Chinook Salmon (Oncorhynchus tshawytscha): COSEWIC assessment and status report 2018 : page 2
Population sizes and trends
Information about population sizes and trends is presented for each DU separately in a series of DU-specific chapters, including extent of occurrence and area of occupancy, habitat trends, sampling effort and methods (for abundance, enhancement, hatchery releases), fluctuations and trends, and threats and limiting factors.
An introduction to the DU-specific chapters describes the type of information provided in each chapter, explains how that information is collected and/or calculated, and identifies general findings across all DUs.
In British Columbia (and elsewhere), enhancement programs were developed to support populations of Chinook Salmon (Brown et al. 2013a) through hatchery releases (Mackinlay et al. 2004). These hatchery releases have been ongoing in BC since 1967, and are largely directed towards increasing or maintaining fishing opportunities (that is, increasing harvest). Few of the programs have been directed at a purely conservation objective (Figure 15).
DFO's Salmonid Enhancement Program (SEP) is responsible for fish production, habitat restoration and community stewardship activities intended to support Chinook Salmon stocks in southern British Columbia. The program was initiated in 1977 and currently consists of 13 DFO-operated facilities, 11 Community Economic Development Program (CEDP) hatcheries, and 25 volunteer-supported Public Involvement Projects (PIP) (Table 6, Figure 16). Most of these sites occur on Vancouver Island and southern mainland DUs, with the remainder located in the Fraser River drainage. Table 7 summarizes the number of nuSEDS sites by level of enhancement for each Designatable Unit. A 'site' is defined in nuSEDS as a freshwater location where an observation has been made of Chinook Salmon spawners. The database contains a separate record for each site and a single stream may contain multiple sites. For those spawning records with a numerical population estimate, site-level data most often represents the population spawning within a particular river system or tributary of a larger river system.
|DU groups||DFO hatcheries||CEDP hatcheries||PIP hatcheries|
|Boundary Bay (DU1)||Not applicable||Not applicable||L Campbell R
|Lower Fraser (DU2, DU3, DU4, DU5, DU6)||Chehalis R||Not applicable||Alouette R|
|(Spring/Summer) (DU3, DU4, DU5, DU6)||Chilliwack R
|Not applicable||Poco Hatchery|
|South Thompson Ocean Summer (DU12)||Shuswap R||Not applicable||Kingfisher Cr|
|Lower Thompson Spring (DU15)||Spius Cr||Not applicable||Not applicable|
|South Coast Georgia Strait (DU18)||Capilano R
|East Vancouver Island – Goldstream (DU21)||Not applicable||Not applicable||Goldstream R|
|East Vancouver Island – Cowichan-Koksilah (DU21)||Not applicable||Cowichan R||Not applicable|
|East Vancouver Island – Georgia Strait (Summer/Fall) (DU20, DU21)||Puntledge R||Nanaimo R||Not applicable|
|East Vancouver Island – Qual/Punt (DU21)||Big Qualicum R
L Qualicum R
|Not applicable||Englishman Enh
|Southern Fjords (DU22)||Not applicable||Not applicable||Gillard Pass|
|Northeast Vancouver Island (DU23)||Quinsam R||Gwa'ni
Woss Comm H
|Southwest Vancouver Island (DU24)||Nitinat R
San Juan R
|West Vancouver Island Ocean Fall Nootka and Kyuquot (DU25)||Conuma R||Not applicable||Nootka Sd Wtrshd Soc
|Northwest Vancouver Island (DU26)||Not applicable||Not applicable||Holberg In
|DU number||DU name||Unknown/no enhancement||Low-moderate enhancement||High enhancement||Total enhanced||Total sites|
|1||Southern Mainland - Boundary Bay, Ocean, Fall population||2||Not applicable||1||1||3|
|2||Lower Fraser, Ocean, Fall population||Not applicable||1||Not applicable||1||1|
|3||Lower Fraser, Stream, Spring population||6||1||Not applicable||1||7|
|4||Lower Fraser, Stream, Summer population||2||Not applicable||Not applicable||Not applicable||2|
|5||Lower Fraser, Stream, Summer population||6||Not applicable||1||1||7|
|6||Lower Fraser, Ocean, Summer population||6||Not applicable||1||1||1|
|7||Middle Fraser, Stream, Spring (FRCany+GStr) population||1||Not applicable||Not applicable||Not applicable||1|
|8||Middle Fraser, Stream, Fall population||1||Not applicable||Not applicable||Not applicable||1|
|9||Middle Fraser, Stream, Spring (MFR+GStr) population||21||2||Not applicable||2||23|
|10||Middle Fraser, Stream, Summer population||20||Not applicable||1||1||21|
