Dataset Information

Classifying Goliath Grouper (Epinephelus itajara) Behaviors from a Novel, Multi-Sensor Tag.

ABSTRACT: Inertial measurement unit sensors (IMU; i.e., accelerometer, gyroscope and magnetometer combinations) are frequently fitted to animals to better understand their activity patterns and energy expenditure. Capable of recording hundreds of data points a second, these sensors can quickly produce large datasets that require methods to automate behavioral classification. Here, we describe behaviors derived from a custom-built multi-sensor bio-logging tag attached to Atlantic Goliath grouper (Epinephelus itajara) within a simulated ecosystem. We then compared the performance of two commonly applied machine learning approaches (random forest and support vector machine) to a deep learning approach (convolutional neural network, or CNN) for classifying IMU data from this tag. CNNs are frequently used to recognize activities from IMU data obtained from humans but are less commonly considered for other animals. Thirteen behavioral classes were identified during ethogram development, nine of which were classified. For the conventional machine learning approaches, 187 summary statistics were extracted from the data, including time and frequency domain features. The CNN was fed absolute values obtained from fast Fourier transformations of the raw tri-axial accelerometer, gyroscope and magnetometer channels, with a frequency resolution of 512 data points. Five metrics were used to assess classifier performance; the deep learning approach performed better across all metrics (Sensitivity = 0.962; Specificity = 0.996; F₁-score = 0.962; Matthew's Correlation Coefficient = 0.959; Cohen's Kappa = 0.833) than both conventional machine learning approaches. Generally, the random forest performed better than the support vector machine. In some instances, a conventional learning approach yielded a higher performance metric for particular classes (e.g., the random forest had a F₁-score of 0.971 for backward swimming compared to 0.955 for the CNN). Deep learning approaches could potentially improve behavioral classification from IMU data, beyond that obtained from conventional machine learning methods.

SUBMITTER: Brewster LR

PROVIDER: S-EPMC8512029 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:The Atlantic Goliath Grouper (Epinephelus itajara) population has rebounded from near extinction to an international status as vulnerable due in part to regional species recovery efforts. The southeastern US population has been recovering with the main spawning locations off the coasts of Florida. Despite their economic importance to the catch-and-release fishery and the dive industry, and their ecological importance as ecosystem engineers resulting in positive impacts on reefs and species richness, baseline health assessment information is very limited in this species to date. The objectives of this study were to: (1) establish reference intervals for hematological and plasma biochemical analytes, and report immune function, oxidative stress, and vitellogenin in mature males and females; (2) evaluate total length, age, and sex in relation to blood analytes in juvenile and mature fish; (3) assess analytes across sampled months in mature male and female fish; and (4) describe the typical light microscopy findings in liver and gill biopsies, including quantitative assessment of pigmented macrophage aggregates. Health indices are reported as reference intervals when applicable, or otherwise descriptively. Blood analyte correlations with length and age, sex differences, and comparisons across months provided relevant physiological considerations, including differences in protein/energy metabolism, tissue growth, sexual maturation, active reproduction, and antigenic stimulation. Liver histology identified changes associated with life stage, active reproduction, or of subclinically to clinically insignificant infectious and/or inflammatory processes. Hepatocellular vacuolation and pigmented macrophage aggregates were prominent. Pigmented macrophage aggregates correlated with total length, presumably from continuous antigenic stimulation and/or metabolic changes as fish grow. Gill histological findings were subtle. The data presented herein provide an essential baseline assessment of a suite of health variables in an iconic marine teleost species, serves as a springboard for future studies relevant to conservation physiology, and allows for population-level applications for conservation management and policy.

Project description:BackgroundOrange-spotted grouper (Epinephelus coioides) is an economically important marine fish cultured in China and Southeast Asian countries. The emergence of infectious viral diseases, including iridovirus and betanodavirus, have severely affected food products based on this species, causing heavy economic losses. Limited available information on the genomics of E. coioides has hampered the understanding of the molecular mechanisms that underlie host-virus interactions. In this study, we used a 454 pyrosequencing method to investigate differentially-expressed genes in the spleen of the E. coioides infected with Singapore grouper iridovirus (SGIV).ResultsUsing 454 pyrosequencing, we obtained abundant high-quality ESTs from two spleen-complementary DNA libraries which were constructed from SGIV-infected (V) and PBS-injected fish (used as a control: C). A total of 407,027 and 421,141 ESTs were produced in control and SGIV infected libraries, respectively. Among the assembled ESTs, 9,616 (C) and 10,426 (V) ESTs were successfully matched against known genes in the NCBI non-redundant (nr) database with a cut-off E-value above 10-5. Gene ontology (GO) analysis indicated that "cell part", "cellular process" and "binding" represented the largest category. Among the 25 clusters of orthologous group (COG) categories, the cluster for "translation, ribosomal structure and biogenesis" represented the largest group in the control (185 ESTs) and infected (172 ESTs) libraries. Further KEGG analysis revealed that pathways, including cellular metabolism and intracellular immune signaling, existed in the control and infected libraries. Comparative expression analysis indicated that certain genes associated with mitogen-activated protein kinase (MAPK), chemokine, toll-like receptor and RIG-I signaling pathway were alternated in response to SGIV infection. Moreover, changes in the pattern of gene expression were validated by qRT-PCR, including cytokines, cytokine receptors, and transcription factors, apoptosis-associated genes, and interferon related genes.ConclusionThis study provided abundant ESTs that could contribute greatly to disclosing novel genes in marine fish. Furthermore, the alterations of predicted gene expression patterns reflected possible responses of these fish to the virus infection. Taken together, our data not only provided new information for identification of novel genes from marine vertebrates, but also shed new light on the understanding of defense mechanisms of marine fish to viral pathogens.

Dataset Information

Classifying Goliath Grouper (Epinephelus itajara) Behaviors from a Novel, Multi-Sensor Tag.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets