Implementation of Fractal Dimension and Self-Organizing Map to Detect Toxic Effects of Toluene on Movement Tracks of Daphnia magna.

Yuedan Liu,Zhongya Fan,Chunlei Xia,Renren Wu,Zuoyi Liu,Xianglin Chen

doi:10.1155/2018/2637209

Abstract

Movement behaviors of an indicator species, Daphnia magna, in response to contaminants have been implemented to monitor environmental disturbances. Complexity in movement tracks of Daphnia magna was characterized by use of fractal dimension and self-organizing map. The individual movement tracks of D. magna were continuously recorded for 24 hours before and after treatments with toluene at the concentration of 10 mg/L, respectively. The general complexity in movement tracks (10 minutes) was characterized by fractal dimension. Results showed that average fractal dimension of movement tracks was decreased from 1.62 to 1.22 after treatments. The instantaneous movement parameters of movement segments in 5 s were input into the self-organizing map to investigate the swimming pattern changes under stresses of toluene. Abnormal behaviors of D. magna are more frequently observed after treatments than before treatments. Computational methods in ecological informatics could be utilized to obtain the useful information in behavioral data of D. magna and would be further applied as an in situ monitoring tool in water environment.

Highlights

Behavior is an organism response to environmental stimuli defined as an action, a reaction, or a function of a system under a specific circumstance [1]
The information involved in animal movement behavior, has been regarded as difficult to exact due to complexity residing in the data, and so effective methods to analyze behaviors without a priori knowledge are important in the behavioral monitoring [9]
This study is aimed at (1) detecting chemical effects of toluene on movement behaviors of Daphnia magna based on general shape and movement patterns of swimming tracks after treatments of toluene in static water environment and (2) verifying the fact that computational analyses could be used to effectively reveal the behavioral information of D. magna for longer and shorter time units

Summary

Introduction

Behavior is an organism response to environmental stimuli defined as an action, a reaction, or a function of a system under a specific circumstance [1]. Automatic detection of behavioral responses of water species has been applied as an efficient tool for biomonitoring in aquatic environment, because animal behaviors would be suitable as indicators for various pollutants [2, 3]. The information involved in animal movement behavior, has been regarded as difficult to exact due to complexity residing in the data, and so effective methods to analyze behaviors without a priori knowledge are important in the behavioral monitoring [9]. The movement tracks of water indicator species could be segmented and patterned to detect the abnormal behaviors occurring when individuals are exposed to chemicals [13, 14]. Movement segmentation and movement patterning, have never been used to comparatively study animal behavioral data for different time scales

Objectives

Methods

Results

Discussion

Conclusion