Target Strength Distributions Research Articles

Classification of acoustic survey data: A comparison between seven teams of experts

Acoustic surveys of mixed fish aggregations often rely on manual allocation of acoustic energy to target species, with auxiliary information such as trawl catches and target strength distributions aiding interpretation. However, different teams of experts may use the same auxiliary data differently. In this experiment, seven teams of experts interpreted the same acoustic data, and we compared their classification to target categories to quantify uncertainty in the manual classification process. The teams largely agreed on the total acoustic energy attributable to organisms, but there was significant variation in how the teams split this energy on different target categories. This was caused by differences in applied thresholds for separating plankton and fish, as well as disagreements in species classification. For all target categories, the variation due to teams was lower than the overall variability across acoustic segments, but when scaled up to the number of segments in a typical survey, the team effect either dominated or was of similar magnitude as the segment variability. These results imply a need for further standardisation and uncertainty estimation of expert evaluations in acoustic surveys involving manual interpretation of echo sounder data.

Assessing the size spectra of marine fish communities with hydroacoustics: examining the challenges of abundant schools, diverse assemblages, and variable orientations

Assessing marine fish community size spectra with hydroacoustics is challenging, as communities are diverse, schooling and swim-bladder-less fishes are common, and fish orientation is variable. We developed an approach to examine these challenges and applied it to data from 51 optic–acoustic surveys of fishes at petroleum platforms throughout the US Gulf of Mexico. When in situ target strength (TS; dB re 1 m2) distributions were used to calculate the density (and subsequently abundance) of schooling fishes, fish lengths and size spectra slopes were significantly smaller than in simulated communities and fish community censuses at platforms (i.e., reference datasets). However, acoustic slopes were comparable to reference slopes when simulated TS values (based on species composition) were used to calculate schooling fish abundance. These findings held regardless of whether specific or general models were used to convert TS to length. Fish orientation was not a useful predictor of TS or slope, but may explain why in situ TS measurements from small groups of fishes around schools were unsuitable for abundance calculations. By examining the challenges associated with assessing size spectra with acoustics, this study aids progress towards using acoustic size spectra metrics for ecological inferences.

Stationary hydroacoustics demonstrates vessel avoidance biases during mobile hydroacoustic surveys of alewife in Lake Ontario

Mobile hydroacoustic surveys are routinely used to estimate pelagic fish abundance. In the Great Lakes, alewife are commonly surveyed with mobile hydroacoustics, however, their behavior often has them associated with epilimnetic habitats which increases the potential for vessel avoidance to bias hydroacoustic observations. Abundance estimates from mobile hydroacoustic surveys are typically made using depth and size thresholds to classify targets to species. In hydroacoustic surveys, fish can appear to be smaller when oriented off horizontal axis, as is common with fish displaying vessel avoidance behaviour. This presents a problem where alewives that are diving may appear too small to be correctly classified. We compared alewife swimming behaviours and target strength distributions between stationary up-looking and mobile down-looking hydroacoustics to quantify how vessel avoidance may bias hydroacoustic observations. Alewives from the mobile survey were found to be diving away from the surface, moving faster and in more linear paths than the undisturbed fish from the stationary survey. In the mobile survey, alewives were observed at smaller target strength values than would be expected if boat avoidance was not occurring. Our results suggest that alewife behavioral changes associated with vessel avoidance affect the number and size of targets classified as alewife from mobile hydroacoustic surveys in Lake Ontario.

Broadband measurements of the acoustic target strength of mesopelagic fishes

The mesopelagic zone (∼200–1000 m) is defined by depths where small amounts of sunlight still penetrate but light levels are insufficient to support photosynthetic activity. This region is one of Earth’s largest ecosystems and is home to a diverse community of marine animals. Biomass estimates for mesopelagic regions, based on acoustic measurements, suggest small fishes at these depths may dominate total fish biomass. These estimates, however, are generally made using surface echosounder measurements at 38 kHz. These measurements are subject to numerous confounding factors including unknown species compositions, physiology, target strength distributions as a function of depth, calibration, and low signal-to-noise measurements. To address these challenges, in addition to questions that require non-acoustic technologies, a towed, integrated sensor platform, the Deep-See, was developed to measure the mesopelagic zone at depth (Lavery et al., 2019). This talk focuses on the acoustics package, which includes commercial and custom split-beam channels from 1–500 kHz. System calibrations and target strength spectra are presented for towed depths up to 700 m, where individual targets are still present in relatively high abundance. Target strength measurements as a function of frequency are considered with an emphasis on their relationship to narrowband, surface-based measurements.

Comparing Fish Density and Echo Strength Distribution Recorded by Two Generations of Single Beam Echo Sounders

Hydroacoustic acquisition was performed by means of two different single beam systems, the Simradmodel EK15 from 2015 and the Simrad model EY-M from the 1980s to explore potential differences of fish density and target strength (TS) distribution between EK15 and EY-M-based estimates. The oligotrophic Lake Storsjøen (48 km2) with steep shores, was chosen for the survey. The pelagic fish stock is dominated by whitefish (Coregonus lavaretus L.), and, recently, illegally stocked smelt (Osmerus eperlanus L.), and a low proportion of Arctic charr (Salvelinus alpinus L.). The total density was estimated at two depth layers (18–32 and 32–48 m), and was quite similar for TS ≥ −56 dB: 755 and 498 fish ha−1, respectively for the EK15-based estimates and similarly 766 and 490 fish ha−1 for the EY-M estimate. Target strength distributions were similar for TS > −48 dB. The proportion of single fish detected with EK15 was negatively affected by the long pulse duration. Six acquisitions from 1986 to 2016 showed a dramatic increase of density of TS = −46 to −44 dB echoes (>10 cm) between 2013 and 2016. This was due to the growth of the introduced smelt population.

Effects of differing ground‐truth data, transect design and statistical analysis on the repeatability of hydroacoustic assessments of pollan Coregonus autumnalis pollan

AbstractPollan, Coregonus autumnalis pollan Thompson, requires an appropriate standardised sampling protocol for conservation assessments. The suitability of hydroacoustics with gillnetting was evaluated, and the effects of sampling design, effort and statistical analysis on the repeatability of results were tested. Summer hydroacoustic surveys appear appropriate as pollan were not abundant in acoustically unsampled areas. However, pollan density estimates were significantly affected by the gillnet sampling design used to ground‐truth the acoustic data. Density estimates from the >12‐ m layer were more robust to gillnet design than estimates from the 3‐ to 12‐ m layer. Estimates from different transect designs yielded some statistically significant results. Comparison of transects common to both transect designs suggests that observed differences reflect temporal changes rather than transect design effects. The reduction in the systematic parallel transect effort by half significantly affected target strength distributions but not pollan density estimates. Density estimates were affected by statistical analysis method where pollan density was highest, with geostatistical analysis providing higher estimates than the arithmetic mean. Hydroacoustic assessments may provide a reliable index of pollan abundance over time and between lakes but future research should investigate the effects of temporal and abiotic factors.

Acoustically Measured Distribution and Abundance of Atlantic Menhaden (Brevoortia tyrannus) in a Shallow Estuary in Long Island, NY

Atlantic menhaden (Brevoortia tyrannus) is an euryhaline forage fish abundant along the eastern coast of the USA that seasonally migrates between the open ocean and shallow water estuaries. These fish are known for forming large, dense schools and are important prey for piscivorous fish, seabirds. Following several large fish kills in the late spring of 2015 in the western Peconic Estuary in Long Island, New York, we conducted multiple surveys during the spring, summer, and fall using scientific echosounders to quantify menhaden abundance in this region. Acoustic backscatter data were used to estimate numeric abundance and biomass of menhaden for the western Peconic Estuary. We found that the 2015 fish kills represented between 7 and 18% of the total local population (~ 2.4 million) that were present at the time. Significant changes in target strength distributions during the study suggest that the size classes of menhaden present in the estuary change as fish grow or migrate in and out of the estuary. This study demonstrates that vertical beam active acoustics is a viable method for quantifying forage fish in a very shallow (< 4 m depth) estuary.

Assessing Target Strength, Abundance, and Biomass for Three Commercial Pelagic Fish Species along the East Coast of Peninsular Malaysia Using a Split-Beam Echo Sounder

ABSTRACT Hashim, M.; Aziz, M.F.H.A.; Hassan, R.B., and Hossain, M.S., 2017. Assessing target strength, abundance, and biomass for three commercial pelagic fish species along the east coast of Peninsular Malaysia using a split-beam echo sounder. Decapterus maruadsi (round scad), Megalaspis cordyla (torpedo scad), and Selar boops (oxeye scad) are commercially important small pelagic fish species of Malaysia. Despite their commercial and ecological value, little is known about their target strength (TS), spatial and vertical distributions, and biomass estimates. A split-beam acoustic survey was performed in the east coast of Peninsular Malaysia (ECPM) in March–June 2009. The calibrated echo sounder allowed determination of TS, estimation of fish density (spatial distribution) and biomass, and produced density maps along depth gradients (vertical distribution). The TS–length relationships indicated that the logarithmic TS was proportional to the squared length. Additional water parameters that included salini...

Using Hydroacoustics to Describe Pelagic Fish Distribution in the Penobscot Estuary, Maine

AbstractTemperate estuaries are inherently variable and productive ecosystems that provide nursery habitat, migration pathways, and forage areas for diadromous, estuarine, and marine fish. We used multifrequency scientific echo sounders (SIMRAD EK60 split‐beam, 38 and 120 kHz) to describe the distribution of pelagic fish in the Penobscot River estuary, Maine, in 2012 and 2013. Differences in responses between frequencies were used to distinguish fish from other biota. Acoustic area backscatter from echo integration (sA [m2/nautical mile2], a common measure proportional to fish density) and target strength (TS; dB re 1 m2, an acoustic measure of fish size) distributions varied with season and salinity. Overall, the sA and TS distributions were similar in both years, with detectable spatial and temporal patterns. The highest value of sA occurred in July of both years, when dense schools of fish were detected in higher‐salinity areas of the lower estuary. The middle estuary had high sA values in April both years, particularly in the vicinity of the seawater–freshwater interface. The mixing area in the middle estuary stratum appears to be important fish habitat; we found fish in this area throughout the year. Fish of variable TS were using this mixing zone throughout the survey period. In full freshwater, upstream from the salinity mixing area, sA was generally low. The majority (~77%) of discrete fish detected had TS values less than −42 dB. The TS distributions varied seasonally, with the highest TS measurements occurring more frequently in April and May and the lowest ones occurring most frequently in July and August. This study demonstrates the efficacy of using a mobile hydroacoustic survey to assess pelagic fish distribution in a complex estuary and may provide a template for long‐term monitoring in dynamic estuarine ecosystems.Received November 3, 2016; accepted March 15, 2017 Published online June 20, 2017

Target strength distributions of Pacific sardine schools: Model results at 500 Hz to 10 kHz

Schools of fish can cause interference for long-range active sonars. The degree of interference depends heavily on the target strengths of the schools. However, there are few measurements and limited modeling of school target strengths in the frequency ranges of these sonars. During the summers of 2009 through 2013, a comprehensive set of measurements of the characteristics of Pacific sardine schools was collected off the west coast of the United States. This data set has enabled model estimates of the target strength distributions of these schools to be made at frequencies between 500 Hz and 10 kHz. This paper describes the sardine school data, its use in the modeling, the modeling procedures, and the target strength distributions obtained. The analysis indicates that, given the rates at which high target strengths occurred at all frequencies of interest, Pacific sardine schools could be a significant source of interference for long-range active sonars operating in their vicinity.

Hippocampal damage impairs recognition memory broadly, affecting both parameters in two prominent models of memory

Declarative memory is thought to rely on two processes: recollection and familiarity. Recollection involves remembering specific details about the episode in which an item was encountered, and familiarity involves simply knowing that an item was presented even when no information can be recalled about the episode itself. There has been debate whether the hippocampus supports only recollection or whether it supports both processes. We approached this issue in a relatively theory-neutral way by fitting two prominent models that have been used to describe recognition memory: dual process signal detection and unequal variance signal detection. Both models yield two parameters of interest when fit to recognition memory data. The dual process signal detection model yields estimates of recollection (r) and familiarity (d'). The unequal variance signal detection model yields estimates of the ratio of the variance of target and foil memory strength distributions (σtarget/σfoil) and the difference in the means of the two distributions (d). We asked how the two parameters of each model were affected by hippocampal damage. We tested five patients with well-characterized bilateral lesions thought to be limited to the hippocampus and age-matched controls. The patients exhibited a broad memory deficit that markedly reduced the value of both parameters in both models. In addition, the pattern of results exhibited by the patients was recapitulated in healthy controls as the delay between learning and testing was extended. Thus, hippocampal damage impairs both component processes of recognition memory.

Effects of pre-experimental knowledge on recognition memory

The influence of pre-experimental autobiographical knowledge on recognition memory was investigated using as memoranda faces that were either personally known or unknown to the participant. Under a dual process theory, such knowledge boosted both recollection- and familiarity-based recognition judgements. Under an unequal variance signal detection model, pre-experimental knowledge increased both the variance and the separation of the target and foil memory strength distributions, boosting hits and correct rejections. Thus, pre-experimental knowledge has profound effects on the multiple, interacting processes that subserve recognition memory, and likely in the neural systems that underpin them.

Estimating the target strength of live, free-swimming chokka squid Loligo reynaudii at 38 and 120 kHz

AbstractSoule, M. A., Hampton, I., and Lipiński, M. R. 2010. Estimating the target strength of live, free-swimming chokka squid Loligo reynaudii at 38 and 120 kHz. – ICES Journal of Marine Science, 67: 1381–1391. Estimates of the target strength (TS) of the loliginid squid Loligo reynaudii at 120 kHz were made from ex situ experiments on 38 caged, but free-swimming animals, and at 38 and 120 kHz from 17 in situ experiments on spawning aggregations in the wild conducted from a research vessel and from small boats. The cage experiments suggested that the backscattering cross section is proportional to mantle length (ML), which would not be expected from simple considerations of ML in relation to wavelength. A similar relationship was found during the in situ experiments conducted from the research vessel. The TS estimates from the research vessel agreed well with those from the cage experiments, for both of the two methods used to extract squid TS distributions from the overall TS distributions, but were some 2–3 dB lower than those from the small boats. This suggests that in situ estimates are affected by the vessel and/or the way in which squid react to it. It is concluded that ex situ experiments on L. reynaudii TS are of most value as a complement to in situ experiments, which should be made concurrently while surveying, using systems capable of having higher volume resolution than in the current experiments, to maximize the applicability of the estimates to the survey data.

Seasonal Estimates of Fish Biomass and Length Distributions Using Acoustics and Traditional Nets to Identify Estuarine Habitat Preferences in Barataria Bay, Louisiana

AbstractWe conducted hydroacoustic, gill‐net, and push trawl surveys to quantify changes in habitat‐specific fish size and biomass in shallow (&lt;2‐m) estuarine waters of Barataria Bay, Louisiana, in order to evaluate essential fish habitat. Surveys were conducted monthly between June 2003 and May 2004 among regions located along a north—south salinity gradient. The fish length distributions derived from the gill‐net and push trawl catches showed moderate concordance with the measured target strength distributions, indicating that our integrated approach more effectively characterized the fish community than using only a single gear type would have. Acoustic estimates showed that biomass was highest during fall (mean ± SE; 2.30 ± 0.27 g/m3) and next highest in spring (1.49 ± 0.20 g/m3), with relatively low biomass during summer (0.70 ± 0.14 g/m3) and winter (0.86 ± 0.14 g/m3); pelagic fish biomass from nets was low during winter (53.9 ± 14.9 grams per unit effort [gpue]) but relatively high in fall (846.1 ± 207.2 gpue), spring (774.3 ± 175.5 gpue), and summer (620.3 ± 140.7 gpue). Oyster habitat supported a greater biomass of pelagic fish (acoustic survey: 1.54 ± 0.15 g/m3; gill‐net survey: 467.3 ± 81.0 gpue) than soft‐bottom habitat (acoustic: 0.94 ± 0.11 g/m3; gill‐net: 315.2 ± 54.8 gpue). Among regions, the greatest biomass of pelagic fish was observed at polyhaline stations (acoustic: 1.78 ± 0.19 g/m3; gill‐net: 654.3 ± 136.5 gpue), followed by mesohaline (acoustic: 1.18 ± 0.15 g/m3; gill‐net: 378.5 ± 79.1 gpue) and oligohaline stations (acoustic: 0.82 ± 0.12 g/m3; gill‐net: 228.3 ± 50.2 gpue). Gill‐net biomass was linearly related to the acoustic biomass estimates of small pelagic fish. The complementary, multigear approach proved to be useful in evaluating habitat use and may be particularly helpful in identifying and monitoring ecosystem reference points to evaluate change and in standardizing ecosystem‐based assessment approaches.

