A10. Acoustic camera

From PRAGMO
Jump to navigation Jump to search
Example of an acoustic camera location © Guillard et Lebourges-Dhaussy, 2014. Pour la science n°436, Février 2014

>> Back to Fisheries Surveys <<

1. What it can be used for

  • Study fish travel speed and direction
  • Monitor fish stock abundance on migration ways after stream recolonization after dam or weir removal
  • Assess fish behavior even at night, in turbid water, or facing obstacles
  • Assess fish behavior facing dams and provide management proposal, with high temporal and spatial resolution.
  • Assess timing and extent of movements of different class size of fish relating to environmental factor

2. The method

Acoustic camera (belonging to the multi-beam sonar or imaging sonar) uses high frequency acoustics beam to create a 2D image covering horizontally the river width. An acoustic camera is a concentrate of technology in a high frequency soundwaves emitting box that lead to an image at the interface of a video and a radar allowing us to see the fish morphology and size (Martignac et al., 2015), know its horizontal position in the river, record its behavior direction and process to species identification (Martignac, Hemon and Blancher, 2015) . The device is light and generates a large amount of data. Specialised software is necessary for beam interpretation to generate a video-like image which can be used to reduce the time-consuming process of visualizing the images and identifying fish. Location of the device has to be well chosen and tailored to each site configuration. Several studies exist, for examples on salmon population monitoring (Maxwell and Gove, 2004) or on downstream migration of eels (Hateley and Gregory, 2006).

Acoustic camera images. Left to right : Silurus glanis, Anguilla anguilla, Alosa alosa. © François Martignac. OFB- INRAE

3. Advantages

*Benefits and limitations can be variable according to the camera model*

  • Non-intrusive method because no handling process or fish disturbance
  • Length measurement and visual observation of fish morphology and behavior even for eels and lamprey (Lenihan, McCarthy and Lawton, 2019) that can help identifying species
  • Can provide data on non-targeted species, behaviors of interaction, erratic or lost individuals in fishways, juvenile school passages etc…
  • Species identification depending on the fish community and morphology characteristics of the targeted species
  • Some software (that aren’t always from the constructor), but not all allow an automatic tracking function. It has reduced by 3 the time for watching sequences in some studies (Daroux, Martignac and Guillard, 2014)
  • Some gears can provide images up to 100 m
  • Can be used on long-term study
  • Easily removable
  • Some background on its use (Atencio and Reichmuth, 2013; and references below)

4. Disadvantages

*Benefits and limitations can be variable according to the camera model*

  • No tools for automatic data analyzing, so very time consuming
  • No 3D position yet. When horizontally orientated, “x” (in which beam the fish is detected) and “z” (distance between the fish and the camera) can be acquired, but not “y” depth of the fish in the water column.
  • Only partial covering of the water column (depending on site shape and depth) and limited covering width
  • Water turbulences and bubbles decrease effective range and increase accuracy of fish size and number (Kock, Perry and Hansen, 2016)
  • Accurate size information only on individuals bigger than 20 cm. Daroux et al., (2019), have observed and over estimation of length for smallest fish (L T < 57 cm) and a overestimation for the larger ones (L T > 57 cm) but only an average error of 2.7cm between measures (5 averaged) and the real size of the fish (study with a DIDSON camera).
  • Huge amount of generated data (sometimes 20 Go for a full day of recording) requiring high storing capacity or to think about a sub-sample protocol (Daroux, Martignac and Guillard, 2014). The amount of data is variable according to the camera model, length of recording, frequency of recording, which frequency is used, and frame rate (Martignac 2020, com. pers.)
  • Several post-processing software need to be used (Hateley, 2005)
  • Fish can be counted two times or more if they don’t have a proper migration behavior and turn around erratically, they have to be actively migrating (Maxwell, 2004)
  • Bias can appear in length measurement between trained and beginners operator (Daroux et al., 2019)
  • Species identification can be hard in stream where the fish community is diversified (Guillard, 2020 com. pers.).

