Using Multiple Gears to Assess Acoustic Detectability and Biomass of Fish Species in Lake Superior

Abstract

Recent predator demand and prey supply studies suggest that an annual daytime bottom trawl survey of Lake Superior underestimates prey fish biomass. A multiple‐gear (acoustics, bottom trawl, and midwater trawl) nighttime survey has been recommended, but before abandoning a long‐term daytime survey the effectiveness of night sampling of important prey species must be verified. We sampled three bottom depths (30, 60, and 120 m) at a Lake Superior site where the fish community included all commercially and ecologically important species. Day and night samples were collected within 48 h at all depths during eight different periods (one new and one full moon period during both early summer and late summer to early fall over 2 years). Biomass of demersal and benthic species was higher in night bottom trawl samples than in day bottom trawl samples. Night acoustic collections showed that pelagic fish typically occupied water cooler than 15°C and light levels less than 0.001 lx. Using biomass in night bottom trawls and acoustic biomass above the bottom trawl path, we calculated an index of acoustic detectability for each species. Ciscoes Coregonus artedi, kiyis C. kiyi, and rainbow smelt Osmerus mordax left the bottom at night, whereas bloaters C. hoyi stayed nearer the bottom. We compared the biomass of important prey species estimated with two survey types: day bottom trawls and night estimates of the entire water column (bottom trawl biomass plus acoustic biomass). The biomass of large ciscoes (>200 mm) was significantly greater when measured at night than when measured during daylight, but the differences for other sizes of important species did not vary significantly by survey type. Nighttime of late summer is a period when conditions for biomass estimation are largely invariant, and all important prey species can be sampled using a multiple‐gear approach.

Can pelagic forage fish and spawning cisco (Coregonus artedi) biomass in the western arm of Lake Superior be assessed with a single summer survey?

“…Limited harvest also occurs during summer with catches sold for human consumption. Recent research on Lake Superior has demonstrated that bottom-trawl sampling provides a poor measure of cisco spawning stocks, and that acoustics combined with midwater trawl sampling is a better approach for estimating spawner densities Yule et al, 2006Yule et al, , 2007. Moreover, sampling has shown that mature cisco utilize offshore waters, at least to some degree, during spring and summer (Hrabik et al, 2006).…”

Status of pelagic prey fishes in Lake Michigan, 2013

“…Similarly, night bottom trawl estimates of rainbow smelt density were ≈ 6% of day estimates. Evidence suggests however that bloater tend to be more demersal; in Lake Superior, night acoustic/midwater trawl sampling may detect only 60% of bloater present (Yule et al 2007). Day-night bottom trawl data from Lake Michigan in 1987 suggested that the availability of bloater to acoustic sampling at night was somewhat higher (mean = 76%, D. M. Warner, unpublished data).…”

“…The initial three strata were retained, but their size was modified based on data collected in 2003 as well as NOAA Coast Watch Great Lakes node maps of sea surface temperature from 2001-2003. In 2007, the number of transects in each stratum was optimized based on stratum area and standard deviation of biomass using methods in Adams et al (2006).…”

Towards a balanced presentation and objective interpretation of acoustic and trawl survey data, with specific reference to the eastern Scotian Shelf

“…In Lake Superior, the pelagic and semipelagic fish species are managed using abundance estimates from bottom-trawl surveys, but observations using midwater trawling and acoustic methods emphasized inconsistencies in these estimates between bottom-trawl and acoustic methodologies and potentially deficiencies in the estimated population trends. In a series of papers by Stockwell et al and Yule et al (Stockwell et al 2006, 2009Yule et al 2007Yule et al , 2008, the quantitative process of evaluating different trawl gear and acoustic instrumentation for detecting and enumerating pelagic fish species in the Laurentian Great Lakes and then combining these data streams is exemplified. The process began with quantifying differences in estimates between bottom-trawl and acoustic methods, continued with studies and experiments comparing and assessing the sampling methodologies, and eventually concluded with the development of sampling and analytical techniques to combine the seemingly disparate data streams in ways to effectively sample the ecosystem and improve understanding of trophic interactions and management strategies.…”