The acoustic properties of Salpa thompsoni

Abstract Wiebe, P. H., Chu, D., Kaartvedt, S., Hundt, A., Melle, W., Ona, E., and Batta-Lona, P. 2010. The acoustic properties of Salpa thompsoni. – ICES Journal of Marine Science, 67: 583–593. Aggregations of the salp Salpa thompsoni were encountered during the Antarctic krill and ecosystem-studies cruise on the RV “G.O. Sars” from 19 February to 27 March 2008. The salp's in situ target strength (TS), size, number of individuals in aggregate chains, and chain angle of orientation were determined. Shipboard measurements were made of Salpa thompsoni's material properties. Individual aggregates were mostly 45.5–60.6 mm in mean length; relatively rare solitaries were ∼100 mm. Chains ranged from 3 to at least 121 individuals, and in surface waters (&lt;20 m), they showed no preferred angle of orientation. Sound-speed contrast (h) ranged from 1.0060 to 1.0201 and density contrast (g) estimates between 1.0000 and 1.0039. The in situ TS distributions peaked between −75 and −76 dB at 38 kHz, with a secondary peak at approximately −65 dB. TS ranged between −85 and −65 dB at 120 and 200 kHz and peaked around −74 dB. The measured in situ TS of salps reasonably matched the theoretical scattering-model predictions based on multi-individual chains. The backscattering from aggregate salps gives rise to TS values that can be similar to krill and other zooplankton with higher density and sound-speed contrasts.

Hydroacoustic Target Strength Distributions of Alewives in a Net‐Cage Compared with Field Surveys: Deciphering Target Strength Distributions and Effect on Density Estimates

AbstractAcoustic estimation of fish abundance requires knowledge of the target strength (TS) distribution associated with the fish present. We measured the TS distribution at 70 kHz for five size‐groups of alewives Alosa pseudoharengus (mean total length = 68‐138 mm) swimming freely in a large net‐cage. Although the mean and mode of the distributions were significantly related to fish length, the TS range obtained from each length‐class was over 25 decibels (dB). The TS distributions were negatively skewed and had a tendency toward bimodality. The TS distributions at 70 and 123 kHz were similar for one alewife size‐group that was measured with both frequencies. Small but inconsistent differences were found between day and night for the same groups of fish. Shapes of the TS distributions were similar when centered on the mode, and we derived a probability density function (PDF) for TS at 70 kHz as a function of fish length to describe these distributions. We used this PDF to predict TS distributions based on alewife gill‐net catch from six lake surveys and compared predictions with in situ TS distributions from each acoustic survey. Observed and predicted distributions were similar for TS values greater than −50 dB, but the observed number of small TS values exceeded the predicted number for several of the surveys. Some of these smaller TS values were probably from invertebrates or smaller fish or were indicative of noise. Estimates of alewife density from lake surveys were highly affected by inclusion or exclusion of these smaller targets. We used the predicted TS distribution derived either from catch data or from the TS values greater than −50 dB to calculate the proportion of smaller TS values that were probably from alewives. Both methods resulted in similar abundance estimates, which were higher than estimates based on an existing TS‐length regression for alewives.