5. Recommendations for method application

Camera and beam direction. © François Martignac. OFB- INRAE
  • The monitoring location has to be downstream from the main spawning grounds and upstream from tide effect (Pipal et al., 2010).
  • The hydromorphology of the site must be in accordance with a acoustic beam shape to avoid acoustic shadow (blind spot) because of depth, width, braided channel and shoal presence. (Figure 2 and 3)
  • Do a precise beam mapping (estimate volume = area of water and depth covered) with a highly detectable object.
  • Keep the camera position fixed in lotic streams
  • If the objective is to record fish migration speed, make sure to choose an obstacle-free area where fish can move linearly
  • Carefully chose the study place to avoid as much as possible water turbulence and bubbles
  • Be careful of misidentification, an adapted protocol has to be set depending on fish community as morphology or behavior can be similar between two species or more
  • Ideally repeat 3 to 5 measurement on each fish to get a more accurate length estimation (Daroux et al., 2019)
  • Choose the device based on site configuration (river width) but also regarding the compatibility with time-saving software.
  • This method alone can not state on population density
  • It’s easier to choose a site close to a road or rapidly accessible in case of flood or other issues and for maintenance (Pipal et al., 2010).
  • Think about electric and internet access and protect the material against vandalism

6. Cost

The cost of these highly specialized units is generally high but a large range of prices exists depending on the constructor Sound Metrics Corp (Didson and ARIS camera), Teledyne (Blueview camera), Tritech (gemini camera), blueprint (oculus camera). Prices range is between 20 000 and 110 000 pounds.

7. Protocol and data analysis

8. Acknowledgement and associated authors

François Martignac - INRAE UMR ESE - Rennes
Jean Guillard - INRAE UMR CAARTEL - Thonons-les-Bains

9. Further reading

  • Atencio, B. and Reichmuth, M. (2013) Monitoring of Adult Coho Salmon and Steelhead Trout Using Dual-frequency Identification Sonar (DIDSON) in Lagunitas Creek, California 2013 Annual Report. Available at: https://irma.nps.gov/DataStore/DownloadFile/504815.
  • Daroux, A. et al. (2019) ‘Manual fish length measurement accuracy for adult river fish using an acoustic camera (DIDSON)’, Journal of Fish Biology, 95(2), pp. 480–489. doi: 10.1111/jfb.13996.
  • Daroux, A., Martignac, F. and Guillard, J. (2014) ‘Utilisation de la caméra acoustique DIDSON pour le suivi en rivière des poissons migrateurs’, Le Cahier des Techniques de l’INRA, 2014(83), pp. 1–8.
  • Hateley, J. (2005) Assessment of acoustic post-processing software products for fisheries surveys, Environment Agency.
  • Hateley, J. and Gregory, J. (2006) ‘Evaluation of a multi-beam imaging sonar system ( DIDSON ) as Fisheries Monitoring Tool : Exploiting the Acoustic Advantage . Technical Report’, p. 7.
  • Kock, T. J., Perry, R. W. and Hansen, A. C. (2016) Survival of Juvenile Chinook Salmon and Coho Salmon in the Roza Dam Fish Bypass and in Downstream Reaches of the Yakima River, Washington, 2016.
  • Lenihan, E. S., McCarthy, T. K. and Lawton, C. (2019) ‘Use of an acoustic camera to monitor seaward migrating silver-phase eels (Anguilla anguilla) in a regulated river’, Ecohydrology and Hydrobiology. Ecohydrology & Hydrobiology, 19(2), pp. 289–295. doi: 10.1016/j.ecohyd.2018.07.001.
  • Martignac, F. et al. (2015) ‘The use of acoustic cameras in shallow waters: New hydroacoustic tools for monitoring migratory fish population. A review of DIDSON technology’, Fish and Fisheries, 16(3), pp. 486–510. doi: 10.1111/faf.12071.
  • Martignac, F., Hemon, A. and Blancher, P. (2015) ‘Hydroacoustique et poissons migrateurs amphihalins : analyse innovante de leur dynamique migratoire’, Sciences Eaux & Territoires, pp. 18–23.
  • Maxwell, S. L. (2004) ‘Hydroacoustics : Rivers’, Rivers, pp. 153–172.
  • Maxwell, S. L. and Gove, N. E. (2004) ‘the Feasibility of Estimating Migrating Salmon Passage Rates in’, Fisheries (Bethesda), (March).
  • Pipal, K. et al. (2010) ‘Operation of dual-frequency IDentification SONar (DIDSON) to monitor adult steelhead (Oncorhynchus mykiss) in the central California coast’, NOAA Technical Memorandum NMFS, (October), pp. v–57.