to attract their pollinators in absence of light and therefore some species developed specialized echoic floral parts. These parts are usually concave shaped and act like acoustic retroreflectors making the flowers acoustically conspicuous to the bats. Acoustic plant specializations only have been described for two bat-pollinated species in the Neotropics and one other bat-dependent plant in South East Asia. However, it remains unclear whether other bat-pollinated plant species also show acoustic adaptations. Moreover, acoustic traits have never been compared between bat-pollinated flowers and flowers belonging to other pollination syndromes. To investigate acoustic traits of bat-pollinated flowers we recorded a dataset of 32320 flower echoes, collected from 168 individual flowers belonging to 12 different species. 6 of these species were pollinated by bats and 6 species were pollinated by insects or hummingbirds. We analyzed the spectral target strength of the flowers and trained a convolutional neural network (CNN) on the spectrograms of the flower echoes. We found that bat-pollinated flowers have a significantly higher echo target strength, independent of their size, and differ in their morphology, specifically in the lower variance of their morphological features. We found that a good classification accuracy by our CNN (up to 84%) can be achieved with only one echo/spectrogram to classify the 12 different plant species, both bat-pollinated and otherwise, with bat-pollinated flowers being easier to classify. The higher classification performance of bat-pollinated flowers can be explained by the lower variance of their morphology.

Original languageEnglish
Article number1009706
Number of pages20
JournalPLOS Computational Biology
Issue number12
Publication statusPublished - 16 Dec 2021




Dive into the research topics of 'Acoustic traits of bat-pollinated flowers compared to flowers of other pollination syndromes and their echo-based classification using convolutional neural networks'. Together they form a unique fingerprint.

Cite this