White bass population demographics in a northwestern South Dakota reservoir

Abstract

The white bass (Morone chrysops) is an important sport fish species in the upper Midwest. As such, understanding white bass dynamics is critical for managing its populations. Thus, we evaluated population dynamics of white bass in a northwestern South Dakota reservoir and attempted to determine whether bass recruitment patterns were more related to reservoir inflows or spring weather patterns. We found that white bass growth increased from 243 mm for age-2 fish, to 322 mm for age-6 fish, and then to 346 mm at age 9. The total annual mortality rate was 29.7%. In terms of year-class strength, white bass recruitment was erratic. The recruitment pattern was more related to spring weather patterns than reservoir inflow. In comparison, white bass population dynamics in this South Dakota irrigation reservoir were more similar to those of white bass populations in Midwestern glacial lakes than those found in southern US reservoirs.

Keywords:

• white bass
• population demographics
• Akaike's information criterion
• recruitment

Introduction

Thorough knowledge of population dynamics (i.e., recruitment, growth, and mortality) is essential for understanding a fish population and the subsequent community. For example, erratic recruitment patterns can be indicative of the lack of suitable spawning habitat or adverse climatic characteristics. Growth can be used to evaluate prey resources or provide insight into population abundance and the balance between prey and predators. Mortality estimation characterizes population longevity, which is affected by both natural and harvest mortality.

The white bass (Morone chrysops) is a predatory fish that often provides a substantial component of the total sport fish harvest in Missouri River reservoirs and eastern glacial lakes in South Dakota (Willis et al. 2002). White bass recruitment (i.e., year-class strength) has been positively related to high spring inflows in three Virginia reservoirs (DiCenzo and Duval 2002) and Missouri River reservoirs in South Dakota (Beck et al. 1997). In contrast, Pope et al. (1997) noted that white bass year-class strength was positively related to spring precipitation and air temperature in eastern South Dakota glacial lakes. However, these relationships have not been evaluated in irrigation reservoirs in the upper Midwestern United States. Thus, the objectives of this study were to: (1) evaluate population dynamics of white bass in a northwestern South Dakota reservoir and (2) determine whether white bass recruitment patterns were more related to reservoir inflows or spring weather patterns.

Materials and methods

We conducted our study at Shadehill Reservoir (SR), a 1899 ha reservoir located on the Grand River in Perkins County, northwestern South Dakota, USA. Maximum and mean depths of SR were 18.3 and 6.7 m, respectively, the shoreline development index was 5.2, and the lake was considered mesotrophic based on the trophic state index (Stueven and Stewart 1996). The SR fish community consisted of black crappie (Pomoxis nigromaculatus), bluegill (Lepomis macrochirus), channel catfish (Ictalurus punctatus), common carp (Cyprinus carpio), emerald shiner (Notropis atherinoides), gizzard shad (Dorosoma cepedianum), northern pike (Esox lucius) smallmouth bass (Micropterus dolomieu), spottail shiner (Notropis hudsonius), walleye (Sander vitreus), white bass, white sucker (Catostomus commersonii), and yellow perch (Perca flavescens). Importantly, the white bass population at SR was self-sustaining.

We collected white bass on 9–10 August 2005 using four, 91.4 m experimental gill nets composed of four equal-sized panels of 12.7, 25.4, 38.1, and 50.8 mm mesh bar measure. We set gillnets using a 24 h soak time at fixed sites used by South Dakota Game, Fish and Parks for fish population assessments. We recorded total length (mm) and weight (g) for each fish collected. We removed sagittal otoliths from 5–10 fish per centimeter length group and placed them in vials to dry for 2 weeks prior to annulus determination and enumeration. We used a subsample of 56 white bass for aging purposes, and then employed an age-length key (DeVries and Frie 1996) to correct the subsample to an assumed age structure based on the length frequency for the entire 257 fish sample.

Following age assignments to individual white bass, we calculated mean length-at-age at the time of capture by cohort to assess growth rates; we compared these data to a statewide average back-calculated length-at-age for white bass (Willis et al. 1997). We estimated total annual mortality from the white bass age structure using a weighted regression analysis for a catch curve (Miranda and Bettoli 2007).

Our quantification of year-class strength followed the residual method proposed by Maceina (1997). We indexed year-class strength by computing the residuals from the catch-curve regression analysis. We used a weighted regression in the catch-curve analysis to deflate the influence of older and/or rarer year classes (Maceina and Pereira 2007). Stronger year classes were identified with positive residuals, and weaker year classes were identified with negative residuals. We used the Fishery Analyses and Simulation Tools (Slipke and Maceina 2002) software to calculate catch-curve residuals and the total annual mortality rate using the weighted (by number) regression option.

To determine the potential relationships between the year-class residuals and environmental data, we obtained air temperature data for northwestern South Dakota from the National Oceanic and Atmospheric Administration (NOAA 2009) and precipitation and reservoir inflow data from the US Bureau of Reclamation (USBOR 2009). We used reservoir inflow, May and June precipitation, and May and June mean air temperature as competing models and fit each model with linear regression. We compared models using Akaike's information criterion (AIC) to determine which model provided the best support for our white bass age-structure sample (AICc, corrected for small sample sizes; Burnham and Anderson 1998).

