We assessed the relative proportions of wild- and hatchery-origin summer steelhead Oncorhynchus mykiss caught by sport anglers in the Deschutes River, Oregon. Our objective was to evaluate if steelhead of wild or hatchery origin were equally or disproportionately captured by sport anglers relative to the overall run composition.

Methods

We conducted access-based angler creel surveys of the Deschutes River steelhead fishery during 2000–2019 and 2021 to observe reported catch of wild and hatchery steelhead. We collated fishery-independent hatchery-wild composition data from three locations (two dams and one fish trap) bounding the fishery area we creel sampled. We then used a series of mixed-effect logistic regression models fit to compare angler catch of wild- and hatchery-origin steelhead relative to their composition at the fishery-independent sampling locations.

Result

The best-fit model suggests that anglers were more than twice as likely to observe a wild steelhead in their catch as compared with the chance of the fishery-dependent sampling sites to observe a wild steelhead. Across all years, wild steelhead were 27–31% of the stock composition at the fishery-independent sampling sites yet averaged 65% of angler catch.

Conclusion

Disproportionately high encounter rates of wild-origin summer steelhead suggest that any adverse impacts associated with catch-and-release fishing may be magnified by this effect. We explore some potential mechanisms that could account for this difference and hypothesize that a difference in catchability between wild and hatchery steelhead is the most likely mechanism driving the observed patterns. Regardless of the mechanism, annual catch of wild steelhead is proportionally greater than expectations, given wild steelhead run size compared with that of hatchery steelhead.

INTRODUCTION

Steelhead Oncorhynchus mykiss are iconic to sport anglers throughout the Pacific Rim along the west coasts of the United States and Canada. However, in the United States, most wild steelhead populations are listed as threatened or endangered under the Federal Endangered Species Act of the United States of America. Sport angling continues to persist in basins with threatened steelhead populations in part via sport fisheries targeting hatchery-origin steelhead for retention while requiring catch and release of wild steelhead. Catch-and-release angling on wild steelhead is thought to be an effective management strategy to provide recreational opportunity without negatively impacting stock recruitment (Hooton 1987), while the hatchery-origin steelhead provide an opportunity for harvest. Thus, most fishery managers mandate anglers to release wild steelhead in mixed-stock fisheries because they assume catch-and-release angling will reduce impacts to threatened steelhead populations. Investigating the frequency in which wild- and hatchery-origin steelhead are captured by anglers is paramount to determine the effectiveness of fish management and balance the contrasting objectives presented by sport angling and wild steelhead conservation.

Sport fisheries in river basins with sympatric wild- and hatchery-origin steelhead have typically been assessed by examining the total recreational harvest of hatchery steelhead as a proxy for catch of wild steelhead (see Lubenau 2022). This approach can be deceiving when the total abundance of hatchery-origin steelhead exceeds the wild stock or when the proportion of wild fish in angler catch is decoupled from their actual relative run size. Consequently, managing these fisheries by examining the recreational harvest of hatchery-origin steelhead alone may be insufficient. There are several potential mechanisms that may cause the proportions of wild fish encountered by sport anglers to differ from their relative abundance to hatchery fish on an annual scale. For example, angling dynamics and the distribution of effort in both time and space may decouple the relative abundance of wild and hatchery fish in angler harvest (Nelson et al. 2005; Feeken et al. 2019). If there is spatial segregation of wild and hatchery fish, such as large concentrations of hatchery fish returning to known release sites, anglers targeting known areas of returns could cause hatchery fish to appear disproportionately high in annual catch (Feeken et al. 2019). Moreover, providing a targeted fishery for a sympatric hatchery steelhead population may also increase fishing effort, which also increases angler interactions with wild steelhead (Hooton 2001). Both spatial and temporal mechanisms have the potential to work synergistically if anglers are both targeting hatchery fish at known release locations and those fish enter systems at different times than their wild counterparts. Similarly, both indicate alternative management tools to focus fisheries on hatchery fish, such as spatial closures or changes in season openings and closing if wild fish are found to be disproportionately prevalent in angler catch.

However, angling dynamics are not the only mechanism that can drive the composition of recreational catch to differ from the composition of wild and hatchery stocks. Catchability is defined as the number of fish caught, per the number of fish available, per angler effort unit and per time unit, which can be interpreted as the probability an individual fish is caught (Hilborn and Walters 1992). The number of times a particular fish is caught may also increase as angler effort increases, especially if there is heterogeneity in catchability.

Although direct comparisons between the catchability of wild- and hatchery-origin steelhead by sport anglers are limited because catchability can often be confounded by fish availability, there is some evidence to suggest that they may differ. These differences have been documented to suggest both that hatchery fish have higher catchability and that wild fish have higher catchability dependent on the system. For example, Nelson et al. (2005) determined that the catchability of winter-run hatchery steelhead where anglers were legally restricted to angling downstream of a hatchery release site was twice that of wild steelhead. In contrast, the addition of hatchery steelhead within Oregon streams has been shown to inadvertently increase the catch of wild stocks, although it is inconclusive whether this is due to differences in their catchability or increased fishing effort (Moring 1993). Among wild stocks, sympatric salmonid species often have variable catchabilities, again illustrating that catchability is a variable trait (Tsuboi and Endou 2008). As a variable trait, catchability itself could drive differences in how angler catch composition reflects the true composition of the underlying system.

The heritable nature of catchability creates the potential for hatcheries to either increase or decrease catchability of hatchery fish across generations. For example, Dwyer 1990 concluded that evolutionary history, as well as hatchery selection, can produce an aggressive fish that is highly susceptible to angling. In this case, a hatchery fish is produced with higher catchability. Observed differences in catchability can occasionally be directly linked to specific hatchery practices, such as selecting stocks with specific growth rates that cause them to be more aggressive (e.g., Brauhn and Kincaid 1982). In addition to selection by hatchery practices, behavioral avoidance by wild fish to evade capture by angling gear can also be a heritable trait, which could cause wild fish catchability to decrease relative to hatchery fish (Philippi et al. 2009; Arlinghaus et al. 2017). In contrast, hatchery practices may also decrease catchability over generations of domestication. For example, Tsuboi et al. (2019) found that offspring of the seventh to ninth generations of hatchery fish exhibited poor vulnerability to angling.

Regardless of the underlying mechanism, if wild steelhead are caught at higher rates, populations could be more susceptible to negative impacts from catch and release. For example, the National Marine Fisheries Service assumes the incidental mortality rate from catch-and-release steelhead fisheries to be 5% when determining impact rates (see National Marine Fisheries Service 1999). However, if steelhead are repeatedly caught, associated mortality rates can be compounded. While a steelhead may have a 95% chance of surviving the first hooking event, the probability that a steelhead survives three independent hooking events decreases to 86%.1 A systematic evaluation that varied the encounter and mortality rates in a catch-and-release fishery for wild steelhead from Fish Creek, Idaho, revealed that in extreme cases, if the incidental mortality rate is high and a large portion of the population is caught and released, it can affect the run at a population level (McCormick et al. 2021). Furthermore, if wild steelhead are found to be more susceptible to being captured than hatchery steelhead in certain fisheries, effects from catch and release of wild steelhead may accumulate disproportionately.

The objective of this study is to use Deschutes River fishery creel data, live captures of wild and hatchery steelhead in a Deschutes River adult fish trap, and visual counts of wild and hatchery steelhead from fish passage facilities at The Dalles and John Day dams on the Columbia River to compare the relative catchabilities of wild- and hatchery-origin steelhead. Our hypothesis is that wild steelhead are captured by sport anglers at higher rates, despite the proportion of hatchery steelhead within the fishery area being much higher than wild steelhead. We hereinafter use “stock composition” or “mixed-origin” to describe the hatchery-origin and wild-origin makeup of the total adult return but not to ascribe steelhead to rivers or populations of origin. We use the term “relative catchability” to compare wild- and hatchery-origin steelhead in this fishery, where hatchery-origin steelhead can be legally harvested but wild-origin cannot. Hence, the sport fishery is sampling available steelhead with partial replacement because all wild steelhead must be replaced but not all hatchery steelhead are replaced. If hatchery-origin steelhead have a lower relative catchability in selective fisheries, then wild-origin steelhead will be caught disproportionately in the fishery, potentially undermining the conservation goals intended by mandatory catch and release if they experience adverse effects from capture.

METHODS

Study site

The Deschutes River basin flows north through central Oregon before joining the Columbia River (Figure 1). The Deschutes River drains most of the eastern slopes of Oregon's Cascade Mountains and the western portion of the Ochoco Mountains. Base summer flows during the summer steelhead fishery average 124.6 m³/s (waterdata.usgs.gov). The Deschutes River is one of the largest coldwater refuges for adult steelhead migrating up the Columbia River basin (U.S. Environmental Protection Agency 2021). The confluence of the Deschutes River and Columbia River (at river kilometer [rkm] 328 on the Columbia River, measuring from its mouth) is between The Dalles Dam (rkm 308 on the Columbia River) and John Day Dam (rkm 348 on the Columbia River), both of which have steelhead and salmonid visual counting stations (Figure 1). The Deschutes River is the only Columbia River tributary between these two dams in which steelhead reside. Bounded by dams, adult salmonids are protected from pinniped predators and are larger than typical avian prey. As they move upstream, they encounter the Sherars Falls adult fish trap (Sherars Falls Trap; rkm 71 on the Deschutes River measuring from its confluence with the Columbia River), which is monitored during summer and fall. The fishery study area was the lowest portion of the Deschutes River, from the creel survey station at Heritage Landing (rkm 0 on the Deschutes River; Figure 1) upstream to the creel survey station on Macks Canyon Road (rkm 70 on the Deschutes River; Figure 1). At these locations, the river flows through a deep canyon; thus, most anglers are forced to enter and exit the fishery study area via Heritage Landing and Macks Canyon Road, where angler interview stations occur.

Details are in the caption following the image — **FIGURE 1**
Open in figure viewer PowerPoint

Map of the lower Deschutes River steelhead fishery area, creel sampling stations, Sherars Falls Trap, and Columbia River dams. Arrows indicate the direction of streamflow. Deschutes River hatchery steelhead smolts are released immediately downstream of Pelton Round Butte Dam.

The Deschutes River supports a mixed-stock summer steelhead fishery that generally allows sport anglers to harvest hatchery-origin steelhead, while Endangered Species Act–listed wild-origin steelhead must be released. The Deschutes sport fishery primarily occurs from July through October and targets premature-migrant adult steelhead (Quinn et al. 2016) that are on migratory routes and still far from spawning locations or hatchery release sites. Summer steelhead in the Deschutes River spawn from February to April, primarily upstream of the fishery area in the main stem and tributaries, from Sherars Falls upstream to the Pelton Dam (rkm 160). Hence, unlike other steelhead catchability studies (e.g., Nelson et al. 2005; Lubenau et al. 2024), anglers in the Deschutes River fishery are not targeting locations near hatchery steelhead release sites but are fishing a large-volume river where numerous hatchery and wild adult steelhead stocks from the Columbia River basin are mixing with Deschutes River hatchery and wild steelhead. The fishery is managed by the Oregon Department of Fish and Wildlife, which has historically evaluated the success of the fishery by examining angler harvest of hatchery fish, catch rates measured in catch per unit effort, spawner escapement, and rates of return to the hatchery (Oregon Department of Fish and Wildlife, unpublished reports).

While the Deschutes basin does support a wild steelhead population and annual releases of 162,000 hatchery steelhead smolts from Round Butte Hatchery (rkm 160), which contribute to the monitored fishery area from rkm 0 to rkm 70, a large proportion of adult steelhead entering the fishery are “temporary” or “permanent” stray steelhead originating from other portions of the Columbia River basin (sensu Keefer and Caudill 2014). The adult steelhead captured in this fishery are a mix of returning Deschutes River–origin adults and “temporary” and “permanent” stray adult steelhead (Hess et al. 2016). Hatchery smolt adipose-clipping rates for Deschutes (Round Butte) hatchery steelhead typically exceed 99%. However, other stocks of hatchery steelhead, which may enter the Deschutes fishery as temporary or permanent strays, have lower marking rates. Among returning hatchery steelhead from 2016 to 2019, 5.7% were unmarked, which translated to 11.8% of the unclipped steelhead being unmarked hatchery steelhead since hatchery steelhead were more numerous than wild steelhead (Washington Department of Fish and Wildlife & Oregon Department of Fish and Wildlife 2023).

Creel surveys and angler catch data

The hatchery–wild composition of sport angler steelhead catch was determined from creel survey data during 2000–2019 and 2021. Each year, we conducted an access-based angler survey (Pollock et al. 1994) as summer steelhead either returned to the Deschutes River or strayed into the Deschutes fishery from the Columbia River. Both boat and bank anglers were required to stop and allow for the creel interview at Heritage Landing (rkm 0) and Macks Canyon Road (rkm 64) at the completion of their angling day (Figure 1). These two survey locations cover nearly all the fishery area (64 km) due to limited access in and out of the Deschutes River Canyon. Surveyors asked steelhead anglers how many wild steelhead were captured and released, hatchery steelhead harvested, hatchery steelhead captured and released, and hours fished during the trip. Harvested steelhead were sampled for fin marks and coded wire tags. During the surveys, two random weekend days and four random weekdays were surveyed in 2-week blocks, from July 1 to October 31. The hatchery–wild composition of catch for this study was determined directly from data observations, not estimated catch.

Creel records included wild summer steelhead captured and released, hatchery steelhead captured and released, and hatchery steelhead harvested. These data were combined to determine the total annual observed catch at each creel station. Total observed catch was used to calculate the proportion of angler-identified wild and hatchery steelhead observed in the catch. The proportion of wild steelhead collected in the total sample and observations at each site was defined as the stock composition of steelhead captured by anglers. Anglers identified hatchery steelhead by the absence of an adipose fin, which was clipped at the hatchery prior to juvenile release. Specific hatchery and wild stocks were not differentiated.

During the 2021 fishing season, an emergency closure ended steelhead fishing and retention of hatchery steelhead on September 1, 2021. However, the Deschutes River remained open to other salmonid angling, and steelhead were incidentally captured during those fisheries. In contrast to the fishery time series from 2000 to 2019, all hatchery steelhead had to be released from September 1 to October 31, 2021. We treated the creel sampling that occurred during these 2 months as an unplanned experiment to investigate the potential impact of mandatory release of hatchery steelhead on our relative catchability comparison. Since partial removal of hatchery steelhead (harvest by anglers) eliminates the opportunity for some hatchery steelhead to be captured multiple times and may hence reduce their perceived “catchability” relative to wild steelhead in our data set, we viewed the mandatory hatchery steelhead release regulations during 2021 as an ad hoc test of whether removing this potential source of bias created a biologically meaningful change in the relative catchability between hatchery and wild steelhead.

Fishery-independent data

Fishery-independent observations of wild- and hatchery-origin adult summer steelhead were taken at the Sherars Falls Trap from early July to October 31 during 2000–2019 and 2021. Fish were captured and sampled at the trap as they ascended a fish ladder, which bypasses Sherars Falls. While natural passage does occur at Sherars Falls, the fish ladder is likely the easiest route of passage for steelhead. Fish were diverted out of the ladder into a holding tank, sampled, tagged, and released upstream of the fish ladder. Steelhead that were captured and sampled spent only a few minutes retained before returning to the river. Trap operation occurred 5 days per week; thus, only a subsample of total steelhead passing upstream was collected. Volitional passage of steelhead above Sherars Falls occurred when the trap was not operating. Steelhead captured were classified as wild- or hatchery-origin by examining fin marks, similar to angler identification in creel surveys. Fish without an adipose fin clip were considered wild, while all others were considered hatchery.

In addition to the Sherars Falls Trap, visual counts of summer steelhead from The Dalles and John Day dams were included as fishery-independent data in 2000–2019 and in 2021 from July 1 to October 31. Both dams have visual counting stations at fish ladders managed by the U.S. Army Corps of Engineers, with data reported by the Fish Passage Center.2 Summer steelhead visually observed at the dams were categorized wild- or hatchery-origin based on presence or absence of an adipose fin. Consistent with observations made by anglers and at the Sherars Falls Trap to identify wild or hatchery steelhead, extraneous marks that may have indicated a hatchery steelhead but did not include an adipose clip were treated as wild in our study.

Analysis

The relative catchability of wild- and hatchery-origin summer steelhead was compared using a candidate set of generalized linear models describing the conditional probability that an observed steelhead was of wild or hatchery origin, given it was observed in the creel versus fishery-independent data sources. Data included steelhead catch and release by origin recorded from creel data and stock compositions (i.e., wild or hatchery) recorded at the Sherars Falls Trap, The Dalles Dam, and John Day Dam concurrent with the summer steelhead fishery from 2000–2019 and 2021 (Table 1).

TABLE 1. Wild and hatchery steelhead data for the Deschutes River fishery sampled from creel surveys and the fishery-independent sites. The estimated number of wild steelhead released (Wild), hatchery steelhead released (Hat Rel), hatchery steelhead harvested (Hat Kept), and the percent of total captured hatchery steelhead released by anglers (% Hat Rel) is summed by year for both Deschutes River fishery creel stations. The number of steelhead sampled at the Sherars Falls Trap or visually observed at The Dalles and John Day dams are noted as wild (Wild) or hatchery (Hat) origin and proportion wild (%Wild) at each site.

Year	Deschutes fishery (n)			Fishery estimated catch				Sherars Falls Trap (n)			The Dalles Dam			John Day Dam
Year	Wild	Hat	%Wild	Wild	Hat Rel	Hat Kept	%Hat Rel	Wild	Hat	%Wild	Wild	Hat	%Wild	Wild	Hat	%Wild
2000	1980	1188	63	4515	586	1783	25	944	1639	37	53,716	151,525	26	56,798	163,530	26
2001	2884	2350	55	6525	1743	3527	33	955	3915	20	125,117	378,210	25	112,335	371,074	23
2002	2384	2258	51	5346	1633	3434	32	1031	3206	24	116,565	271,355	30	112,755	277,545	29
2003	1435	753	66	3269	589	1138	34	654	1663	28	86,636	195,738	31	88,078	216,964	29
2004	1351	910	60	2997	574	1390	29	289	937	24	73,815	167,264	31	70,056	163,375	30
2005	1282	752	63	3049	643	1338	32	366	1485	20	72,298	171,915	30	66,287	165,816	29
2006	2254	2115	52	5306	1876	3524	35	574	2653	18	59,199	190,024	24	55,305	176,050	24
2007	1709	1457	54	4021	1293	2180	37	498	2125	19	61,499	185,461	25	63,053	174,464	27
2008	1798	1056	63	4108	937	1658	36	504	1600	24	78,849	198,791	28	82,851	194,311	30
2009	4036	2188	65	8140	1457	2953	33	958	2711	26	143,489	371,622	28	163,259	418,386	28
2010	2790	1327	68	6714	983	2203	31	815	1395	37	122,543	210,490	37	103,991	179,821	37
2011	2709	1651	62	6205	1241	2399	34	496	1283	28	103,123	195,884	34	89,873	167,472	35
2012	2243	1087	67	5336	970	1624	37	302	752	29	71,236	130,404	35	61,631	100,452	38
2013	2773	676	80	6326	474	1155	29	493	891	36	80,347	110,403	42	66,072	89,670	42
2014	3962	1093	78	9704	1106	1627	40	579	995	37	101,146	154,545	40	78,691	124,545	39
2015	2371	665	78	5454	392	1097	26	392	1352	22	76,073	140,945	35	64,512	114,994	36
2016	1067	503	68		374	854	30	109	556	16	39,602	105,425	27	37,823	92,803	29
2017	726	302	71	1754	328	417	44	65	381	15	26,017	70,995	27	24,918	58,335	30
2018	858	264	76	2011	222	469	32	140	445	24	23,636	57,081	29	23,603	49,749	32
2019	1100	218	83	2607	153	372	29	177	274	39	27,532	33,459	45	24,500	26,586	48
2021^a	314	59	84	559	81	73	53	48	148	24	20,192	41,569	33	17,422	36,673	32

^a Retention of hatchery steelhead allowed from July 1 to August 31; mandatory release of all steelhead from September 1 to October 31, 2021.

