Quantification of Walleye Spawning Substrate in a Northern Minnesota River using Side-Scan Sonar

Evaluating aquatic habitats is an important component of many ecological studies and natural resource assessments, but traditional habitat evaluations are time and labor intensive and do not provide continuous data. Side-scan sonar (SSS) can provide a low-cost method that collects continuous aquatic habitat data. We used SSS mapping to quantify suitable spawning substrate available toWalleye Sander vitreus during the 2015 spring spawningmigration in a 10.8-km reach of the Tamarac River, Minnesota. The SSS map had 78.0% agreement with reference points classified in the field, and the proportion of reference points predicted as suitable using the SSS map was not significantly different than the proportion of reference points observed to be suitable. Suitable substrate forWalleye spawning comprised 8.4% (26,392 m) of the total area mapped. The estimated number of females that suitable substrate could support was lower than the number that likely migrate up the Tamarac River and suggests that access to spawning substrate may sometimes limit reproductive success. This study demonstrates that a relatively inexpensive SSS unit can be used to efficiently map aquatic habitat while acquiring quantitative and qualitative data. The assessment of aquatic habitats provides useful information for a variety of ecological applications. Assessing the components of abiotic habitat present in a system, and where specific conditions occur, yields insight into biotic community structure and locations where certain taxa are likely to be found (Knapp et al. 1998; Jackson et al. 2001; Bornette and Puijalon 2011). Further, locating suitable habitat aids in estimating the ability of certain taxa to grow, develop, survive, and reproduce (Knapp et al. 1998; Jackson et al. 2001; Hafs et al. 2014). The availability of suitable spawning habitat is an important component of reproduction in fish, with reduced quantity and quality of spawning habitat linked to reduced egg deposition, egg survival, age-0 abundance, and recruitment to the adult population (Johnson 1961; Dombeck et al. 1984; Knapp et al. 1998). Traditionally, river habitat has been evaluated and mapped using techniques that consist of taking measurements visually or physically at discrete points, transects, or grids throughout the river (e.g., Wright et al. 1981; Fitzpatrick et al. 1998; Maddock 1999; Hafs and Gagen 2010). These techniques can require extensive time and effort in thefield and result in a trade-off between the level of detail and the size of the area being mapped (Maddock 1999). Furthermore, traditional techniques typically do not produce continuous data; thus, areas not evaluated must be inferred from areas that have been evaluated. The evaluations may also be limited by environmental conditions, including water depth and transparency. Techniques have been developed that use relatively inexpensive recreational side-scan sonar (SSS) units and GIS software to effectivelymap and evaluate freshwater habitats (Kaeser andLitts 2008, 2010; Kaeser et al. 2013). Side-scan sonar provides continuous *Corresponding author: [email protected] Present address: Boise State University, 1910 University Drive, Boise, Idaho 83725-1535, USA. Received July 6, 2016; accepted November 5, 2016 420 North American Journal of Fisheries Management 37:420–428, 2017 © American Fisheries Society 2017 ISSN: 0275-5947 print / 1548-8675 online DOI: 10.1080/02755947.2017.1280568 data and is not limited by deep or turbid water. Previous freshwater SSS studies have evaluated and/or quantified variables, including sedimentation (Manley and Singer 2008), large woody debris (Kaeser and Litts 2008), substrate type (Kaeser and Litts 2010; Kaeser et al. 2013), fish abundance (Barton 2000), and fish spawning habitat (Edsall et al. 1989;Walker andAlford 2016). One of the most recent SSS studies mapped and accurately classified spawning habitat for Walleye Sander vitreus in Wisconsin lakes (Richter et al. 2016). When using SSS to evaluate substrate type in freshwater habitats, the reported overall accuracy ranges from 29% to 93% (Kaeser et al. 2013; Richter et al. 2016; Walker and Alford 2016). Kaeser and Litts (2010) reported time in the field for data collectionwhen rivermapping usingSSS to be 11min/km and total map production time to be ~3 h/km, which was about one tenth of the time required when using transect-based techniques. Although these studies suggest that SSS has distinct advantages when mapping freshwater habitats, not all studies in the literature provide accuracy assessments and studies that produce SSS maps directed toward a specificmanagement interest remain sparse. Thus, studies evaluating SSS habitat mapping techniques for freshwater fish in systems with varying morphology are necessary to test the broad