Towards simplicity and accuracy: Assessing traditional and new estimators of orb-web capture thread length
| Autor: | Rafael Rios Moura, Gabriel Máximo Xavier, Marcelo O. Gonzaga |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
0106 biological sciences
Computer science media_common.quotation_subject 05 social sciences Foraging Prey capture Silk Estimator Spiders General Medicine Thread (computing) 010603 evolutionary biology 01 natural sciences Asymmetry Behavioral Neuroscience Predatory Behavior Animals 0501 psychology and cognitive sciences Animal Science and Zoology 050102 behavioral science & comparative psychology Algorithm media_common |
| Zdroj: | Behavioural processes. 173 |
| ISSN: | 1872-8308 |
| Popis: | Before using estimators, it is essential to consider their efficiency in order to avoid bias in results. Due to the architectural and structural complexity of spider webs, some important variables involved in prey capture are usually estimated based on a few measurements obtained from photographs. One of these variables is the capture thread length (CTL), which can provide valuable information on foraging behaviours and the energetic investment in prey capture. However, many of the webs found in the field are damaged, and there is no automatic method to measure the CTL. Therefore, the determination of a simple and accurate estimator of this variable is important to several studies involving spider foraging strategies. In this study, we assessed the accuracy of traditional and new CTL estimators and their vulnerability to web shape and asymmetry. Our results validated the accuracy of the previous estimators. However, we also presented a simple new estimator that can be even more accurate, irrespective of whether the webs exhibit circular shapes or asymmetry in thread investment between superior and inferior web parts. Moreover, we presented an accurate CTL estimator for non-circular orb webs, for which the traditional ones are not applicable. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect