Spawning-related movement patterns of a unique rainbow trout (Oncorhynchus mykiss) population in a South Dakota headwater stream

Abstract

The objectives of this study were to determine the spawning locations and movement patterns, and provide insight to the success of a naturalized rainbow trout (Oncorhynchus mykiss) population that occurs in a 3 km section of Spearfish Creek, South Dakota. Twelve rainbow trout from Spearfish Creek, South Dakota were implanted with radio transmitters and tracked from 6 December 2006 to 19 June 2007. Initiation of rainbow trout spawning-related movement was correlated with increasing discharge (r = 0.65, p = 0.0009). Mean gross movement was significantly lower during the pre-spawn period (mean = 236 m, SE = 59) compared with the spawning period (mean = 1661 m, SE = 313) (Kruskal–Wallis, F _1,30 = 56.3, p < 0.0001). Mean net movement was 25 m during the pre-spawn period and −9 m (downstream) during spawning. All rainbow trout used in this study moved into Cleopatra Creek (a tributary of Spearfish Creek) during the spawning period even though other nearby tributaries was available. The presence of brown trout (Salmo trutta) in the other tributaries is a possible reason why the rainbow trout did not use those areas for spawning.

Keywords:

• rainbow trout
• spawning
• movement

Introduction

The Black Hills of South Dakota contain approximately 1287 stream kilometers of suitable trout habitat (Erickson and Koth 2000), nearly all of which has naturally reproducing and recruiting populations of either brown trout (Salmo trutta) or brook trout (Salvelinus fontinalis). Although the rainbow trout (Oncorhynchus mykiss) has been stocked extensively in some streams of the Black Hills and age-0 rainbow trout have been observed on occasion in these streams, recruitment beyond age-0 has not been usually observed during South Dakota Game, Fish and Parks (SDGFP) stream electrofishing surveys. In the Black Hills’ Spearfish Creek watershed, rainbow trout have not been stocked since the 1970s (B. Fletcher, SDGFP, personal communication). However, a small naturally reproducing and recruiting population of rainbow trout exists in a small (about 3 km) section of Spearfish Creek, which makes this population unique in the Black Hills.

Reasons for why rainbow trout maintain a self-sustaining population in the small section of Spearfish Creek and not other Black Hills streams or the remainder of Spearfish Creek are unknown. Spawning habitat in the Black Hills does not seem to be a limiting factor as age-0 rainbow, brown, and brook trout are observed in several Black Hills streams. Rainbow trout spawn in fine gravels during the spring, generally in smaller tributaries of the rivers or lakes in which they reside (Scott and Crossman 1973; Becker 1983). During spawning, rainbow trout display a high degree of homing by returning to their natal stream to spawn after reaching an age of 3–5 years (Scott and Crosman 1973; Becker 1983; Kwain 1983). Juvenile rainbow trout then spend up to 2 years in their natal stream before migrating to the main river or lake the adults reside in once a certain size has been reached (Stauffer 1972; Scott and Crossman 1973; Becker 1983; Mellina et al. 2005).

Based on existing knowledge of rainbow trout spawning behavior (Scott and Crosman 1973; Becker 1983; Kwain 1983; Mellina et al. 2005), I hypothesized that rainbow trout in Spearfish Creek must use local tributaries, rather than Spearfish Creek itself, for spawning. I further hypothesized that tributaries provide rearing habitat for juvenile rainbow trout. Spawning locations and juvenile rearing areas of Spearfish Creek rainbow trout were unknown. Therefore, the objectives of this study were to determine the locations of rainbow trout spawning areas, quantify pre-spawning and spawning-related rainbow trout movement patterns, and provide insight for reasons of rainbow trout reproduction and recruitment success in Spearfish Creek.

Materials and methods

Spearfish Creek is an unregulated, headwater stream located in the northern Black Hills of South Dakota and drains approximately 360 km² (United States Geological Survey [USGS] 2008). Discharge in Spearfish Creek is maintained through precipitation and natural springs. Mean width of the creek in the study section was approximately 9 m. Water temperature generally ranged from approximately 0.1–19°C throughout the year. Cleopatra Creek, a tributary of Spearfish Creek, is an unregulated headwater stream that bisects the portion of Spearfish Creek that supports the natural population of rainbow trout. Cleopatra Creek drains approximately 18 km² (Williamson and Hayes 2000). Iron Creek is a headwater tributary of Spearfish Creek located approximately 4 km above Cleopatra Creek. The drainage area of Iron Creek is unknown. A few other smaller tributaries of Spearfish Creek are also present in the watershed.

Discharge data were obtained from the United States Geological Survey (USGS) gauging station 06430900 in Spearfish Creek above Spearfish, South Dakota. Data were downloaded from the USGS website (www.usgs.gov).

Due to the low estimated abundance and difficulty of capturing adult rainbow trout in this section of Spearfish Creek in 2006 (80 adults >20 cm total length (TL) per km; Shearer and James 2007), only 12 rainbow trout within the 3 km section of Spearfish Creek were implanted with radio transmitters and subsequently tracked with radio-telemetry. Four fish from three different sites (approximately 0.8 km apart) within this section were captured by backpack electrofishing (Smith-Root LR-24) and implanted with transmitters. Fish were anesthetized and implanted with transmitters using the shielded-needle technique (Ross and Kleiner 1982). The transmitters (Model 357, Advanced Telemetry Systems (ATS), Isanti, MN, USA) were 13 × 7 × 24 mm in size, weighed 4 g, and each had a unique frequency. An ATS receiver (Model R4000) with a Yagi Antenna (AF Antronics, Inc., Model F150-3FB 05464) was used to detect transmitter signals and a Trimble handheld global positioning system (GPS) was used to collect coordinate locations of fish locations. Surgeries were conducted on 5 December 2006 and fish were tracked until the transmitter batteries expired (approximately 180 d). All rainbow trout that received transmitters were released at sites of capture.

Fish were tracked one to two times per week during daylight hours in the winter and two to three times per week during the spawning period in April and May. Fish locations were determined by walking along the stream bank until the transmitter signal was located at its strongest point. To disturb fish minimally, in-stream fish locations were not recorded; rather, a GPS point was recorded from the stream bank perpendicular to the strongest signal location. Location accuracy was <5 m, determined by locating hidden transmitters. Movement was defined as the distance between one rainbow trout location and the next recorded location of the tagged fish. These movements represent the minimum linear distance a fish moved during the time period from one tracking event to another. The minimum movement was calculated by measuring the linear distance from one GPS location to the next recorded GPS location (according to date) using ArcMap Geographical Information Systems software.

The number of movements, gross movement, net movement, and mean movement for each individual fish were calculated for the pre-spawn and spawning time periods. Spawning period movements were defined as those that began with a larger than normal movement from its previous location to the time it returned to near its pre-spawn location. Total, or gross, movement was defined as the sum of all recorded distances moved by an individual fish during its tracking period regardless of its upstream or downstream direction (Brown et al. 2001). Net movement was the sum of all movements made during the tracking period for an individual fish, where the upstream movements were given a positive value and downstream movements a negative value, similar to the approach of Brown et al. (2001). The positive and negative movement values were summed to calculate the net movement. The net movement approximates a home range value as it indicates the distance and direction from the fish's original location. The mean distance per movement for an individual fish was calculated as the average of all individual movements, regardless of downstream or upstream direction for that individual fish.

A correlation analysis was conducted to determine if discharge in Spearfish Creek was related to the initiation of rainbow trout movement. Additionally, a Kruskal–Wallis test was used to determine if movement patterns were different between pre-spawn and spawning time periods. Statistical analyses were performed with SAS 9.2 (SAS Institute Inc., Cary, NC, USA).

Results

Rainbow trout implanted with radio-transmitters had a mean TL of 283 mm (SE = 15.3; range, 224–362 mm) and mean weight of 249 g (SE = 34.4; range, 116–474 g). Transmitters averaged 1.7% (SE = 0.23; range, 0.7–3.0%) of implanted fish body weight. Surgery times were short (mean = 4 m and 14 s; SE = 12 s; min = 3 m and 32 s; max = 5 m and 48 s).

Of the 12 rainbow trout implanted with transmitters, only five were used for analysis. These five trout were used because they were active during the spawning time period and spawning-related movement was the focus of this study. Of the tagged fish that were not used for analysis, one apparently was consumed by a predator; its transmitter was found on the stream bank and had teeth marks on the antenna. Two other fish were found far downstream of their original tagging location and did not subsequently move. I assumed that these two fish did not recover from surgery. The other four fish made small movements throughout the study period near their tagging locations but were not observed making large-scale movements during the spawning period. It is possible that I missed these four fish's spawning movements due to my sampling schedule; these fish did not make large spawning movements or these fish spawned at their tagging locations although no redds were observed in their immediate areas. For the five rainbow trout used for analysis, a total of 132 individual locations was recorded from 5 December 2006 to 19 June 2007.