Mixed-effect models were used to compare the proportion of wild fish observed in the fishery to the proportion of wild fish observed across fishery-independent data sources. The proportion of wild steelhead collected in the total sample and observations at each site was defined as the stock composition. Within these models, year was always included as a random effect, while the data source and location were incorporated as either fixed or random effects in alternative models. Models were also fit by pooling and not pooling the fishery-independent data sources. In all models, the input data included the total number of fish observed at each location as well as the total number of wild steelhead (i.e., the total number of successful Bernoulli trials as described by the binomial distribution). Proportions were bound by estimating each model using a logit-link function. In the first model (M1), data source (either fishery dependent or fishery independent) was considered a fixed effect estimated with a logit-transformation, i.e., M1.A.

{W}_F\sim Binomial\left({p}_F,{N}_F\right),

logit\left({\hat{W}}_F\right)={B}_0+{B}_F\times {I}_F+{b}_y,

{b}_y\sim N\left(0,{\upsigma}_y^2\right),

{\widehat{W}}_{F} = \frac{{e}^{\mathit{\text{logit}}\left({\widehat{W}}_{F}\right)}}{\left[1+{e}^{\mathit{\text{logit}}\left({\widehat{W}}_{F}\right)}\right]} ,

where F is an index for data source, which includes values of fishery-independent or fishery-dependent data;

{p}_F

is the true proportion of wild fish caught by data source F;

{N}_F

is the true total number of all fish caught by data source F;

{w}_F

is the observed proportion of fish in data source F that are wild;

{\hat{W}}_F

is the predicted proportion of wild fish based on the model using the observed proportion (i.e., fitted value); I is an indicator variable indexing data source;

{B}_F

is the fixed effect of the data source on the proportion of wild fish observed; and

{b}_y

is a random effect created by year that contains some shared, year-level variance.

In order to accommodate potential differences between the dams and the Sherars Falls Trap, M1 was also fit by adding an additional level to the explanatory value by categorizing the fishery-independent data source into dam or fish trap, allowing the catchability at Sherars Trap to be different than the catchability of the dams, essentially fitting the same model but having an additional level for fishery data source, i.e., M1.B.

{W}_F\sim Binomia\mathrm{l}\left({p}_F,{N}_F\right),

logit\left({\hat{W}}_F\right)={B}_0+{B}_F\times {I}_F+{(b)}_y,

{b}_y\sim N\left(0,{\upsigma}_y^2\right),

{\widehat{W}}_{F} = \frac{{e}^{\mathit{\text{logit}}\left({\widehat{W}}_{F}\right)}}{\left[1+{e}^{\mathit{\text{logit}}\left({\widehat{W}}_{F}\right)}\right]} ,

where F is an index for data source, which includes values of dam, trap, or fishery;

{p}_F

is the true proportion of wild fish caught by data source F;

{N}_F

is the true total number of all fish caught by data source F;

{w}_F

is the observed proportion of fish in data source F that are wild;

{\hat{W}}_F

is the predicted proportion of wild fish based on the model using the observed proportion (i.e., fitted value); I is an indicator variable indexing data source;

{B}_F

is the fixed effect of the data source on the proportion of wild fish observed; and

{b}_y

is a random effect created by year that contains some shared, year-level variance (

{\upsigma}_y^2

In a second model the data source was not considered and each location was considered an independent fixed effect, i.e., M2.

{W}_L\sim Binomial\left({p}_L,{N}_L\right),

logit\left({\hat{W}}_L\right)={B}_0+{B}_L\times {I}_L+{(b)}_y,

{b}_y\sim N\left(0,{\upsigma}_y^2\right),

{\widehat{W}}_{L} = \frac{{e}^{\mathit{\text{logit}}\left({\widehat{W}}_{L}\right)}}{\left[1+{e}^{\mathit{\text{logit}}\left({\widehat{W}}_{L}\right)}\right]} ,

where L is an index for data source, which includes values of Sherars Trap, The Dalles Dam, Bonneville Dam, Macks Canyon Creel Station, and Heritage Landing Creel Station;

{p}_L

is the true proportion of wild fish caught by data source L;

{N}_L

is the true total number of all fish caught by data source L;

{w}_L

is the observed proportion of fish in data source L that are wild;

{\hat{W}}_L

is the predicted proportion of wild fish based on the model using the observed proportion (i.e., fitted value); I is an indicator variable indexing data source;

{B}_L

is the fixed effect of the data source on the proportion of wild fish observed; and

{b}_y

is a random effect created by year that contains some shared, year-level variance (

{\upsigma}_y^2

A final model (M3) took components of each of the other two, treating the data source as a fixed effect and location as a random effect, i.e., M3.A.

{W}_F\sim Binomial\left({p}_F,{N}_F\right),

logit\left({\hat{W}}_F\right)={B}_0+{B}_F\times {I}_F+{b}_y+{b}_l,

{b}_l\sim N\left(0,{\upsigma}_l^2\right),

{b}_y\sim N\left(0,{\upsigma}_y^2\right),

{\widehat{W}}_{F} = \frac{{e}^{\mathit{\text{logit}}\left({\widehat{W}}_{F}\right)}}{\left[1+{e}^{\mathit{\text{logit}}\left({\widehat{W}}_{F}\right)}\right]} ,

where F is an index for data source, which includes values of fishery-dependent or fishery-independent data;

{p}_F

is the true proportion of wild fish caught by data source F;

{N}_F

is the true total number of all fish caught by data source F;

{w}_F

is the observed proportion of fish in data source F that are wild;

{\hat{W}}_F

is the predicted proportion of wild fish based on the model using the observed proportion (i.e., fitted value); I is an indicator variable indexing data source;

{B}_F

is the fixed effect of the data source on the proportion of wild fish observed;

{b}_y

is a random effect created by year that contains some shared, year-level variance (

{\upsigma}_y^2

); and

{b}_l

is a random effect created by year that contains some shared, location-level variance (

{\upsigma}_l^2

Similar to M1, M3 was also fit by recategorizing the fishery-independent data into dam and trap, allowing the catchability at Sherars Trap to be different than the dams, i.e., M3.B.

{W}_F\sim Binomial\left({p}_F,{N}_F\right),

logit\left({\hat{W}}_F\right)={B}_0+{B}_F\times {I}_F+{b}_y+{b}_l,

{b}_L\sim N\left(0,{\upsigma}_l^2\right),

{b}_y\sim N\left(0,{\upsigma}_y^2\right),

{\widehat{W}}_{F} = \frac{{e}^{\mathit{\text{logit}}\left({\widehat{W}}_{F}\right)}}{\left[1+{e}^{\mathit{\text{logit}}\left({\widehat{W}}_{F}\right)}\right]} ,

where F is an index for data source, which includes values of trap, dam, and fishery;

{p}_F

is the true proportion of wild fish caught by data source F;

{N}_F

is the true total number of all fish caught by data source F;

{w}_F

is the observed proportion of fish in data source F that are wild;

{\hat{W}}_F

is the predicted proportion of wild fish based on the model using the observed proportion (i.e., fitted value); I is an indicator variable indexing data source;

{B}_F

is the fixed effect of the data source on the proportion of wild fish observed;

{b}_y

is a random effect created by year that contains some shared, year-level variance (

{\upsigma}_y^2

); and

{b}_l

is a random effect created by year that contains some shared, location-level variance (

{\upsigma}_l^2

Stock composition from fishery-independent data sources was assumed to represent the true proportion of wild fish within the Deschutes River fishery, allowing us to interpret any differences in the proportions of wild fish in the creel data to represent a difference in relative catchability between the two stock types.

Akaike information criterion (AIC) was used to evaluate candidate models (Akaike 1973; Burnham and Anderson 2002). Akaike weights (w_i) were used to assess the relative support for each candidate model. The best-fit model was used to determine the probability a captured fish was wild given each source (trap, dam, or fishery) and back-transformed using the inverse-logit function. Profile confidence intervals were calculated for all coefficients and exponentiated to be interpretable as the log-odds ratio. All analyses were conducted using the lmer4 package in the R statistical computing language (R Core Team 2020; Bates et al. 2015). Confidence intervals around predicted values were calculated using the LmerTools package.

Assumptions

The Sherars Falls Trap provides a representative sample of the in-river stock composition (hatchery and wild).
Summer steelhead species and origin (hatchery and wild) are correctly identified and fully reported at both The Dalles Dam and John Day Dam fish ladder counting stations.
Sport anglers in the Deschutes fishery can successfully identify and differentiate hatchery and wild summer steelhead.
Anglers interviewed in the creel are not biased in recalling their fishing day or trip.
Anglers interviewed in the creel are honest.
Nearly all anglers fishing from the selected creel access point were intercepted and participated in an interview, providing a census during each sampling event.
There is not an abundance of wild-origin summer steelhead spawning downstream of Sherars Falls that would substantively change the wild–hatchery ratio observed at Sherars Falls relative to what was present in the fishery area.

