Seasonal sampling dynamics of smallmouth bass (Micropterus dolomieu) in northeastern South Dakota

Abstract

Numerous studies on smallmouth bass (Micropterus dolomieu) have focused on movements, diets, and capture rates among habitat types, but we are aware of no study that has documented capture rates and size structure by month to aid in assessment and management of this fish. The objective of this study was to determine monthly sampling dynamics and population size structure of smallmouth bass in four northeastern South Dakota glacial lakes. Smallmouth bass were collected monthly (May–September) from Enemy Swim Lake and Roy Lake in 2007 and Clear Lake and Pickerel Lake in 2008 using modified-fyke nets and night electrofishing in rocky habitats. Monthly mean catch-per-unit-effort (CPUE) values for smallmouth bass collected with fyke nets were variable but generally low (0.2–4.7 fish/net night) across months and lakes. Monthly mean CPUE values for smallmouth bass collected by electrofishing were variable in all four lakes (99.7–771.8 fish/h); however, in Clear Lake and Enemy Swim Lake CPUE significantly increased from May through September, primarily due to age-0 bass becoming vulnerable to electrofishing. For all lakes except Clear Lake, there were no significant differences in monthly mean CPUE values of 180 mm and longer smallmouth bass. Length distributions varied by month, but in general May and June samples contained a higher proportion of larger individuals than later sampling months. Overall mean relative weight (Wr) values were greater than 80, but relative weight varied among lakes and months. Our results suggest that spring (May to June) sampling of smallmouth bass using night electrofishing in rocky habitats will result in the broadest length range of captured bass.

Keywords:

• smallmouth bass
• sampling
• electrofishing
• population structure

Introduction

Smallmouth bass (Micropterus dolomieu) is a top predator and among the most popular sportfishes in the United States (Scott and Angermeier 1998; Olson and Young 2003). In South Dakota, smallmouth bass was originally confined to the Minnesota River drainage (Bailey and Allum 1962) but was introduced throughout the state by South Dakota Department of Game, Fish and Parks in the 1980s, which resulted in several self-sustaining populations (Milewski 1990). During the summer of 2006, 14% of anglers interviewed as part of an angler use and harvest survey were targeting smallmouth bass at Enemy Swim Lake in northeastern South Dakota, whereas in 2005 only 2% were targeting smallmouth (Blackwell et al. 2007). Although smallmouth bass angling appears to be increasing in popularity, sampling strategies have been ineffective at collecting fish of all sizes. Electrofishing samples are often dominated by small fish, while anglers catch larger fish not sampled by electrofishing. Therefore, improved sampling strategies are needed to effectively assess and manage these populations.

Capture rate and size selectivity for smallmouth bass can be highly variable among sampling gears, seasons, and habitats. In South Dakota, modified-fyke nets have been used on occasion to sample smallmouth bass but are generally considered ineffective due to low sample sizes (<1 fish/net night). However, Milewski and Willis (1991) reported that the mean total length (TL) of smallmouth bass collected in modified-fyke nets was significantly higher (mean = 287 mm TL) than night boat electrofishing (mean = 198 mm TL). Night electrofishing [mean catch-per-unit-effort (CPUE) = 94 fish/h] had a significantly higher catch rate than day electrofishing (mean CPUE = 46 fish/h) in the Maquoketa River, Iowa (Paragamian 1989). Electrofishing captures a broader size range of individuals than other sampling methods, typically capturing smaller fish that would escape from nets (Beamesderfer and Rieman 1988; Milewski and Willis 1991; Odenkirk and Smith 2005). Effectiveness of electrofishing can be influenced by water clarity and conductivity (Reynolds 1996). Beamesderfer and Rieman (1988) compared five sampling gears for collecting smallmouth bass in John Day Reservoir in Oregon and Washington and found that night boat electrofishing captured the most individuals (N = 6408 bass) and the broadest length range (50–450 mm TL).

Although night boat electrofishing may be the most utilized sampling gear for bass, sampling season may affect capture vulnerability. Sammons and Bettoli (1999) collected black bass from Normandy Reservoir, Tennessee and reported significantly higher catch rates of smallmouth bass in the fall (mean CPUE = 6.9 fish/100 m) than in the spring (mean CPUE = 4.3 fish/100 m), although fish collected during fall were smaller and younger than those in spring collections. Additionally, habitat at sampling sites may influence capture as few individuals are generally collected over silt/organic substrates (Milewski 1990; Sammons and Bettoli 1999), while coarse woody structure (Newbrey et al. 2005) or rocky habitats (i.e., boulders, rip-rap, etc.; Pflug and Pauley 1989; Milewski 1990; Scott and Angermeier 1998; Sammons and Bettoli 1999) are primary capture locations for smallmouth bass.

Prior to 2009, smallmouth bass in South Dakota were collected using fall nighttime electrofishing. The fall samples tended to have high catch rates, but were dominated by small individuals, potentially causing population size and age structure to be underestimated. Therefore, the objective of this study was to determine the monthly sampling dynamics and size structure in an attempt to improve inferences made from collected data and ultimately improve smallmouth bass management in South Dakota waters.

Materials and methods

This study was conducted on four glacial lakes in northeastern South Dakota. Clear Lake (Marshall County) is 474 ha with a mean depth of 3.8 m and a maximum depth of 6.7 m (Stukel 2003). Enemy Swim Lake (Day County) is 870 ha with a mean depth of 5.0 m and a maximum depth of 8.5 m (Weimer 2004). Pickerel Lake (Day County) is 397 ha with a mean depth of 4.8 m with a maximum depth of 12.5 m (Stukel 2003). Roy Lake (Marshall County) is 831 ha with a mean depth of 3.0 m and a maximum depth of 6.4 m (Stukel 2003). The sportfish communities are dominated by black crappie (Pomoxis nigromaculatus), bluegill (Lepomis macrochirus), northern pike (Esox lucius), smallmouth bass, walleye (Sander vitreus), and yellow perch (Perca flavescens).

Smallmouth bass were collected from Enemy Swim Lake and Roy Lake in 2007 and Clear Lake and Pickerel Lake in 2008 using modified-fyke nets and night boat electrofishing. Monthly sampling from May through September was conducted in the littoral zone; in particular, rocky habitats were selected as sample sites. Even though rocky habitats were selected as preferred sampling locations, other habitat types such as sand, submersed and emergent aquatic vegetation, and woody debris were often interspersed within each sample site. Each month, 10 fyke nets were set in fixed locations overnight with at least one net site within each electrofishing station. Fyke nets had 0.9 × 1.5 m double frames with 19 mm mesh (bar measure) and 18.3 m leads. Night boat electrofishing was conducted with two dippers on the bow of a Smith-Root boat generating pulsed-DC current at 60 pulses/sec, 500 V, and 7–9 A. Six 10 min electrofishing transects were conducted at fixed stations each month. The netting and electrofishing were 1 week apart so that operation of one gear did not influence the other. All captured bass were measured for (TL; mm) and wet weight (g) and then released. Surface water temperature, conductivity, and water transparency (Secchi disk) were measured monthly from May through September during fish sampling at each lake (Table 1).

Table 1. Monthly conductivity (µS/cm), Secchi disk transparency (cm), and surface water temperature (°C) for Enemy Swim Lake and Roy Lake in 2007 and Clear Lake and Pickerel Lake, South Dakota in 2008.

Lake	Variable	May	June	July	August	September
Clear	Conductivity	*	469	*	495	*
	Transparency	137	229	*	122	130
	Temperature	12.3	19.5	*	21.6	*
Enemy Swim	Conductivity	325	351	366	370	300
	Transparency	427	161	162	159	*
	Temperature	17.7	20.7	22.7	23.6	16.1
Pickerel	Conductivity	*	370	452	467	*
	Transparency	137	274	152	*	107
	Temperature	10.9	16.4	22.7	24.6	17.7
Roy	Conductivity	552	*	616	614	456
	Transparency	320	191	84	107	*
	Temperature	16.9	*	24.1	23.9	14.8
Note: *Data were not collected.

Relative abundance was indexed as mean CPUE for netting and electrofishing. Fyke net CPUE was calculated for all collected smallmouth bass and expressed as number of bass per net night, while electrofishing CPUE was calculated for all bass collected and for stock-length (≥180 mm) bass and was expressed as number of fish per hour of electrofishing. To assess differences in monthly mean CPUE, repeated measures analysis of variance (ANOVA) was used [MIXED procedure; Statistical Analysis Software (SAS) Institute 2004; Neumann and Allen 2007] when sample sizes were sufficient. Repeated measures ANOVA was used to test if there was an effect of capture by sample location and month for each lake. Pair wise comparisons between months were made using a least-squares means multiple range test when the overall ANOVA was significant.