applicability of the technique and inform future researchers. Mapping spawning habitat requires an understanding of the environmental variables that constitute suitable spawning habitat. The U.S. Fish and Wildlife Service synthesizes habitat suitability indices for many game and nongame species based on compiled literature and expert reviews (Schamberger et al. 1982). The habitat suitability index constructed for Walleyes suggests that the quality and quantity of suitable substrate is one of the primary factors affecting reproductive success (McMahon et al. 1984). The substrates most suitable for Walleye spawning are those with diameters between 2.0 and 250.0 mm (McMahon et al. 1984), which are categorized as gravel, pebble, and cobble using the sediment classification scheme from Wentworth (1922). Egg production in female Walleyes is proportional to body mass, with a typical production of 60,000 eggs/kg (Nickum 1986) and a mean egg diameter of 2.0 mm (Smith 1941; Colby et al. 1979). The stacking of eggs has been shown to reduce fertilization rates in laboratory settings (Moore 2003) and would likely increase the chance of transport and ensuing siltation, abrasion, and predation, which are the presumed causes for reduced egg survival on finer substrates (Bozek et al. 2011). Literature estimates ofWalleye egg densities range from 65 to 7,047 eggs/m (Johnson 1961; Corbett and Powles 1986; Manny et al. 2007). The Walleye fishery of the Red Lakes, Minnesota, is economically important and supports popular recreational and commercial fisheries. For example, combined recreational and commercial harvest exceeded 770,000 Walleyes in 2015 (Brown and Kennedy 2016). The fishery is currently near record-high levels of Walleye abundance and is completely supported by natural reproduction (Kennedy 2016). Although spawning migrations have only been quantified in the Tamarac River, the largest tributary to Upper Red Lake, it is presumed (based on local knowledge) that this tributary supports the largest Walleye spawning migration. A hatchery was previously operated at the Tamarac River, where the mean Walleye catch between 1932 and 1979 was 203,066 individuals, with catch reaching as many as 646,161 fish (Groshens 2000). However, these collections did not capture the entire duration of the migration because nets were often disabled once holding pens were filled to capacity, and therefore the true migration size was larger. Hatchery operations at the Tamarac River, which provided total catch data, have been suspended since 1979, but annual electrofishing surveys are currently conducted during spawning migrations. The mean electrofishing survey catch rate over the last 10 years was 638 Walleye/h (Brown and Kennedy 2016). The magnitude of historic migrations and current electrofishing catch rates provide evidence to suggest that the Tamarac River may be an important component of Red Lakes Walleye reproduction. Two previous cursory evaluations by Minnesota Department of Natural Resources staff have been conducted to assess Walleye reproduction in the Tamarac River (Fraune and Scidmore 1963; Groshens 2000). However, reproductive success in the Tamarac River is largely unknown and in-depth assessments on the quantity and quality of spawning habitat in the river are lacking. The ability to acquire accurate and continuous substrate data regardless of water depth or clarity makes SSS mapping an ideal technique to describe and quantify Walleye spawning substrate in the Tamarac River. Our objectives were to (1) quantify suitable Walleye spawning substrate in the 10.8-km reach accessible toWalleyes in the Tamarac River during spring 2015, (2) assess the accuracy of the SSS map, (3) estimate the number of females that could be supported by available suitable substrate, and (4) evaluate the implications. METHODS Study site.—The Red Lakes cover 116,550 hectares and comprise the largest body of water contained within Minnesota borders (Figure 1). The Tamarac River is the main tributary to the upper basin of the Red Lakes and flows 34.9 km into the northeastern corner of the basin (Groshens 2000). The river’s drainage encompasses 815 km, including a portion of a 1,295-km peat bog, the largest in the lower 48 states (Groshens 2000). The drainage is primarily wetlands, but approximately 35% of the watershed is forested (Groshens 2000). The river has a low gradient with substrates ranging from silt to cobble but is dominated by silt and sand (Fraune and Scidmore 1963; Groshens 2000). Water depths near the mouth can exceed 3 m, but the majority of the river is typically 1 m deep. Beaver Castor canadensis dams are common in the river and have blocked Walleye migrations in the past (Fraune and Scidmore 1963; Groshens 2000). This study focused on the 10.8-km reach from the river mouth up to a large beaver dam (Figure 1). QUANTIFICATION OF WALLEYE SPAWNING SUBSTRATE 421