To examine whether differences in catchability could be reasonably explained by the partial removal of hatchery steelhead, the observed proportion of wild steelhead in angler catch relative to their true proportion in the system was simulated in a simplified conceptual framework using an individual-based model with three primary scenarios. Full details are provided in Appendix A, but in brief, the purpose of each scenario was to determine the overall impact that retention of hatchery steelhead may have had in interpreting the observed stock composition in angler catch and whether or not that could be ascribed to true differences in catchability. In the first, or baseline scenario, hatchery and wild steelhead had the same catchability. In the second scenario, hatchery and wild steelhead catchability was equal, but there was a subset of hatchery steelhead flagged as “biters” in the population. In the third scenario, wild steelhead had a higher catchability than hatchery steelhead.

We also examined the within-year variability of fishing effort and the proportion of wild fish passing The Dalles Dam to explore patterns of fishing effort coupled and run timing of wild and hatchery steelhead. A series of panel plots is provided that illustrate the 2-week interval estimates of angler catch, angler effort, and the proportion of wild fish in each 2-week estimate within angler catch for each year, paired with the daily and 2-week cumulative proportions of wild fish observed at the Dalles Dam. Longer cumulative time periods were not computed, and as fish are thought to pass through the system in between 2–15-day time periods (Keefer et al. 2009).

RESULTS

Across all our candidate models, the odds of observing a wild fish were greater in the fishery than at the fishery-independent locations. The transformed odds estimate for the probability of a sport-caught steelhead being wild was 65%, compared with only 30% of the total stock composition at either of the dams or the Sherars Falls Trap being wild (Table 1). Although the proportion of wild fish was greater than their annual average in the beginning of each year, angler catch and effort was maximized at the period of highest hatchery returns and the proportion of wild fish in angler catch within each year was always greater than at the location of the closest dam (Figure 2). The total volume of fish crossing the dam was the greatest during the peak of the hatchery fish run (Figure 3). This result was consistent with observed creel catch per unit effort within the Deschutes fishery on an annual scale (Figure 4).

Interannual patterns across all of the fishery independent sites were similar (Figure 5). The difference in stock composition among the two Columbia River dams and the Sherars Falls Trap was marginal, with the Sherars Falls Trap consistently having a lower proportion of wild-origin fish (Figure 6). Differences in the estimated proportion of wild-origin steelhead between the dams and trap ranged between 0.02 and 0.07. Moreover, among the candidate set of models, the model with the highest support included a fixed effect that differentiated the fishery-independent data sources between dams and traps (Model M1.B; see Tables 1, 2 and Figure 6). Similarly, the model with the second most support (Model M3.B) also treated the dams and the trap separately.

TABLE 2. Estimated coefficients (Coef) for the fixed effects in each candidate model with 95% profile confidence intervals (CI) on the logit scale. The solution for the conditional probability a fish is wild given the parameter (

p \left[W\mid B\right]

) is also provided on the back-transformed scale. In any particular year, the estimated probability will differ depending on the random effects. Model degrees of freedom (df) and goodness-of-fit comparison are provided via the change in Akaike information criterion scores (

∆ AIC

Model	Model structure	Fixed-effect parameters	Coef	CI	p(W\|B)	df	$\Delta \mathrm{AIC}$
M1.B	$logit\left({\hat{W}}_F\right)={B}_0+{B}_F\times {I}_F+{(b)}_y$	${B}_{F,2}$ = dam	0.46	0.40–0.52	0.314	4	0
		${B}_{F,2}$ = fishery	4.10	4.05–4.14	0.652
		${B}_{F,2}$ = trap	0.78	0.78–0.82	0.269
M3.B	$logit\left({\hat{W}}_F\right)={B}_0+{B}_F\times {I}_F+{b}_y+{b}_l$	${B}_{F,2}$ = dam	−0.78	−0.90 to −0.66	0.314	5	2
		${B}_{F,2}$ = fishery	1.41	1.39–1.42	0.652
		${B}_{F,2}$ = trap	−0.22	−0.24 to −0.12	0.269
M2		${B}_L$ = The Dalles Dam	−1.41	−1.42 to −1.39	0.314	7	3
	$logit\left({\hat{W}}_L\right)={B}_0+{B}_L\times {I}_L+{(b)}_y$	${B}_L$ = John Day Dam	−1.41	−1.42 to −1.39	0.314
		${B}_L$ = Sherars Falls Trap	−1.63	−1.66 to −1.6	0.269
		${B}_L$ = Macks	1.45²	−0.02 to +0.05	0.655
		${B}_L$ = Heritage	−4.8e⁻²	−0.03 to +0.02	0.651
M3	$logit\left({\hat{W}}_F\right)={B}_0+{B}_F\times {I}_F+{b}_y+{b}_l$	${B}_F$ = fishery dependent	0.63	0.48–0.79	0.653	4	28
M3		${B}_F$ = fishery Independent	−1.48	−1.64 to −1.33	0.299
M1	$logit\left({\hat{W}}_F\right)={B}_0+{B}_F\times {I}_F+{b}_y$	${B}_F$ = fishery dependent	0.63	0.50–0.75	0.652	3	379
M1		${B}_F$ = fishery independent	−1.41	−1.42 to −1.40	0.314
M0	Null	Intercept only				2	60,614

There was no significant difference [p(t) < 0.05] in catch composition between the two creel access sites, despite some hatchery steelhead being removed from the system by anglers as they move upstream. The average proportion of wild fish reported by anglers at the two locations were nearly identical, with no significant longitudinal change detected. We estimated 67% of observed steelhead being wild origin and 33% hatchery origin at Heritage Landing and 68% being wild origin and 32% hatchery origin at Macks Canyon Road. Throughout the study period, the proportion of captured hatchery steelhead that were voluntarily released by anglers ranged from 25% to 44%, with a long-term mean of 33% (Table 1). There was no evidence for a temporal trend in the proportion of hatchery steelhead voluntarily released through the years in our study.

During the 2021 sportfishing season, retention of hatchery steelhead was allowed in July and August but legally prohibited in September and October to protect wild steelhead, therefore requiring the release of hatchery steelhead. Over the entire 2021 fishery period (July 1–October 31) wild steelhead composed 33% of the total population observed at The Dalles Dam. During the months of July and August, 83% of the reported catch was wild. For the months of September and October, 65% of captures were wild steelhead. Both the within-season trend in proportion wild reported in the creel (higher proportion wild early, then decreasing later in the season) and the absolute values of proportion wild in the creel during 2021 were generally consistent with observations from prior years that allowed hatchery steelhead harvest (Figure 2).

The individual-based model scenarios suggested that harvesting hatchery steelhead alone was unlikely to cause the discrepancy between the proportion of wild steelhead observed in recreational catch versus the stock composition observed at all three fishery-independent sites. When hatchery steelhead were harvested, the proportion of wild fish observed in the catch may be positively biased but not enough to reflect what was observed in the Deschutes River fishery. Hatchery steelhead abundance was approximately three times higher than wild steelhead abundance within the fishery, with a long-term mean of 33% of the angler-captured hatchery steelhead released. As one example of our simulation results, in the baseline scenario, when all captured hatchery steelhead were retained (0% hatchery steelhead released) from a 12,000-steelhead run, the proportion of wild fish observed in the catch increased only from 25% to 30% (see Appendix A). This slight increase was due to the inability to capture any of the hatchery steelhead more than once.

DISCUSSION

Relative catchability of wild summer steelhead was significantly greater than that of hatchery summer steelhead in the Deschutes River recreational sport fishery, and wild fish were encountered disproportionately by anglers. Overall, we found that wild steelhead were 2.1 times more likely to be captured by anglers than expected based on their proportional abundance in the run. These findings contrast with the findings of Lubenau et al. (2024), who found no significant difference in encounter rate between hatchery and wild steelhead across two run years in the Snake River basin upstream from Lower Granite Dam. The different temporal and geographic contexts of the Deschutes and Snake basin fisheries likely drive these observed differences in catchability. The Deschutes River fishery occurs from July to October and is a “migratory route” fishery, where steelhead are far from their spawning locations in both space and time. There are no hatchery steelhead release sites adjacent to the fishery area we sampled or other physical features to separate hatchery and wild steelhead (e.g., major wild steelhead spawning tributaries). Thus, various stocks of both hatchery and wild steelhead were mixed within the fishery. In contrast, the fishery area integrated by the mark–resight sampling of Lubenau et al. (2024) encompasses hatchery release sites and occurs later in the adult steelhead freshwater migration phase (October–April). These differences create temporal and geographic sorting between wild and hatchery steelhead before and during the spawning season (e.g., Feeken et al. 2019; Zendt et al. 2023). Steelhead anglers respond to the spatial–temporal sorting of hatchery steelhead prior to the spawning season and largely focus effort on areas with hatchery steelhead (Feeken et al. 2019). Similarly, the fishery evaluated by Nelson et al. (2005) also included a steelhead hatchery location that anglers were legally required to fish downstream of. Hence, the differences in relative catchability between hatchery and wild among these studies were likely driven by different levels of spatiotemporal sorting by steelhead and anglers.

Partial removal of hatchery steelhead, where not all hatchery fish were harvested when captured by anglers, occurred in all but 1 year of this study. While partial removal creates bias in comparison of relative catchability between wild and hatchery steelhead, there are several lines of evidence in our study, both observational and computational, that suggest that harvest of hatchery steelhead was a minor source of bias. We observed that interannual variation in the proportion of hatchery steelhead released by anglers had little effect on the proportion of wild steelhead reported in the fishery. For instance, the proportion of wild steelhead within the fishery was nearly identical between fishery year 2000 and fishery year 2017 (Figure 6), but there was a 1.9-fold difference in the percent of hatchery steelhead released between those 2 years (Table 1). Despite the nearly doubled proportion of hatchery steelhead released, the proportion of wild steelhead caught in the fishery was nearly identical in both years (Figure 6). Similarly, widespread fishery closures to retention of hatchery steelhead in both the Deschutes and Columbia rivers occurred in 2021, but these closures did not dramatically change the ratio of hatchery to wild steelhead captured in the Deschutes fishery. The proportion of wild steelhead in the angler catch remained largely unchanged from the long-term averages, with 65% of reported angler captures being wild steelhead. While angler effort in 2021 was not perfectly analogous to other years of our study, it still provides insight into the potential influence of full replacement of hatchery steelhead on their catchability relative to wild steelhead. Similarly, the results of our individual-based model simulation (see Appendix A) support the observational evidence that partial removal of hatchery steelhead was insufficient to alter our conclusions. In sum, these observations and the simulation model suggest that variation in the proportion of hatchery steelhead removed from the fishery had little influence on the proportion of wild steelhead captured in the fishery.

The level of individual heterogeneity in propensity to bite remains uncertain (i.e. wild steelhead caught multiple times by anglers). This uncertainty can not be resolved from existing creel sampling data and may be best approximated through recapture of marked steelhead. Multiple hooking events of wild steelhead occur in other mixed origin fisheries with mandatory wild release (Hooton 1987; Thorstad et al. 2020; Lubenau 2022). Hooton (1987) reported that greater than 30% of wild steelhead caught and released with various gear types on Vancouver Island, Canada, were repeat captures, suggesting that if these fish were removed from the population, the overall catchability observed might be lower. Similarly, Lubenau et al. (2024) observed that of 166 steelhead released with an external anchor tag across two fishing seasons, 22% were reported to be captured a second time, but sample sizes precluded evaluating what proportion were captured three or more times. These frequencies of repeat captures occurred in a fishery with a 50% mean encounter rate for hatchery steelhead (Lubenau et al. 2024).

Removal of hatchery steelhead by angler harvest does not appear to impact our observation of wild steelhead being captured at higher rates in the Deschutes River. During the early years of our study period, Keefer et al. (2009) estimated that 27% of steelhead crossing The Dalles Dam (dam counts provided in Table 1, this paper) entered the Deschutes River fishery area. These steelhead were a mix of returning Deschutes-origin steelhead, temporary strays that ultimately left the Deschutes River, and steelhead that became permanent strays into the Deschutes River. Using this multiyear mean fraction of the total Columbia River run entering the Deschutes River to approximate the abundance of hatchery steelhead in the Deschutes River coupled with our creel sampling data, it is plausible that 6% of the hatchery steelhead entering the Deschutes fishery are encountered by anglers. This approximate encounter rate (inclusive of retained and released hatchery steelhead) for hatchery steelhead is an order of magnitude lower than in the Snake River terminal area fishery (Lubenau et al. 2024). Since 33% of the landed hatchery steelhead were released, it is likely that only approximately 4% of the hatchery steelhead present in the fishery area were harvested, which is consistent with the negligible differences in proportion wild among The Dalles Dam, John Day Dam, and the Sherars Falls Trap (Figure 6)—not enough hatchery steelhead were removed by the fishery to substantively alter the hatchery–wild ratio. Given this lack of impact to the overall stock composition, we conclude that missed opportunity for multiple captures of harvested hatchery steelhead did not create substantial negative bias in our comparison of relative catchability.

While the top two logistic regression models we evaluated found a small difference in stock composition between the Sherars Falls Trap and the Columbia River dams, we believe that stock composition from The Dalles and John Day dams reasonably represent the stock composition in the majority of the Deschutes River steelhead fishery. Probability of straying, whether temporary or permanent, is influenced both by environmental conditions during adult migration (behavioral thermoregulation; Keefer et al. 2005; High et al. 2006) and by artificial transportation during smolt migration (e.g., Keefer et al. 2008). With respect to our relative catchability analysis, Tattam and Ruzycki (2020) found no difference between the probability of wild and hatchery steelhead from the Snake River straying into the John Day River (the next steelhead-bearing tributary to the Columbia River upstream from the Deschutes River) when barge-transported as smolts. Although the frequency of adult steelhead straying into the Deschutes River is greater than into the John Day River, data suggest that there is also no difference in probability of straying between wild and hatchery steelhead (I. A. Tattam, unpublished data). The temporary and permanent stray portions of the steelhead abundance in the Deschutes River fishery were likely composed primarily of adult steelhead that were barge-transported as smolts. There was generally little difference within each year in the probability of smolt transport between wild and hatchery steelhead smolts in the Snake River (Tattam and Ruzycki 2020). Hence, we expect the annual stock composition of the stray steelhead in the Deschutes River fishery to mirror the stock composition at The Dalles Dam.

Evaluating hatchery practices can be an approach to ensure that the intended trade-offs between conservation risk and fishing opportunity are not compromised by lower catchabilities of hatchery fish present in mixed-stock fisheries. In cases where hatchery fish are less catchable than wild fish, managers could place resources into determining what practices during the broodstock selection or the fish rearing process might be changed to increase their vulnerability to capture by sport anglers. Propagation across generations can lead to domestication (Dwyer 1990; McClure et al. 2008). Using angler-caught broodstock to avoid selection for timid fish has been proposed as a strategy to increase propensity to bite, despite not always having the anticipated effect (e.g., Johnson et al. 2023). Reducing these types of unintended effects on propagated steelhead, such as domestication and timidity selection, could ameliorate lower catchability of hatchery-origin steelhead.

Our results relied on accurate and unbiased data reflective of the true hatchery–wild steelhead compositions within the fishery and angler catch. At the fishery-independent dam sites along the Columbia River, it is standard practice for counters at viewing windows to be blind tested and rated for accuracy monthly, with counters held to a minimum 95% accuracy rating (C. Peery, U.S. Army Corps of Engineers, personal communication). On the Deschutes River, there is no evidence to suggest a large number of wild-origin summer steelhead spawning downstream of Sherars Falls that would substantively change the wild–hatchery ratio observed at the Sherars Falls Trap, relative to what was present in the fishery area. The only area known to consistently produce steelhead downstream of Sherars Falls is Buckhollow Creek, where previous monitoring estimated that spawner escapement averaged 331 wild adults during 2009–2016 (Faber et al. 2020), which was about 5–10% of the basin's total wild steelhead production.

In the Deschutes River creel data, we assumed anglers were honest and could remember their catch accurately. Generally, intentional deception occurs when anglers believe their response could directly influence a management decision to close a fishery or when it becomes a matter of prestige (Pollock et al. 1994). Wild steelhead are currently mandatory to release in this fishery, so it is unlikely that anglers perceive any personal risk to their own future fishing experience by reporting catch. Meanwhile, hatchery steelhead harvested are visually validated by the creel surveyors, which makes it unlikely any fish are omitted. Although prestige bias can occur in some fisheries with low catch rates (e.g., Sullivan 2003), many anglers sampled in our study reported no catch. Honest inaccuracies are also unlikely in this fishery. The inception of mandatory-release policies on the Deschutes River began in 1980, which created a standard practice for anglers to differentiate between wild and hatchery steelhead. Moreover, as part of Oregon licensing requirements, anglers must report their catch on harvest cards by identifying species and stock type. This requirement makes them legally liable for misidentification of species or stock origin (wild or hatchery) and provides an incentive to be accurate. Educational materials to help with identification are available in the Oregon Sport Fishing Regulations, which is readily available to all anglers. In addition, recall bias can lead to systematic error when anglers do not remember things accurately (Sudman and Bradburn 1983) but is typically associated with a time lag between when fishing occurred and when an angler is contacted, such as with off-site surveys (Lewin et al. 2021). In the creel survey used for our analysis, anglers were interviewed on the same day that they concluded their fishing trip, making it unlikely that they would forget their catch. As mentioned, catch of steelhead in the Deschutes fishery consisted of very few fish daily (generally single digit or zero catches per day) for most anglers, making each fish more memorable than in fisheries with higher daily catches (e.g., resident trout or bass fisheries). That said, memory recall may have been more problematic in estimating effort because anglers tend to round the time they started or ended their fishing day (i.e., digit bias; Pollock et al. 1994). However, although we did not use effort in our modeling approach, we did assume that all anglers fishing were intercepted from the creel survey access locations and participated in an interview, providing a census during each sampling period. This was important in our analysis because we relied on proportions versus absolute numbers of fish; thus, missing a few anglers or incorrect recall of the number of hours fished would not impact our results about the proportions of wild and hatchery steelhead captured. In the fishery area, the river flows through a deep canyon, forcing anglers to enter and exit the fishery study area via Heritage Landing and Macks Canyon Road creel stations. By virtue of this geography and the creel being conducted by a government agency, angler approach of and compliance with stopping at the creel station is nearly 100% due to a perceived and actual legal obligation to report catch data.

Our annual creel sampling of this steelhead fishery, while expensive and labor intensive, was critical in demonstrating that the relative catchability of wild summer steelhead in the Deschutes River sport fishery was significantly higher than the relative catchability of hatchery summer steelhead. As such, using hatchery steelhead catch as a direct surrogate to wild steelhead catch is inappropriate in the Deschutes River because potential impacts to the wild population are much greater than what is represented by hatchery catch alone. Using only hatchery steelhead harvest as an evaluation metric would not fully account for potential impacts of the Deschutes River sport fishery to wild-origin steelhead. As budgets for creel sampling shrink, using electronic harvest reporting (“eCreel”; Riggers and Jones 2022) may become a necessary substitute for expensive access-based creel sampling. However, an eCreel approach that solely estimates harvested catch is unlikely to appropriately monitor fishery impacts to wild steelhead in mixed-origin fisheries unless there is a statistically significant and temporally stable relationship between hatchery and wild steelhead catchability that can predict wild steelhead catch estimates from reported hatchery steelhead catches.