Diel changes in acoustic and catch estimates of krill biomass

Abstract Simard, Y., and Sourisseau, M. 2009. Diel changes in acoustic and catch estimates of krill biomass. – ICES Journal of Marine Science, 66: 1318–1325. Krill-biomass estimates can be compromised by diel variabilities in acoustic backscatter and the catch efficiencies of various nets. This paper describes an effort to quantify these variabilities at fine temporal and spatial scales during a three-day experiment at a fixed location, using high-resolution, stratified Bioness samples and echo-integration, and assuming a fixed distribution of krill orientations. Night-time catches in the krill scattering layer (SL) were 15 times the acoustic estimates. The situation was reversed during daytime, when the acoustic estimates in the SL were 5 times larger than the catches. This collectively resulted in a ±10‐dB gradual diel cycle in the difference of vertically integrated biomass from both sampling methods. Use of a strobe light on the Bioness reduced avoidance of the net by krill and significantly increased (×10) daytime catches in the SL, but had no significant effect on night-time catches. The difference in volume-backscattering strength at 120 and 38 kHz (ΔSv120–38) in the densest parts of the SL agreed with predictions using a target-strength (TS) model and an assumed normal distribution of tilt (mean θ = 11°; s.d. = 4°). The ΔSv120–38 was smaller for lower densities and during night-time. It appears that the θ and, therefore, TS distributions of krill significantly change during their diel vertical migrations. At twilight and at night, when they are feeding and swimming vertically, they exhibit lower mean TS and ΔSv120–38 and react less to strong strobe-light pulses, in contrast to daytime. Diel patterns in TS and net avoidance should be taken into account in krill-biomass assessments that use round the clock acoustic-survey data and multi-frequency TS models for target classification.

Hydroacoustic assessment of fish abundance in the presence of gas bubbles

Methane ebullition from bottom sediments is abundant in many productive water bodies. The occurrence of the methane bubbles in the water column may cause erroneous quantification of fish densities in aquatic ecosystems when using hydroacoustic methods, since the target strength (TS) of bubbles strongly overlaps with that of fish. In this study, an approach that allows separation of TS distributions of fish and bubbles is proposed. The method was applied to data collected in a shallow reservoir with a 120‐kHz echo sounder. Separation between fish and bubbles was based on the difference in their trajectories on echograms. To distinguish the target trajectories, the water column was sampled at low boat speed at restricted locations. Sampling was also carried out at high boat speed to assess the abundance and TS distribution of all targets in the entire reservoir. The information from the two different sampling modes was integrated in a simple mathematical algorithm, which allowed calculating bubble and fish densities in the reservoir. The total abundance of gas bubbles was higher than the density of fish. The elaborated method can be useful for quantification of dominant objects when spatial distributions are heterogeneous.

Measuring fish abundance in a weir trap using an acoustical-optical platform

Data recorded by a bottom mounted survey platform deployed within the opening of a fishing weir were used to calculate species specific abundance estimates for comparison to the weir catch. Abundance estimates were calculated from the combination of sonar and video information recorded by the Acoustical-Optical Platform (AOP). Echo counting was used to detect individual moving targets in the sonar images with the application of a background removal technique utilizing a moving average filter. Video images provided species identification of acoustic targets. Video images and differences in target strength distributions reflected a change in dominant species from each deployment which was confirmed with the weir catch. The algorithm used to calculate AOP abundance estimates was most accurate in predicting abundance for species comprising at least 13% of the overall catch by weight. Close agreement between the species specific AOP estimates and absolute abundances of each species suggests that the combination of acoustic and video data is a powerful combination for accurately identifying fish species and predicting abundance.

Evaluation of an acoustical-optical platform for measuring species specific abundance in a weir trap

Data recorded by a bottom mounted survey platform deployed within the opening of a fishing weir were used to calculate absolute abundance estimates for comparison to the weir catch. Species specific abundance estimates were calculated from a combination of sonar and video information recorded by the acoustical-optical platform (AOP). Individual targets were detected in the sonar images using an echo counting technique. Video images were used to provide species identification of acoustic targets within an approximate 10 m range. The number and identification of acoustic targets detected during each deployment was positively correlated with the total catch weights and dominant species in each deployment. The algorithm used to calculate AOP abundance estimates tended to underestimate abundance when more than 50 individuals of the same species were caught and overestimate abundance when less than 50 individuals were caught. Target strength, bearing, and range distributions for individual targets are presented in addition to the species specific percent difference between the AOP estimates and absolute abundances of the weir catch. The agreement between the species specific AOP estimates and absolute abundances of each species indicate that the AOP has the potential to provide data for accurate estimates of multispecies abundance. [Work supported by NOAA.]