|11||Upper Fraser, Stream, Spring population||22||2||Not applicable||2||24|
|12||South Thompson, Ocean, Summer population||6||1||1||2||8|
|13 and 14||BC South Thompson Stream Summer 1.3 and 1.2||7||Not applicable||1||1||8|
|15||Lower Thompson, Stream, Spring population||4||2||2||4||8|
|16||North Thompson, Stream, Spring population||10||Not applicable||Not applicable||Not applicable||10|
|17||North Thompson, Stream, Summer population||7||Not applicable||Not applicable||Not applicable||7|
|18||South Coast - Georgia Strait, Ocean, Fall population||5||2||7||9||14|
|19||East Vancouver Island, Stream, Spring population||1||Not applicable||Not applicable||Not applicable||1|
|20||East Vancouver Island, Ocean, Summer population||1||Not applicable||2||2||3|
|21||East Vancouver Island, Ocean, Fall population||2||1||9||10||12|
|22||South Coast - Southern Fjords, Ocean, Fall population||10||1||1||2||12|
|23||East Vancouver Island, Ocean, Fall (EVI + SFj) population||4||2||4||6||10|
|24||West Vancouver Island, Ocean, Fall (South) population||16||3||12||15||31|
|25||West Vancouver Island, Ocean, Fall (Nootka and Kyuquot) population||29||1||8||9||38|
|26||West Vancouver Island, Ocean, Fall (WVI + WQCI) population||5||Not applicable||3||3||8|
|27||Southern Mainland, Ocean, Summer population||1||Not applicable||Not applicable||Not applicable||1|
|28||Southern Mainland, Stream, Summer population||1||Not applicable||Not applicable||Not applicable||1|
Pie diagram illustrating the proportions of southern BC Chinook Salmon hatchery releases by production objective for brood year 2013.
Map showing the locations of major Salmonid Enhancement Program hatcheries, spawning channels and economic development projects.
Production of Chinook Salmon at SEP facilities directly supports the delivery of several departmental priorities, which include:
- harvest – enhancement for fisheries that are reliant on enhanced production, and would disappear or become severely constrained in the absence of enhancement. This includes harvest opportunities for First Nations, recreational, and commercial fisheries. When the objective is to provide a targeted-fishery opportunity, production targets may be set to consider both natural spawning and harvest requirements
- assessment – fish produced for mark-recapture analysis where stock assessment information contributes to Pacific region assessment priorities, such as the Pacific Salmon Treaty. This information may also contribute to assessment as defined under the regional stock assessment framework, Area stock assessment priorities and regional SEP assessment priorities, that is, those produced for program performance measurement. Fish produced for assessment generally address other objectives as well but, in a few instances, fish are produced solely for marking for assessment purposes
- conservation – enhancement of a stock at high risk of extirpation, or a vulnerable stock that has been identified as a regional priority (for example, populations which have an approved conservation/recovery strategy). This includes re-establishing locally extinct populations according to transplant guidelines (Fedorenko and Shepherd 1986) and rebuilding populations at high risk of extirpation
- rebuilding – enhancement of a stock that is below apparent carrying capacity. This includes rebuilding depleted populations and mitigating for habitat loss
- stewardship and education – small numbers of fish produced to provide a stewardship or educational opportunity. Production for these purposes is assessed based on contribution to stewardship and educational goals and not on production levels or contribution to harvest or escapement
Production planning occurs annually as part of the Integrated Fisheries Management Planning process. Internal priorities from all DFO sectors, including Fisheries Management, Science Stock Assessment and SEP are brought forward and integrated with partner and stakeholder priorities in the development of a comprehensive production plan. The annual SEP production plan identifies production targets by species, stock, release site, and release strategy in order to meet specific production objectives. Production targets are calculated using current bio-standard survival rates by species and release stage, and are set at a level intended to produce a number of returning adult salmon that will support harvest, conservation or assessment goals.
In 2012, the SEP Production Plan for southern BC Chinook Salmon included a total production target of nearly 40 million Chinook Salmon juveniles. Total Chinook Salmon production in 2011 was approximately 34 million juveniles (Brown et al. 2013a).
Since 1995 there has been active enhancement in 19 of the 28 DUs in southern BC, while the remaining eight DUs have had no active enhancement. Over this period, mean annual production has been 49.3 million juvenile Chinook Salmon, although this has decreased during the most recent generation (2007-2011) to 39.7 million per year (Brown et al. 2013a). Although direct estimates of enhanced production as a percentage of total DU production cannot be calculated, the relative scale of enhancement by DU over the past three generations is summarized in Table 8.
|Conservation unit||DU number||Mean juvenile Chinook Salmon releases, in millions (1995-2015 avg)||Mean juvenile Chinook Salmon releases, in millions (2011-2015 avg)|
|Lower Fraser – fall||DU2||1.2||0.3|
|Lower Fraser – spring||DU3||0.1||0.0|
|Lower Fraser – summer||DU5||0.4||0.5|
|Lower Fraser – Maria Slough||DU6||<0.1||0.0|
|Middle Fraser – spring – age 1.3||DU9||<0.1||0.0|
|Middle Fraser River – summer – age 1.3||DU10||0.1||0.0|
|Upper Fraser River – spring||DU11||0.1||0.0|
|South Thompson – Shuswap R||DU12||0.9||0.7|
|South Thompson – summer – age 1.3||DU13||0.1||0.1|
|Lower Thompson – spring – age 1.2||DU15||0.4||0.3|
|South Coast-Georgia Strait – fall||DU18||1.0||0.6|
|East Vancouver Island – Qualicum/Puntledge||DU21||8.4||6.8|
|Cowichan – Koksilah||DU21||1.6||0.8|
|East Vancouver Island – Nanaimo – fall||DU21||0.5||0.4|
|East Vancouver Island – Goldstream||DU21||0.1||0.0|
|South Coast – Southern Fjords||DU22||0.1||0.2|
|Northeast Vancouver Island||DU23||3.4||3.3|
|Southwest Vancouver Island||DU24||13.5||13.0|
|Nootka – Kyoquot||DU25||3.1||3.4|
|Northwest Vancouver Island||DU26||0.5||0.6|
Hatchery-released juvenile Chinook Salmon can be unfed fry, fed fry, smolts, or 'super smolts'. In most cases releases attempt to mimic indicator stock characteristics (for example, size, timing). However, super smolts are fish for which release is delayed by 2-3 months, which is also likely to alter freshwater life histories, increase straying, and shorten the number of years they spend at sea (Clarke et al. 2016; Westley et al. 2013). These fish are grown at a fast rate to achieve an unusually large size at release for the actual age. Most releases occur close to when naturally produced juveniles are migrating to the ocean and undergoing the smolting process (not the case for unfed fry or fry). Therefore, release timing is dependent on life-history type and variations therein: juveniles of ocean-type variants are generally released in the spring as 0+ smolts (that is, underyearlings), while those of stream-type variants are released in the spring as 1+ smolts (that is, yearlings).
SEP has developed several options for releasing Chinook Salmon progeny from hatcheries into the natural environment. Each strategy has advantages and disadvantages and is selected to best meet the enhancement objective as per the production planning framework. The release strategies are shown in Table 9, in order from the youngest life-history stage to the oldest. The information in this table has been adapted from HSRG (2004) and California HSRG (2012).
|Unfed Fry (incubator boxes)||
|Fall, late or delayed release||
Assessment and monitoring
As part of ongoing program development and monitoring, SEP employs many program- and project-level tools to guide operations and planning. At the program level, several integrated planning tools guide management and decision-making. These include SEP Production Planning: A Framework (DFO 2012), the Biological Risk Management Framework for Enhancing Salmon in the Pacific Region (DFO 2013b), and an infrastructure strategy. These tools will be integrated into a long-term planning process that will focus on program-level strategic management as well as directing annual program and project planning. This will ensure that enhanced salmon production objectives of the Department and stakeholders are being met.
At the operational level, Fish Health Management Plans (FHMPs) have been implemented at all DFO-operated hatcheries. These plans were summarized by Stephen et al. (2011), as part of Cohen Commission Technical Report 1A. In addition to the FHMPs, SEP implements Enhancement Guidelines for Salmon Enhancement Programs (DFO 2005c), which provide guidance at the operational level to ensure that any genetic, disease and ecological risks of enhancement are managed appropriately. Hatchery operations are evaluated as a component of periodic program review processes, such as the 2004/5 SEP Facility Operations Review (FORT) (DFO 2005b).
Each spawning pair of Chinook Salmon has a reproductive potential of between 2,000 and 17,000 progeny, depending primarily on the size of the female (McPhail 2007). In the case of declining populations, fewer than two of these progeny survive to adulthood. The actual number of offspring that survive to reproduce the next generation is dependent on the threats (direct human-induced interactions) and limiting factors (such as natural fluctuations in the environment) that are encountered at each stage in the Chinook Salmon life history. These threats can be natural such as predation or food restrictions, direct human interaction such as fishing, or habitat pressures exacerbated by human activities such as pollution, forestry or urbanization (Brown et al. 2013a).
Primary threats to Chinook Salmon include harvest, changes in freshwater or marine habitat, hatcheries, pathogens/aquaculture, and climate change.
The threat that harvest poses to southern BC Chinook Salmon should be considered on a DU by DU basis. Total fishing impacts vary according to geographical regions where fisheries occur and intercept Chinook Salmon during their oceanic migrations. These impacts also vary for each of the southern BC indicators according to fishery type, for example, troll, net, and recreational fisheries (Brown et al. 2013a).
The total landed catch of Chinook Salmon in BC has declined considerably since the mid-1970s (Figure 17). Harvests of Chinook Salmon originating from southern BC streams have declined (Figure 18), with reduced commercial landings accounting for most of this decline (Figure 19). Some of these reductions are at least in part attributable to changes in the marine environment driven by large scale climate oscillations (for example, NPGO, PDO) (Braun et al. 2015; Dorner, 2017).
Table showing trends in total landed catch of Chinook Salmon in three regions (BC, Washington-Oregon, and Alaska) by catch year, regardless of the geographic origin of the fish.
Chart showing trends in estimated coast-wide ocean fishery landings (ocean net and excluding ocean net) of Chinook Salmon originating only from streams in southern British Columbia.
Chart showing trends in total landed catch of Chinook Salmon in southern BC ocean fisheries by type of fishery (commercial, sport, First Nations).
Total exploitation rates, expressed as the proportion of adults harvested in fisheries over a brood year's complete life span, are computed based on coded-wire tag (CWT) recovery data for indicator stocks (Figure 20). Total exploitation rates declined substantially over brood years 1973-1993 for both far-north migrating and locally distributed stock types, from an average of approximately 75 percent to an average of about 45 percent. A rate in the range of 70 percent to 80 percent is likely well above what is necessary for maximum sustainable yield (MSY) during periods of average productivities (PSC-CTC 2017a). Total exploitation rates for all three ocean distribution types have been similar since about the 1993 brood year and have ranged from about 25 percent to 50 percent. Despite these dramatic reductions in total exploitation rates and ocean fishery landings, many stocks have experienced declines in spawning escapements over the past three generations.