Regardless of the exact mechanisms driving differences in catchability between hatchery and wild steelhead in the Deschutes River (i.e., whether it is an effect of removing a small fraction of the hatchery return that are “biters,” a broader inherent difference in propensity to bite between wild and hatchery steelhead, or a combination of mechanisms), the functional effect to these wild populations in the lower Deschutes is the same with higher proportional catches of wild steelhead in this migratory route fishery. The difference in relative catchability in this fishery between hatchery and wild steelhead creates higher catch-and-release impacts to the wild population than would be expected if catchability were equal. Hence, fisheries with uncontrolled angler effort, mixed-origin steelhead, and partial removal of hatchery steelhead, such as the Deschutes River, could create unintended angler impacts on wild steelhead, such as potential reduced reproductive success of caught-and-released female salmonids (e.g., Bouchard et al. 2022).

Temporary presence of hatchery and wild out-of-basin steelhead within the Deschutes River sport fishery complicates estimation of management metrics, such as exploitation and impact rates. While estimates of catch, harvest, and spawner abundance are generated annually by monitoring programs, it is not appropriate to pair catch and harvest estimates with Deschutes River spawner abundance because of the high prevalence of temporary strays within the Deschutes River steelhead fishery. Until a spatiotemporal model quantifying the distribution of temporary stray steelhead across the Deschutes riverscape is developed, it will be impossible to distinguish angler catches of temporary strays, permanent strays, and Deschutes-origin wild steelhead to accurately estimate angler impacts to the Deschutes River or other wild steelhead populations. If wild steelhead stocks approach low or critical population abundance levels, this specific fishery will require cautious management to avoid adverse population level consequences. At a broader scale, fishery managers should use caution when allowing high angler effort in migratory route steelhead fisheries with mixed-origin steelhead, which could result in a higher catch of wild steelhead than hatchery steelhead due to differences in catchability.

ACKNOWLEDGMENTS

Special thanks to Rod French for his ideas, support, and management of the programs and fisheries in the lower Deschutes River. We thank Josh McCormick for his ideas and initial analysis of steelhead catchability in the Deschutes River. Nadine Craft and Jon Bowers assisted with graphics. Also, we thank Jerry George, Corey Heath, Mike Gauvin, Tom Stahl, and Lindsay Powell for review, support, and feedback. Thank you to the U.S. Fish and Wildlife Service for funding these monitoring programs on the Deschutes River and the Confederated Tribes of the Warm Springs Reservation as comanagers of Deschutes River fisheries. Finally, thanks to the numerous field crews who collected the creel and trap data from the Deschutes River that we used in this study, especially Jim Burgett, who operated the Sherars Falls Trap for most of this study period.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

We did not capture or handle live steelhead during the course of this work and hence did not need a fish sampling permit.

APPENDIX A: Individual-Based Model of Capture Probability

Specific input parameters within the individual-based model used to define the scenarios included the total number of wild steelhead in the run ( ${N}_W$ ), total number of hatchery steelhead in the run ( ${N}_{H,}$ ), the proportion of hatchery steelhead whose probability of capture was higher than that for baseline hatchery steelhead (i.e., “biters”) ( $H{\prime}_p$ ), the total number of days in the fishery ( ${n}_{seas}$ ), and the parameters defining the probability distributions of being caught for each type of steelhead (wild, hatchery, biter). For each individual wild steelhead, the probability of being caught any day within the fishing season was drawn from a normal distribution with mean ${\upmu_P}_w$ and standard deviation ${\upsigma_P}_w$ . The probability for each hatchery steelhead to be caught was drawn from a normal distribution with mean ${\upmu_P}_H$ and standard deviation ${\upsigma_P}_H.$ In scenarios with a secondary set of hatchery steelhead representing biters in the population, the probability of being caught was drawn from a normal distribution defined by mean ${\upmu_P}_{H\prime }$ and standard deviation ${\upsigma_P}_{H\prime }$ .

The model operated by initializing the true parameters prior to running any simulations. At the start of the simulation, the specified number of individuals for each steelhead type (wild, hatchery, hatchery-biter) was created. Each steelhead was assigned a unique capture probability drawn from the distributions defined by the parameter set. The capture probabilities drawn for each individual at the beginning of the simulation remained constant across the entire fishing season and were used to define success in a series of Bernoulli trials for each day of the fishery. When a hatchery steelhead was successfully caught during a simulated full-retention fishery, the capture probability was overridden to zero for the rest of the time series. At the end of each fishing season, the total number of wild and hatchery steelhead caught was tallied to create a single proportion of wild steelhead observed in the catch across the entire season. Total catch for each steelhead type was also recorded (Figure A.1).

TABLE A.1. Symbol definitions used in an individual-based model to examine the impact of hatchery steelhead removal on the observed proportion of wild steelhead in angler catch.

Symbol	Definition
${n}_{sim}$	Total number of simulations in scenario
${n}_{seas}$	Days in fishing season
i	Index indicating individual
s	Index for simulation number
d	Index for day number in fishing season
${N}_W$	Total number of wild steelhead
${N}_H$	Total number of hatchery steelhead
$H{\prime}_p$	Total percent of “biters” among hatchery steelhead
${\upmu_P}_w$	Average probability a wild steelhead is caught on day d
${\upsigma_P}_w$	Standard deviation of probability a wild steelhead is caught
${\upmu_P}_H$	Average probability a hatchery steelhead is caught on day d
${\upsigma_P}_H$	Standard deviation of probability a hatchery steelhead is caught
${\upmu_P}_{H\prime }$	Average probability a “biter” is caught on day d
${\upsigma_P}_{H\prime }$	Standard deviation of probability a “biter” is caught
$\overset{\rightharpoonaccent }{W}$	Vector of capture probabilities representing each individual wild steelhead
$\overset{\rightharpoonaccent }{H}$	Vector of capture probabilities representing each individual hatchery steelhead
$\overset{\rightharpoonaccent }{H^{\prime }}$	Vector of capture probabilities for each individual hatchery steelhead with higher catchability
${C}_{d_{W,i}}$	Vector indicating whether a wild steelhead was caught by an angler drawn daily using capture probability
${C}_{d_{H,i}}$	Vector indicating whether a hatchery steelhead was caught by an angler on day d using capture probabilities from $\overset{\rightharpoonaccent }{H}$
${P}_{obs}$	Proportion of wild steelhead in total catch across entire simulated fishing season

Model Scenarios

Using the model, three different scenarios were simulated 1000 times under the condition that all hatchery steelhead were released, then repeated under the condition all hatchery steelhead were retained. Across scenarios, wild steelhead were 25% of the true stock composition, with run sizes of 4000 individuals for wild steelhead and 12,000 individuals for hatchery steelhead. The fishing season was set to 120 days across all scenarios, reflecting the 4-month fishery period that is creel sampled in the Deschutes River. The purpose of each scenario was to determine the overall impact that retention of hatchery steelhead may have had in interpreting the observed stock composition in angler catch and whether or not that could be ascribed to true differences in catchability.

In the first, or baseline, scenario, hatchery and wild steelhead had the same catchability. More specifically, the probability of capture was drawn from a normal distribution with mean 0.005 and standard deviation 0.001. By keeping the catchability the same between wild and hatchery steelhead, the impact of hatchery steelhead removal on the observed proportion of wild steelhead in the catch could be isolated.

In the second scenario (Heterogeneous Hatchery), hatchery and wild steelhead catchability was also drawn from the same distribution with mean 0.005 and standard deviation 0.001 except for a subset of hatchery steelhead that represented “biters” in the population. The probability of capture for the biters was drawn from a distribution with mean 0.01, or twice that of baseline hatchery steelhead. While the catchability itself remained the same across each scenario within this set, alternative proportions of biters among the hatchery steelhead were compared, including simulations with 30%, 40%, and 50% of the hatchery steelhead being considered “biters.” This was designed to illustrate the impact of highly catchable steelhead in the population that might be removed first in a full retention fishery.

The third scenario (Wild Steelhead Higher) reflected our interpretation of our creel data collection results. Wild steelhead in this scenario actually did have a higher catchability than hatchery steelhead. In the first case within this scenario, the mean catchability for wild steelhead was set to 0.006, which was three times that of hatchery steelhead, which had a mean capture probability of 0.002. In the second case, the catchability of wild steelhead was increased to five times that of hatchery steelhead, or 0.01 (Table A.2).

TABLE A.2. Parameter sets used to define alternative scenarios in which hatchery and wild fisheries ran simultaneously over a 120-day fishery. Each parameterization was run twice, first with a release-only hatchery fishery and then with a fully retained hatchery steelhead. Wild steelhead were always released.