Sample sizes from fyke nets were insufficient for statistical analyses; therefore, only length-frequency distributions were constructed for monthly electrofishing. TL data were grouped each month by length groups where minimum stock, quality, and preferred lengths are 18, 28, and 35 cm, respectively (Gabelhouse 1984). A chi-square test for differences in probabilities (FREQ procedure; SAS Institute 2004) was used to assess differences in proportional size distribution [PSD; percentage of stock length (180 mm) fish that also exceed quality length (280 mm); Guy et al. (2007)] for smallmouth bass among months. If a difference in PSD was detected, individual chi-square tests were used to determine the differences between months (Neumann and Allen 2007).

Body condition was indexed as mean relative weight (Wr; Kolander et al. 1993; Anderson and Neumann 1996) with standard errors (SEs) (Blackwell et al. 2000) calculated by length groups (15–18, 18–28, 28–35, and >35 cm; Gabelhouse 1984) for bass collected by electrofishing. Length categories that had fewer than five fish were omitted from Wr calculations. An ANOVA was used to test for differences in length groups by date for each lake; if differences were detected, a Bonferroni mean comparison was used to separate means (MIXED procedure; SAS Institute 2004). All statistical analyses were conducted in SAS Institute (2004) with a significance level of 0.05.

Results

Monthly mean fyke net CPUE values were low and variable among months and lakes, ranging from 0.2 fish/net night in Pickerel Lake to 4.7 fish/net night in Enemy Swim Lake (Figure 1). For all lakes, sample sizes from fyke nets were low, precluding statistical tests.

Figure 1. Mean CPUE (fish per net night) and ± SE by sampling date for smallmouth bass captured using modified-fyke nets in 2007 for Enemy Swim Lake and Roy Lake and in 2008 for Clear Lake and Pickerel Lake, South Dakota.

Mean electrofishing CPUE for all sizes of smallmouth bass ranged from 99.7 to 771.8 fish/h, while mean CPUE for stock-length bass ranged from 53.4 to 286.5 fish/h (Figure 2). There was no significant difference in sampling location for all lakes and all smallmouth collected, but there was an influence of sampling location for stock-length bass in Pickerel Lake (F = 4.56; p = 0.04) and Roy Lake (F = 10.97; p = 0.02) (Table 2). Monthly mean CPUE for all smallmouth bass was significantly different from May to September in Clear Lake (F = 6.17; p = 0.01), Enemy Swim Lake (F = 7.68; p = 0.01), and Pickerel Lake (F = 4.33; p = 0.02), but not significantly different in Roy Lake (F = 0.42; p = 0.78). There was a slight decline in stock-length CPUE from May to September in Enemy Swim Lake, Pickerel Lake, and Roy Lake. However, there was a significant difference in Clear Lake for mean CPUE of stock-length bass captured in May and June, which then declined to September.

Figure 2. Mean CPUE (fish per h) and ± SE for all smallmouth bass and stock-length (≥180 mm) smallmouth bass by sampling date collected using night electrofishing in 2007 for Enemy Swim Lake and Roy Lake and in 2008 for Clear Lake and Pickerel Lake, South Dakota. Means with similar letters above or below their SE bars are not significantly different (α = 0.05).

Table 2. Summary statistics for mean CPUE of all smallmouth bass and stock length (≥180 mm) smallmouth bass collected with night boat electrofishing using repeated measures ANOVA to test for effects of sampling location and month from Clear Lake, Enemy Swim Lake, Pickerel Lake, and Roy Lake in northeastern South Dakota.

		All smallmouth bass		Smallmouth bass > 180 mm
Lake	Effects	F	p	F	p
Clear	Location	3.38	0.39	1.74	0.30
	Month	6.17	0.01	4.02	0.02
Enemy Swim	Location	1.66	0.34	0.99	0.48
	Month	7.68	0.01	2.48	0.10
Pickerel	Location	2.88	0.12	4.56	0.04
	Month	4.33	0.02	1.67	0.21
Roy	Location	5.02	0.33	10.97	0.02
	Month	0.42	0.78	0.54	0.71

For all lakes, monthly PSD values were significantly different, and individual chi-square tests were used to assess differences. Length distributions varied by month, but in general, May and June samples contained a higher proportion of larger individuals than later sampling months (Figure 3). Substock bass collected in August and September were primarily less than 100 mm TL; these were age-0 bass that became vulnerable to electrofishing.