Spawning began in late April and continued till the end of May (Figure 1). Rainbow trout spawning activity was isolated in the Cleopatra Creek tributary. Spawning activities (i.e., presence of redds, observation of redd construction by rainbow trout, and observation of rainbow trout exhibiting spawning behavior) were observed in Cleopatra Creek during the time period rainbow trout were present in Cleopatra Creek.

Figure 1. Mean daily discharge (solid line) and time period of observed rainbow trout movement during spawning (diagonal lines) for radio-tagged rainbow trout in Spearfish Creek, South Dakota, from 1 January to 30 June 2007.

Mean daily discharge in Spearfish Creek from 1 January to 30 June 2007 ranged from 1.02 to 6.31 m³/s (mean = 1.93, SE = 0.08; median = 1.44) (Figure 1). Minimum mean daily discharge occurred on 11 February 2007, and the maximum occurred on 2 May 2007. Rainbow trout movement was correlated to increases in mean daily discharge (r = 0.65, p = 0.0009; Figure 2).

Figure 2. Mean daily discharge versus rainbow trout movement in Spearfish Creek, South Dakota. Pearson's correlation coefficient and significance value are given.

All five radio-tagged trout moved into Cleopatra Creek during the spawning period regardless of their original location upstream or downstream from the confluence of the tributary. Three of the five tagged fish moved from their original location in Spearfish Creek (500 km upstream of the Cleopatra Creek confluence) downstream to the confluence and then moved upstream to Cleopatra Creek (400–700 m). Two of these fish returned to within 4 m of their pre-spawn location. The third fish returned to the confluence of Spearfish Creek and Cleopatra Creek when the signal was lost; I assumed that the transmitter battery expired at that point. The fourth tagged fish originated at the Spearfish Creek/Cleopatra Creek confluence and moved approximately 500 m upstream in Cleopatra Creek where it remained for about 30 days. The signal was lost near mid-May when I assumed the transmitter battery failed. The fifth tagged rainbow trout moved from its original location in Spearfish Creek (700 m downstream of the Cleopatra Creek confluence) upstream into Cleopatra Creek roughly 200 m. After presumably spawning, this fish moved back to within 3 m of its pre-spawn location.

Movement summaries were calculated for two time periods, pre-spawn and spawning (Table 1). A mean of 22 locations per rainbow trout was calculated for the pre-spawn period. Mean gross movement per fish was 236 m, mean net movement was 25 m, and distance per movement was 10 m. During the spawning period, rainbow trout gross movement averaged 1661 m, net movement averaged −9 m (downstream), and distance per movement was 523 m. Rainbow trout moved significantly more during the spawning period than during the pre-spawn time period (Kruskal–Wallis, F _1,30 = 56.3, p < 0.0001).

Table 1. Summary of rainbow trout movement for the pre-spawn and spawning periods in Spearfish Creek, South Dakota from 5 December 2006 to 19 June 2007.

Time period	Fish frequency	TL (mm)	W (g)	N movements	Total movement (m)	Net movement (m)	Distance/ movement (m)
Pre-spawn	151.453	232	144	25	377	50	15
	151.493	224	116	22	178	25	8
	151.632	318	302	18	93	−5	5
	151.874	282	251	26	165	50	6
	151.972	331	367	20	371	5	16
	Mean	277 (22)	236 (47)	22 (2)	236 (59)	25 (11)	10 (2)
Spawning	151.453	232	144	3	2429	1	810
	151.493	224	116	5	1678	−532	336
	151.632	318	302	7	520	485	74
	151.874	282	251	2	1870	4	935
	151.972	331	367	4	1807	−3	463
	Mean	277 (22)	236 (47)	4 (1)	1661 (313)	−9 (161)	524 (157)
Notes: Total movement is sum of distance moved regardless of direction. Net movement indicates directional movement (a negative value is downstream movement, while a positive value represents upstream movement). Standard errors are reported in parentheses.

Discussion

Spearfish Creek rainbow trout followed general winter and spring movement patterns described by previous research – little winter movement (Cargill 1980; Simpkins et al. 2000; Mellina et al. 2005) with larger springtime migrations (Mellina et al. 2005). Rainbow trout in Spearfish Creek had net movements of <50 m during non-spawning periods, indicating that these fish spent the late-winter within the same 50 m of stream.

Springtime movement behavior of rainbow trout was similar to that of their counterparts in other North American populations. Rainbow trout in Spearfish Creek made sudden, large-scale movements (244–1208 m) into Cleopatra Creek in the spring; these movements coincided with increases in daily discharge. Rainbow trout were observed in spawning activities (i.e., redd construction, trout engaged in spawning, presence of redds) in Cleopatra Creek during this time and movement decreased when discharge began decreasing in Spearfish Creek. Erman and Hawthorne (1976) suggested that increases in spring discharge may contribute to the initiation of rainbow trout reproductive behavior, and thus movement. Rainbow trout in British Columbia lake-headed streams made long-range movements in the spring because of increases in discharge and generally ceased when the onset of summer low flows began (Mellina et al. 2005).