Parameter	Baseline	Heterogeneous Hatchery			Wild Steelhead Higher
Parameter	Baseline	1.A	1.B	1.C	2.A	2.B
${N}_W$	4000	4000	4000	4000	4000	4000
${N}_H$	12,000	12,000	12,000	12,000	12,000	12,000
$H{\prime}_p$	0	30%	40%	50%	0	0
${\upmu_P}_w$	0.005	0.005	0.005	0.005	0.006	0.01
${\upsigma_P}_w$	0.001	0.001	0.001	0.001	0.001	0.001
${\upmu_P}_H$	0.005	0.005	0.005	0.005	0.002	0.002
${\upsigma_P}_H$	0.001	0.001	0.001	0.001	0.001	0.001
${\upmu_P}_{H\prime }$		0.01	0.01	0.01
${\upsigma_P}_{H\prime }$		0.001	0.001	0.001

Individual-Based Model Results

Our simulations demonstrated that the observed proportion of wild steelhead in recreational fisheries does not always correlate directly to their stock composition but can reflect differences in catchabilities that can be biased when steelhead are differentially removed from the system (i.e., hatchery steelhead are retained while wild steelhead are released). However, despite the potential for some bias in how catch composition reflects catchability, retaining hatchery steelhead alone was unlikely to cause the extreme differences in the proportion of wild steelhead observed in recreational catch versus the true stock composition present in the Deschutes River unless their catchability was also higher.

As evidenced in the baseline scenario, when wild and hatchery steelhead share the same catchability and are fully released, the proportion of steelhead in the catch does reflect the underlying stock composition across the populations (Figure A.2). Conversely, when hatchery steelhead are retained, the proportion of wild steelhead observed in the catch may be positively biased but not enough to reflect what was observed in the Deschutes River fishery, where hatchery steelhead run sizes are around three times the size of wild steelhead runs and upwards of 30% of hatchery steelhead caught are released. In our simulation, the fully retained steelhead from a 12,000-steelhead run caused the proportion of wild steelhead observed in the catch to increase from 25% to 30%, caused by the inability to capture any of the hatchery steelhead more than once (Figure A.2).

Changing the hatchery run to contain a subset of hatchery steelhead with higher catchabilities created a pronounced decrease in the proportion of wild steelhead observed in catch relative to the true stock composition when all hatchery steelhead were released. By including a subset of hatchery steelhead with higher catchability, this decrease accurately reflected the increased encounter rates between hatchery steelhead and anglers. However, by switching to a retention-only hatchery fishery, the highly catchable steelhead in the hatchery population were removed first and could only be caught once. Consequently, similar to the baseline scenario, the proportion of wild steelhead observed in the catch was higher than their true proportion in the system, despite having the same, or lower, catchability. Comparing the proportions of wild steelhead in the catch during a retention fishery between the baseline scenario and scenarios with biters in the population, the bias in the scenario with biters was less than the baseline because the presence of highly catchable biters compensated for some of their removals. For the same reason, though not statistically significant, as the hatchery steelhead with high catchability made up a larger proportion of the hatchery population, the relative upward shift in the observed proportion of wild steelhead in the catch decreased because even with removals there were plenty of highly catchable steelhead left in the system (Figure A.3).

Increasing the catchability of wild steelhead increased their proportion in angler catch relative to the actual stock composition, similar to the increased proportion of wild steelhead observed within our creel data for the Deschutes River recreational steelhead fishery. Although the proportion shifted upward when hatchery steelhead were retained, the shift was marginal at best because hatchery steelhead were less likely to be caught anyway (Figure A.4). Among all the scenarios tested, when wild steelhead truly had higher catchabilities, their proportion was much higher in angler catch and the impact of retaining hatchery steelhead was the least significant (Table A.3).

TABLE A.3. Mean proportion of wild steelhead observed in angler catch (Observed P_W) across each scenario as defined below. Each scenario was run 1000 times.

Scenario summary	Initial defining parameters	Observed P_W
Scenario summary	Initial defining parameters	Release mean (SD)	Retain mean (SD)
Baseline		0.25 (0.004)	0.31 (0.005)
1.Subset of hatchery steelhead with high catchability	A. H′ = 30%; ${\upmu_P}_{H\prime }$ = 0.01	0.20 (0.004)	0.28 (0.004)
1.Subset of hatchery steelhead with high catchability	B. H′ = 40%; ${\upmu_P}_{H\prime }$ = 0.01	0.19 (0.004)	0.27 (0.004)
1.Subset of hatchery steelhead with high catchability	C. H′ = 50%; ${\upmu_P}_{H\prime }$ = 0.01	0.18 (0.003)	0.26 (0.004)
2. Wild catchability > hatchery catchability	A. ${\upmu_P}_w$ = 0.006	0.50 (0.007)	0.53 (0.006)
2. Wild catchability > hatchery catchability	B. ${\upmu_P}_w$ = 0.01	0.63 (0.005)	0.65 (0.005)

Open Research

DATA AVAILABILITY STATEMENT

Creel sampling data are property of the Oregon Department of Fish and Wildlife but can be made available upon reasonable request.

