Underwater video and still-image dataset of fishes and other aquatic animals in Lake Biwa, Japan, observed via carp-mounted video loggers.

Study Extent

The study area was located in the northern part of the North Basin of Lake Biwa, Japan, and was enclosed within the following coordinates: 35.287–35.454 N, 136.018–136.225 E. This study was conducted from Oct 2016 to Dec 2018. Samplings were made through five field experiments as follows: (1) an experiment in Oct 2016, (2) another experiment in Nov 2017, and (3) three experiments from Nov to Dec 2018.

Sampling

Samplings were made through five field experiments using nine fish as follows: (1) an experiment using one fish in Oct 2016, (2) another experiment using one fish in Nov 2017, and (3) three experiments using six fish from Nov to Dec 2018. Animal-borne video loggers (DVL200L; 27 mm width, 116 mm length, 10 mm height and 57 g in air, or DVL400M130; 21 mm width, 68 mm length, 22 mm height and 47 g in air, Little Leonardo, Tokyo, Japan) were used to collect video footage underwater. The logger was molded into a float with an accelerometer (ORI400-D3GT; 12 mm diameter, 45 mm length and 9 g in air, or W190-PD3GT; 21 mm diameter, 116 mm length and 60 g in air, Little Leonardo), a VHF transmitter (MM-130B; 16 mm diameter, 60 mm length and 20 g in air, Advanced Telemetry Systems, MN) and a time-release system (RT4; 16 mm diameter, 25 mm length and 16 g in air, Little Leonardo), and was then mounted on the back of each carp using a plastic cable tie. The float was shaped to reduce drag and it provided just enough buoyancy to return the instrument package to the surface upon release. All carp were released near the site at which they were captured. The video loggers were set to start recording in the morning (6:00 or 7:00 JST) three (or five) days after the fish’s release. The recording duration was 6 hours for DVL200L and 11–12 hours for DVL400M130. On the fourth (or sixth) day, the float detached from the fish, floated to the water surface, and was retrieved.

Quality Control

Scientific names followed the GBIF backbone taxonomy (https://www/gbif.org/species/) and the Union of Japanese Societies for Systematic Biology (2003). To identify avian species in the study area, daily occurrence reports of birds, provided by Kohoku Wild-Birds Center, (Bird News; http://www.biwa.ne.jp/~nio/newsindex.html) were used as a reference.

Method steps

1. The video data was formatted in MP4 (3 or 6 Mbps, 30 fps) format. A total of 64 h of video was taken. Images that contained any aquatic animals were cropped from obtained video footage using iMovie software (ver.10.1.9, Apple, CA). Color values of the images were adjusted using the “Automatic Color Adjustment” tool in the software.
2. All species were identified by the authors or by professional experts. If we could not obtain sufficient information to identify a species from the videos (e.g., a blurring due to fish movements), we classified the organism to a higher taxonomic level (e.g., order, class, and family) that could be specified with certainty.

This dataset includes occurrence records of 11 fish species, three avian species and a crustacean.

1. Cyprinus
   rank: genus
2. Carassius
   rank: genus
3. Hemibarbus
   rank: genus
4. Pseudogobio
   rank: genus
5. Rhinogobius
   rank: genus
6. Micropterus
   rank: genus
7. Silurus
   rank: genus
8. Gymnogobius
   rank: genus
9. Lepomis
   rank: genus
10. Gobiidae
    rank: family
11. Aves
    rank: class
12. Caridea
    rank: infraorder

Lake Biwa in Shiga, Japan