I hypothesized that rainbow trout would spawn in nearby tributaries of Spearfish Creek. Of the five rainbow trout tracked in Spearfish Creek, all spawned in the Cleopatra Creek tributary regardless of their original positions (upstream or downstream) of the confluence, even though Iron Creek and other tributaries were nearby. Although the sample size of five fish is low, the movement patterns suggest that rainbow trout spawn in Cleopatra Creek more than other nearby tributaries. Rainbow trout have demonstrated preferential spawning site selection in other areas. Preferential site selection by rainbow trout seemed apparent in the Big Horn River, where the only known spawning location in a 12 km study reach was in a side channel in the upper 4 km of the study reach (Simpkins et al. 2000). Erman and Hawthorne (1976) found that 39–47% of adult rainbow trout spawned in Kiln Meadow Tributary, an intermittent tributary of Sagehen Creek, while only 10–15% of adults spawned in one of three other nearby permanent tributaries. The remaining rainbow trout were presumed to spawn in the upper reaches of the main stem Sagehen Creek.

Presence of other species may be a reason for spawning in one tributary over another. An intermittent portion of Kiln Meadow Tributary was thought to be a primary rainbow trout spawning location because brook trout were absent there, while brook trout were present in the surrounding permanent tributaries (Erman and Hawthorne 1976). A similar situation could be occurring in Spearfish Creek. Stream survey estimates for the Iron Creek tributary of Spearfish Creek indicated a brown trout (>200 mm, TL) biomass of 236 kg/ha and a rainbow trout (<200 mm, TL) biomass of 0 kg/ha in 2004 (James and Erickson 2006). In 2008, brown trout (>200 mm, TL) biomass was estimated at 195 kg/ha while rainbow trout (<200 mm, TL) biomass was 2.7 kg/ha (Bucholz and Wilhite 2009). Conversely, Cleopatra Creek brown trout (>200 mm, TL) biomass estimates were 0 kg/ha in 1997 and 2007 and rainbow trout (<200 mm, TL) estimates in Cleopatra Creek 129 and 55 kg/ha in 1997 and 2007 (Meester 2008; Wilhite and Carreiro 2008, respectively). From 2000 to 2008, Spearfish Creek supported high brown trout biomass (range, 59–279 kg/ha).

Evidence of juvenile rainbow trout in Iron Creek (Bucholz and Wilhite 2009) suggests that spawning does occur in tributaries besides Cleopatra Creek. Brown trout predation on rainbow trout does occur (Stuber et al. 1985; Tabor and Wurtsbaugh 1991) and may be a reason rainbow trout spawn less in areas with high numbers of brown trout. I have observed brown trout predation on trout species during diet studies in Spearfish Creek. Predation on rainbow trout would reduce success of recruitment in Iron Creek and other tributaries. Given that rainbow trout return to natal streams to spawn (Scott and Crossman 1973), rainbow trout may spawn in Cleopatra Creek more than other tributaries due to higher survival of juveniles, and thus a higher number of adults will return to spawn.

Although I was only able to use five rainbow trout in this study, which would make it difficult to detect a low occurrence of use in other tributaries, the movement of these five fish provided useful insight into the rainbow trout population dynamics in Spearfish Creek. The Cleopatra Creek tributary seems to be an important component for maintenance of a locally rare naturally reproducing and recruiting rainbow trout population in the Black Hills. Cleopatra Creek is not unlike important tributaries in other areas in that a specific tributary can be vital to the success of a local fish population. To enhance fisheries that are dependent on specific tributaries, fisheries managers might consider closing these areas to recreational angling (i.e., closed seasonal fishing) during critical time periods (spawning). Managers could also initiate predator removal programs if predation is found to be problematic.

Acknowledgements

I thank G. Simpson, B. Miller, J. Shearer, and G. Galinat of the SDGFP for their assistance with field work and trout surgeries. I also thank D. Willis (South Dakota State University Department of Wildlife and Fisheries) and S. Chipps (South Dakota State University Department of Wildlife and Fisheries, U.S. Geological Survey Cooperative Wildlife and Fisheries Research Unit) for reviewing this manuscript. Funding was provided by Federal Aid in Sport Fish Restoration Project F-21-R-38; SDGFP.