Results

We captured a total of 257 white bass comprised of six cohorts (Table 1); however, year-class strength was erratic. The age-4 cohort (i.e., 2001 year class) dominated the population age structure. Three cohorts (i.e., year classes 1997, 1998, and 2004) were not sampled during our study, indicating a likely weak or missing year class. It is probable that the 2004 year class had not recruited to our gear at the time of sampling; thus, data from 2004 were not used in our analyses. The total annual mortality rate was 29.7%. Mean length by age group increased from 243 mm for age-2 fish, to 322 mm for age-6 fish, and then to 346 mm at age 9 (Table 1).

Table 1. Age structure and growth for white bass collected from SR, South Dakota, in August 2009.

Year	Age group	N	Age structure	Residual	Length	SE	State-wide	Temperature	Precipitation	Inflow
2003	2	5	5	−1.402	243	3.4	243	14.4	12.1	3,403,176
2002	3	4	11	−0.356	256	9.2	298	15.5	6.3	−532,864
2001	4	40	217	2.896	278	2.0	339	15.7	12.3	27,284,618
2000	5	3	18	0.808	292	9.1	358	15.2	15.2	5,320,007
1999	6	2	4	−0.174	322	2.0	384	14.7	19.1	27,595,456
1998	7	0	0	−1.431				14.6	16.7	13,695,349
1997	8	0	0	−1.078				15.2	13.4	50,201,477
1996	9	2	2	0.373	346	4.5		14.1	20.4	69,368,552
Notes: N = number of white bass that were aged; Age structure = corrected age structure (age-length key); Residual = year-class strength index; Length = mean total length (mm) by cohort; SE = standard error of the mean length; Statewide = South Dakota average total length (mm) at age for white bass (Willis et al. 1997); Temperature = mean air temperature (°C) during May and June; Precipitation = total precipitation (cm) during May and June; Inflow = volumetric inflow during May and June (ha-m). The US Bureau of Reclamation database did record May and June combined inflow in 2002 as a negative number. Blank cells represent no data.

The most supported model in our AIC analysis was May and June mean temperature (Table 2), indicating that of the four competing models, white bass recruitment was best related to mean air temperature in May and June (W_i  = 1.00). The May and June precipitation, reservoir inflow, and global (full) models were unsupported by our data (W_i  = 0.00; Table 2).

Table 2. Rankings of a priori models to explain variation in year-class strength for white bass in SR, South Dakota.

Model	N	K	AIC	AICc	ΔAICc	Wi
Temperature	8	3	6.92	12.92	0.00	0.999992
Precipitation	8	3	8.06	14.06	25.43	0.000003
Inflow	8	3	8.00	14.00	25.50	0.000003
Global (Temp + Precip + Inflow)	8	5	9.50	39.50	26.58	0.000002
Notes: N = sample size; K = number of parameters (main factors plus error and intercept); AIC = Akaike's information criterion; AICc = AIC corrected for small sample size; ΔAICc = difference in AICc between each model and the most supported model; Wi  = relative weight of evidence for each model; Temperature (Temp) = mean air temperature (°C) during May and June; Precipitation (Precip) = total precipitation (cm) during May and June; Inflow = volumetric inflow during May and June (ha-m).

Discussion

The maximum age of white bass that we collected in SR was 9 years, which is lower than the maximum ages of 12 and 14 years previously reported from eastern South Dakota glacial lakes (Soupir et al. 1997; Willis et al. 2002). In southern US, waters maximum age apparently is near age 7 (Colvin 2002; Lovell and Maceina 2002), although fish older than age-4 are rare (Muoneke 1994; Colvin 2002). Furthermore, white bass growth was below the statewide average at age 2 (Table 2), although direct comparisons are tenuous because we collected our SR sample in August, when the bass had completed a partial year of growth beyond age 2. However, growth slowed considerably for older white bass (i.e., ages 4–9, which were all below the statewide average despite the additional growing season advantage). Mean length of age-9 white bass only exceeded mean length of age-6 white bass by 22 mm. The size structure of white bass in SR was similar to that in Nebraska reservoirs that also contained a gizzard shad prey base and where most bass were <300 mm (Bauer 2002). However, white bass in eastern South Dakota glacial lakes commonly exceed 400 mm in length (Willis et al. 2002).

Our original research objective was to determine whether white bass recruitment in SR was more affected by spring weather patterns, similar to eastern South Dakota glacial lakes (Pope et al. 1997), or more related to reservoir inflow as in mainstem hydropower reservoirs in Virginia (DiCenzo and Duval 2002) and South Dakota (Beck et al. 1997). According to our AIC analysis, the most supported model for white bass recruitment in SR was spring temperature, while there was no support for models involving spring precipitation or reservoir inflow. Thus, white bass in this impoundment apparently were functioning more like those in the eastern South Dakota glacial lakes. Our literature review on the Virginia and South Dakota reservoirs only involved mainstem reservoirs and all were constructed with hydropower as a primary function. In contrast, SR was originally constructed as an irrigation storage reservoir, although no irrigation district developed. Thus, the reservoir has relatively stable water levels. We suggest that limited discharge pattern of SR may have created a stable environment that was more lake-like than riverine, and white bass subsequently responded as they would in lacustrine habitats.