REFERENCES

Arlinghaus, R., Laskowski, K. I., Alos, J., Klefoth, T., Monk, C. T., Nakayama, S., & Shroder, A. (2017). Passive gear-induced timidity syndrome in wild fish populations and its potential ecological and managerial implications. Fish and Fisheries, 18, 360–373. https://doi.org/10.1111/faf.12176
10.1111/faf.12176
Web of Science® Google Scholar
Bates, D., Maechler, M., Bolker, B., & Walker, S. (2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67, 1–48. https://doi.org/10.18637/jss.v067.i01
10.18637/jss.v067.i01
Web of Science® Google Scholar
Bouchard, R., Wellband, K., Lecomte, L., Bernatchez, L., & April, J. (2022). Effect of catch-and-release and temperature at release on reproductive success of Atlantic Salmon (Salmo salar) in the Rimouski River, Québec, Canada. Fisheries Management and Ecology, 29, 888–896. https://doi.org/10.1111/fme.12590
10.1111/fme.12590
Web of Science® Google Scholar
Bradburn, N. M. (1983). Response effects. In P. H. Rossi, J. D. Wright & A. B. Anderson (Eds.), Handbook of survey research (pp. 289–328). Academic Press. https://doi.org/10.1016/B978-0-12-598226-9.50014-8
10.1016/B978-0-12-598226-9.50014-8
Google Scholar
Brauhn, J. L., & Kincaid, H. (1982). Survival, growth and catchability of Rainbow Trout of four strains. North American Journal of Fisheries Management, 2, 1–10. https://doi.org/10.1577/1548-8659(1982)2<1:SGACOR>2.0.CO;2
10.1577/1548-8659(1982)2<1:SGACOR>2.0.CO;2
Google Scholar
Burnham, K. P., & Anderson, D. R. (2002). Model selection and inference: A practical information-theoretic approach ( 2nd ed.). Springer-Verlag.
10.1111/j.1095-8312.2005.00458.x
Web of Science® Google Scholar
Dwyer, W. P. (1990). Catchability of three strains of Cutthroat Trout. North American Journal of Fisheries Management, 10, 458–461. https://doi.org/10.1577/1548-8675(1990)010<0458:COTSOC>2.3.CO;2
10.1577/1548-8675(1990)010<0458:COTSOC>2.3.CO;2
Google Scholar
Faber, D. M., Wilson, W. H., & Ruzycki, J. R. (2020). Investigation of relative reproductive success of stray hatchery and wild steelhead in the Deschutes River subbasin, 1/1/2017–12/31/2018. https://www.cbfish.org/Document.mvc/Viewer/P170496
Google Scholar
Feeken, S. F., Bowersox, B. J., Dobos, M. E., Corsi, M. P., Quist, M. C., & Copeland, T. (2019). Distribution and movement of steelhead and anglers in the Clearwater River, Idaho. North American Journal of Fisheries Management, 39, 1056–1072. https://doi.org/10.1002/nafm.10345
10.1002/nafm.10345
Web of Science® Google Scholar
Hess, M. A., Hess, J. E., Matala, A. P., French, R. A., Steele, C. A., Lovtang, J. C., & Narum, S. R. (2016). Migrating adult steelhead utilize a thermal refuge during summer periods with high water temperatures. ICES Journal of Marine Science, 73, 2616–2624. https://doi.org/10.1093/icesjms/fsw120
10.1093/icesjms/fsw120
Web of Science® Google Scholar
High, B., Peery, C. A., & Bennett, D. H. (2006). Temporary staging of Columbia River summer steelhead in coolwater areas and its effect on migration rates. Transactions of the American Fisheries Society, 135, 519–528. https://doi.org/10.1577/T04-224.1
10.1577/T04-224.1
Web of Science® Google Scholar
Hilborn, R., & Walters, C. J. (1992). Quantitative fisheries stock assessment: Choice, dynamics, and uncertainty. Chapman and Hall. https://doi.org/10.1007/978-1-4615-3598-0
10.1007/978-1-4615-3598-0
Google Scholar
Hooton, R. (2001). Facts and issues associated with restricting terminal gear types in the management of sustainable steelhead sport fisheries in British Columbia. British Columbia Ministry of Environment.
Google Scholar
Hooton, R. S. (1987). Catch and release as a management strategy for steelhead in British Columbia. In R. A. Barnhart & T. D. Roelofs (Eds.), Proceedings of catch and release fishing—A decade of experience (pp. 143–156). Humboldt State University.
Google Scholar
Johnson, M. A., Jones, M. K., Falcy, M. R., Spangler, J., Couture, R. B., & Noakes, D. L. G. (2023). Can angler-assisted broodstock collection programs improve harvest rates of hatchery-produced steelhead? Environmental Biology of Fishes, 106, 1079–1092. https://doi.org/10.1007/s10641-023-01401-5
10.1007/s10641-023-01401-5
Google Scholar
Keefer, M. L., Boggs, C. T., Peery, C. A., & Caudill, C. C. (2008). Overwintering distribution, behavior, and survival of adult summer steelhead: Variability among Columbia River populations. North American Journal of Fisheries Management, 28, 81–96. https://doi.org/10.1577/M07-011.1
10.1577/M07-011.1
Web of Science® Google Scholar
Keefer, M. L., & Caudill, C. C. (2014). Homing and straying by anadromous salmonids: A review of mechanisms and rates. Reviews in Fish Biology and Fisheries, 24, 333–368. https://doi.org/10.1007/s11160-013-9334-6
10.1007/s11160-013-9334-6
Web of Science® Google Scholar
Keefer, M. L., Peery, C. A., Firehammer, J., & Moser, M. L. (2005). Straying rates of known-origin adult Chinook Salmon and steelhead within the Columbia River basin, 2000–2003 (Technical Report 2005-5). University of Idaho.
Google Scholar
Keefer, M. L., Peery, C. A., & High, B. (2009). Behavioral thermoregulation and associated mortality trade-offs in migrating adult steelhead (Oncorhynchus mykiss): Variability among sympatric populations. Canadian Journal of Fisheries and Aquatic Sciences, 66, 1734–1747. https://doi.org/10.1139/F09-131
10.1139/F09-131
Web of Science® Google Scholar
Lewin, W. C., Weltersbach, M. S., Haase, K., Riepe, C., Skov, C., Gundelund, C., & Strehlow, H. V. (2021). Comparing on-site and off-site survey data to investigate survey biases in recreational fisheries data. ICES Journal of Marine Science, 78, 2528–2546. https://doi.org/10.1093/icesjms/fsab131
10.1093/icesjms/fsab131
Web of Science® Google Scholar
Lubenau, W. J. (2022). Encounter rates and catch-and-release mortality of steelhead in the Snake River basin (Master's thesis, University of Idaho). University of Idaho Digital Collections. https://www.lib.uidaho.edu/digital/etd/items/lubenau_idaho_0089n_12278.html
Google Scholar
Lubenau, W. J., Johnson, T. R., Bowersox, B. J., Copeland, T., McCormick, J. L., & Quist, M. C. (2024). Encounter rates and catch-and-release mortality of steelhead in the Snake River basin. North American Journal of Fisheries Management, 44, 3–20. https://doi.org/10.1002/nafm.10965
10.1002/nafm.10965
Google Scholar
McClure, M. M., Utter, F. M., Baldwin, C., Carmichael, R. W., Hassemer, P. F., Howell, P. J., Spruell, P., Cooney, T. D., Schaller, H. A., & Petrosky, C. E. (2008). Evolutionary effects of alternative artificial propagation programs: Implications for viability of endangered anadromous salmonids. Evolutionary Applications, 2008, 356–375. https://doi.org/10.1111/j.1752-4571.2008.00034.x
10.1111/j.1752-4571.2008.00034.x
Web of Science® Google Scholar
McCormick, J. L., Dobos, M. E., Bowersox, B. J., & Copeland, T. (2021). Evaluation of management strategies for an incidental catch and release steelhead fishery. North American Journal of Fisheries Management, 41, 498–512. https://doi.org/10.1002/nafm.10558
10.1002/nafm.10558
Web of Science® Google Scholar
Moring, J. R. (1993). Effect of angling effort on catch rate of wild salmonids in streams stocked with catchable-size trout. North American Journal of Fisheries Management, 13, 234–237. https://doi.org/10.1577/1548-8675(1993)013<0234:EOAEOC>2.3.CO;2
10.1577/1548-8675(1993)013<0234:EOAEOC>2.3.CO;2
Google Scholar
National Marine Fisheries Service. (1999). Endangered and threatened species; Proposed rule governing take of threatened Snake River, Central California Coast, South/Central California Coast, Lower Columbia River, Central Valley California, Middle Columbia River, and Upper Willamette River evolutionarily significant units (ESUs) of West Coast steelhead. Federal Register, 64:250 (30 December 1999), 73485.
Google Scholar
Nelson, T. C., Rosenau, M. L., & Johnston, N. T. (2005). Behaviour and survival of wild and hatchery-origin winter steelhead spawners caught and released in a recreational fishery. North American Journal of Fisheries Management, 25, 931–943. https://doi.org/10.1577/M04-192.1
10.1577/M04-192.1
Web of Science® Google Scholar
Philipp, D. P., Cooke, S. J., Claussen, J. E., Koppelman, J. B., Suski, C. D., & Burkett, D. P. (2009). Selection for vulnerability to angling in Largemouth Bass. Transactions of the American Fisheries Society, 138, 189–199. https://doi.org/10.1577/T06-243.1
10.1577/T06-243.1
Web of Science® Google Scholar
Pollock, K. H., Jones, C. M., & Brown, T. L. (1994). Angler survey methods and their application in fisheries management (Special Publication 25). American Fisheries Society.
Google Scholar
Quinn, T. P., McGinnity, P., & Reed, T. E. (2016). The paradox of “premature migration” by adult anadromous salmonid fishes: Patterns and hypotheses. Canadian Journal of Fisheries and Aquatic Sciences, 73, 1015–1030. https://doi.org/10.1139/cjfas-2015-0345
10.1139/cjfas-2015-0345
Web of Science® Google Scholar
R Core Team. (2020). R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org/
Google Scholar
Riggers, B. L., & Jones, M. K. (2022). Paradigm shift: Applying capture–recapture techniques to electronic licensing system data to estimate Chinook Salmon harvest. North American Journal of Fisheries Management, 42, 815–827. https://doi.org/10.1002/nafm.10778
10.1002/nafm.10778
Web of Science® Google Scholar
Sullivan, M. G. (2003). Exaggeration of Walleye catches by Alberta anglers. North American Journal of Fisheries Management, 23, 573–580. https://doi.org/10.1577/1548-8675(2003)023<0573:EOWCBA>2.0.CO;2
10.1577/1548-8675(2003)023<0573:EOWCBA>2.0.CO;2
Web of Science® Google Scholar
Tattam, I. A., & Ruzycki, J. R. (2020). Smolt transportation influences straying of wild and hatchery Snake River steelhead into the John Day River. Transactions of the American Fisheries Society, 149, 284–297. https://doi.org/10.1002/tafs.10228
10.1002/tafs.10228
CAS Web of Science® Google Scholar
Thorstad, E. B., Diserud, O. H., Solem, Ø., Havn, T. B., Bjørum, L. O., Kristensen, T., Urkd, H. A., Johansen, M. R., Lennox, R. J., Fiske, P., & Uglem, I. (2020). The risk of individual fish being captured multiple times in a catch and release fishery. Fisheries Management and Ecology, 27, 248–257. https://doi.org/10.1111/fme.12407
10.1111/fme.12407
Web of Science® Google Scholar
Tsuboi, J., & Endou, S. (2008). Relationships between catch per unit effort, catchability, and abundance based on actual measurements of salmonids in a mountain stream. Transactions of the American Fisheries Society, 137, 496–502. https://doi.org/10.1577/T06-259.1
10.1577/T06-259.1
Web of Science® Google Scholar
Tsuboi, J., Kaji, K., Baba, S., & Arlinghaus, R. (2019). Trade-offs in the adaptation towards hatchery and natural conditions drive survival, migration, and angling vulnerability in a territorial fish in the wild. Canadian Journal of Fisheries and Aquatic Sciences, 76, 1757–1767. https://doi.org/10.1139/cjfas-2018-0256
10.1139/cjfas-2018-0256
Web of Science® Google Scholar
U.S. Environmental Protection Agency. (2021). EPA Region 10 Columbia River cold water refuges plan (EPA-910-R-21-001). U.S. Environmental Protection Agency.
Google Scholar
Washington Department of Fish and Wildlife & Oregon Department of Fish and Wildlife. (2023). Joint staff report: Stock status and fisheries for fall Chinook Salmon, Coho Salmon, Chum Salmon, summer steelhead and White Sturgeon. Washington Department of Fish and Wildlife; Oregon Department of Fish and Wildlife.
Google Scholar
Zendt, J. S., Allen, M. B., Kock, T. J., Perry, R. W., & Pope, A. C. (2023). Spatial and temporal overlap between hatchery- and natural-origin steelhead and Chinook Salmon during spawning in the Klickitat River, Washington, USA. North American Journal of Fisheries Management, 43, 1687–1701. https://doi.org/10.1002/nafm.10945
10.1002/nafm.10945
Google Scholar

1 Multiplication theorem of probability states p(A and B) = p(A)×p(B).
2 fpc.org (last accessed October 18, 2022).

Citing Literature

Volume44, Issue5

October 2024

Pages 1041-1061

Differential relative catchability of wild- and hatchery-origin steelhead in the Deschutes River, Oregon

Abstract

Objective

Methods

Result

Conclusion

INTRODUCTION

METHODS

Study site

Creel surveys and angler catch data

Fishery-independent data

Analysis

Assumptions

RESULTS

DISCUSSION

ACKNOWLEDGMENTS

CONFLICT OF INTEREST STATEMENT

ETHICS STATEMENT

APPENDIX A: Individual-Based Model of Capture Probability

Model Scenarios

Individual-Based Model Results

Open Research

DATA AVAILABILITY STATEMENT

REFERENCES

Citing Literature

Figures

References

Information

About Wiley Online Library

Help & Support

Opportunities

Connect with Wiley

Differential relative catchability of wild- and hatchery-origin steelhead in the Deschutes River, Oregon

Abstract

Objective

Methods

Result

Conclusion

INTRODUCTION

METHODS

Study site

Creel surveys and angler catch data

Fishery-independent data

Analysis

Assumptions

RESULTS

DISCUSSION

ACKNOWLEDGMENTS

CONFLICT OF INTEREST STATEMENT

ETHICS STATEMENT

APPENDIX A: Individual-Based Model of Capture Probability

Model Scenarios

Individual-Based Model Results

Open Research

DATA AVAILABILITY STATEMENT

REFERENCES

Citing Literature

Figures

References

Related

Information