Figure 3. Smallmouth bass length distribution and sampling date by length group [Gabelhouse 1984; substock < 180 mm; stock-quality = 180–279 mm; quality-preferred = 280–349 mm; preferred and longer ≥ 350 mm]. Smallmouth bass were collected with night electrofishing from Enemy Swim Lake and Roy Lake, South Dakota in 2007, while Clear Lake and Pickerel Lake, South Dakota bass were sampled in 2008. Bars with similar letters represent proportional size distributions that are not significantly different (α = 0.05).

Trends in mean Wr were highly variable between years and among length groups (Table 3, Figure 4). Substock length and smaller bass Wr peaked in mid-summer for Enemy Swim Lake and Roy Lake (both sampled in 2007), whereas for the other lakes (sampled in 2008) mean Wr increased from May through September. Mean Wr of stock-quality length bass followed a similar yearly trend as substock length and smaller bass mean Wr. Due to low sample sizes for Enemy Swim Lake, mean Wr could not be assessed for fish longer than quality length after June. Mean Wr for quality-preferred length bass was similar from May through September and peaked slightly in August for Clear Lake and Pickerel Lake, while there was a decline from May to June that was similar in Roy Lake. Trends in mean Wr for preferred and longer length bass were similar among lakes where mean Wr was the highest in May, sharply declined to June, and then increased through September.

Figure 4. Mean Wr and ± 95% confidence intervals for smallmouth bass by sampling date and length group [Gabelhouse 1984; substock = 150–179 mm; stock-quality = 180–279 mm; quality-preferred = 280–349 mm; preferred and longer ≥ 350 mm] for bass collected using night electrofishing in Clear Lake (2008), Enemy Swim Lake (2007), Pickerel Lake (2008), and Roy Lake (2007), South Dakota.

Table 3. Monthly mean Wr values for smallmouth bass by length group [Gabelhouse 1984; substock < 180 mm; stock-quality = 180–279 mm; quality-preferred = 280–349 mm; preferred and longer ≥ 350 mm] for bass collected using night electrofishing in Clear Lake (2008), Enemy Swim Lake (2007), Pickerel Lake (2008), and Roy Lake (2007), South Dakota.

				Monthly mean Wr values
Lake	Year	Size	p-value	May	June	July	August	September
Clear	2008	Substock	<0.001	89 a	91 ab	97 abcd	103 c	110 cd
		Stock – quality	<0.001	94 a	95 a	104 b	108 b	108 b
		Quality – preferred	0.130	101	101	105	106	105
		Preferred and longer	<0.001	110 a	94 b	98 b	99 b	99 b
Enemy Swim	2007	Substock	<0.001	90 a	96 a	112 b	108 b	103 c
		Stock – quality	<0.001	91 a	97 b	109 c	104 d	104 d
		Quality – preferred	*
		Preferred and longer	*
Pickerel	2008	Substock	<0.001	84 a	97 ab	99 ab	107 b	116
		Stock – quality	<0.001	85 a	88 b	93 c	99 d	101 d
		Quality – preferred	0.774	87	87	86	89	89
		Preferred and longer	0.024	89 a	84 a	85 a	88 a	90 a
Roy	2007	Substock	<0.001	94 a	100 ab	102 ab	105 b	94 a
		Stock – quality	<0.001	96 a	100 ab	111 c	109 c	104 b
		Quality – preferred	0.228	105	98	99	102	100
		Preferred and longer	<0.001	106 a	97 b	99 b	96 b	91 ab
Note: Means with the same letter are not significantly different (α = 0.05) based on ANOVA and a Bonferroni mean comparison was used to separate means. *Fewer than five fish collected.

Discussion

Our study demonstrated that modified-fyke nets are a less effective sampling gear for smallmouth bass in northeastern South Dakota lakes when compared to night boat electrofishing. Similarly, Milewski and Willis (1991) collected substantially fewer smallmouth bass in eastern South Dakota lakes using spring fyke netting than electrofishing. Fyke nets were considered an ineffective sampling gear for collecting smallmouth bass in our study because too few individuals were collected. As a result of the low catches, we recommend that fyke nets not be used for smallmouth bass population assessment.

