Bewaard in:
| Hoofdauteurs: | , , , , , , , |
|---|---|
| Formaat: | Recurso digital |
| Taal: | |
| Gepubliceerd in: |
Zenodo
2021
|
| Onderwerpen: | |
| Online toegang: | https://doi.org/10.5281/zenodo.13527089 |
| Tags: |
Voeg label toe
Geen labels, Wees de eerste die dit record labelt!
|
Inhoudsopgave:
- (Uploaded by Plazi for the Bat Literature Project) Bat-pollinated flowers have toattract their pollinators inabsence oflight 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 tothe bats. Acoustic plant specializations only have been described for two bat-pollinated species inthe Neotropics and one other bat-dependentplant inSouth East Asia. However, itremains 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 toother pollination syndromes. To investigate acoustic traits ofbat-pollinated flowers we recorded adataset of32320 flower echoes, collected from 168 individual flowers belonging to12 different species. 6ofthese species were pollinated by bats and 6species were pollinated by insects orhummingbirds. We analyzed the spectral target strength ofthe flowers and trained aconvolutional neural network (CNN) on the spectrograms ofthe flower echoes. We found that bat-pollinated flowers have asignificantly higher echo target strength, independent oftheir size, and differ intheir morphology, specifically inthe lower variance oftheir morphological features. We found that agood classification accuracy by our CNN (up to84%) can be achieved with only one echo/spectrogram toclassify the 12 different plant species, both bat-pollinated and otherwise, with bat-pollinated flowers being easier toclassify. The higher classification performance ofbat-pollinated flowers can be explained by the lower variance oftheir morphology.