Bimodal peaks in electrofishing CPUE of stock-length fish have been documented in the spring and fall for various fish species (Pope and Willis 1996). In Clear Lake and Enemy Swim Lake, a peak CPUE for all smallmouth bass in the fall was primarily due to increased numbers of bass smaller than 180 mm, likely age-0 bass, while mean CPUE for stock-length bass was highest in the spring (May and/or June). McKibbin (2002) sampled largemouth bass (Micropterus salmoides) monthly with night electrofishing in eastern South Dakota lakes and found that the mean CPUE of quality-length (≥30 cm) largemouth bass was highest in the late spring–early summer when surface water temperature was between 16°C and 22°C. Similarly, Lott (2000) concluded that electrofishing at Lake Oahe, South Dakota in May and June when water temperature was 15–20°C captured more large smallmouth bass (i.e., those with TL >350 mm) than other sampling months. Both Lott (2000) and McKibbin (2002) attributed the greater proportion of large individuals captured in the spring to onshore–offshore movements likely related to spawning activity.

Trends in smallmouth bass size structure were similar in our four study lakes. A higher proportion of larger smallmouth bass (quality length) was captured in the spring, and smaller bass (substock; primarily age-0 individuals) dominated samples in August and September. Similarly, Sammons and Bettoli (1999) collected more smallmouth bass in the fall over rubble and riprap habitats in Lake Normandy, Tennessee; however, the bass were smaller and younger than fish collected in the spring. Milewski and Willis (1991) collected smallmouth bass in the spring from Clear Lake and Pickerel Lake in 1989 using night electrofishing and reported their samples were biased toward a high proportion of stock-quality length bass. We similarly found a higher proportion of stock-quality length bass collected in the spring but an underrepresentation of large (quality length) bass in the summer and fall. Additionally, in August and September particularly in Clear Lake, gear saturation (electrofishing CPUE > 700 fish/h) likely occurred with high catch rates of age-0 bass affecting the subsequent capture of larger individuals. At this point, dippers simply could not collect all smallmouth bass that surfaced.

Smallmouth bass mean Wr values were greater than 80 in all study lakes but varied among months. Fish condition can be a short-term indicator of fish health and is primarily influenced by food availability and gonadal growth (Pope and Willis 1996). It is likely that seasonal variation in the smaller fish (TL < 280 mm) can be attributed to differences in food availability. McKibbin (2002) found similar results for largemouth bass in Enemy Swim Lake where condition increased from spring through summer before peaking in July 2000 and August 2001 but then decreased throughout the fall; differences in condition were attributed to food availability between years. Larger smallmouth bass in our study followed similar trends for spring spawning adult fish where condition was highest in spring prior to spawning and was followed by an immediate decline after spawning and an increase through summer and fall (Pope and Willis 1996). Smallmouth bass condition will likely change throughout the year, and proper interpretation of condition during spring sampling will require knowledge of the stage of bass spawning.

Our study was conducted over 2 years, but two of the four lakes were sampled in different years, which could result in temporal variability in abiotic factors (i.e., climate, lake level) that could affect the bass population. Despite the yearly differences in each population and the lakes, the capture results were similar; more large (quality length) bass were collected during the spring (May to June). Therefore, we recommend that smallmouth bass standardized sampling occur during the spring (May to June) using night electrofishing when surface water temperature is between 15°C and 20°C. Selection of sampling sites composed primarily of rock (i.e., boulders, rip-rap, and rubble) will increase sample sizes needed for assessment with less required effort. Spring sampling of smallmouth bass using night electrofishing over rocky habitats should result in a higher mean CPUE for stock-length bass, an increased sample size structure, and a variable Wr value depending on the stage of spawning activity of adult bass. If spring electrofishing is used to sample smallmouth bass, managers need to consider the effect that electrofishing may have on spawning adults as it could potentially reduce a year-class due to additional spawning stress (Siepker et al. 2007). Therefore, careful planning should occur for unstable bass populations or those with small population size in smaller waters so that sampling does not affect subsequent year classes. Sampling prior to spawning would likely be best as many larger smallmouth bass move offshore following spawning activities, which results in samples dominated by small bass.

Acknowledgements

The authors thank the many assistants, especially, N. Pool and J. Grote, who helped with field collection and laboratory analysis. This study was supported in part by the Federal Aid to Sportfish Restoration program administered by the South Dakota Department of Game, Fish and Parks for Project number F-15-R, study number 1508.