Chart showing trends in average estimated total exploitation rates by brood year for adipose fin-clipped and coded-wire tag (CWT) indicator stocks by ocean distribution pattern: far north-migrating, locally distributed, and offshore.
Under low productivity, harvest is a significant threat for many DUs. Riddell et al. (2013) note that although the total exploitation rate has reduced considerably since about the 1993 brood year, a corresponding decline in stock-specific productivities could substantially reduce the MSY. This decreased stock-specific productivity could be due to low marine survival, and suggests that even the lowered exploitation rates could be too high. Realized exploitation rates can be compared to approximate total exploitation rates for MSY (EMSY) to understand whether fishing has been an important stressor on southern BC Chinook Salmon (Table 10). Based on existing estimates of productivity, only Dome Creek (DU11) has exploitation rates that could be regarded as 'stressful'. However, relatively strong evidence from the marine habitat suggests that productivity has decreased coast-wide since the early 1990s (Kilduff et al. 2014; Dorner et al. 2018), and that estimates of exploitation rates for EMSY are therefore overestimated. Recent evidence suggests changed maturation schedules, decreased size at age, and decreased fecundity at age all contributed to decreased productivity (Bailey pers. comm. 2018). When EMSY was adjusted to account for lower productivity, six of 12 indicator stocks showed total exploitation rates that exceeded EMSY values. It is therefore possible that many stocks are being over-exploited as their productivity falls (Bailey pers. comm. 2018).
|Location||CU||Indicator stock||Mean 1995-2008 brood exploit. rate||Assumed ricker 'a'||EMSY||Adjusted ricker 'a'||Adjusted EMSY|
|Far north migrants||SWVI||Robertson||0.578a||2.03||0.73||1.015||0.44|
|Far north migrants||NEVI||Quinsam||0.405||2.03||0.73||1.015||0.44|
|Far north migrants||Qual-Punt Falls||Big Qualicum||0.415||2.03||0.73||1.015||0.44|
|Far north migrants||Mid ECVI Summer||Puntledge||0.302||2.03||0.73||1.015||0.44|
|Far north migrants||Shuswap Summer 0.3||Lower Shuswap||0.486||2.07||0.74||1.035||0.44|
|Far north migrants||Thompson Summer 0.3||Lower Shuswap||0.486||1.59||0.62||0.80||0.35|
|Locally-distributed||Lower Fraser Fall||Harrison (Chehalis)||0.355||1.67||0.64||1.34||0.54|
|Locally-distributed||Lower Fraser Fall||Chilliwack||0.301||1.67||0.64||1.34||0.54|
|Offshore||Lower Thompson Spring (1.2)||Nicola||0.238||1.51||0.60||0.75||0.34|
|Offshore||Upper Fraser Spring||Dome||0.698||1.65||0.63||0.82||0.36|
aThe total exploitation rate for the Robertson Creek stock is probably unusually high due to the intensive terminal fisheries targeting these hatchery-origin fish.
Habitat modification and degradation can cause a decline in southern BC Chinook Salmon abundance: (1) directly through continuous deterioration of freshwater habitats; and/or (2) indirectly through interactions between human-induced declines in habitat quality and other stressors whereby populations from poorer freshwater habitats are more vulnerable to other stressors (Riddell et al. 2013). Evidence suggests that the freshwater life history of Chinook Salmon is plastic, with juveniles undergoing a variety of migration and rearing strategies in their first year (Bradford and Taylor 1997) in response to environmental variation. This diversity in life history reduces the likelihood that a single environmental or biological forcing agent will be able to generate coherent trends in freshwater survival across a broad spatial scale.
Dams can alter flow regimes, impede access to spawning and rearing habitats, affect the quality and quantity of those habitats, and create stress on migrating Chinook Salmon. Identified dams that affect Chinook Salmon include hydroelectric facilities on the Puntledge, Bridge, Seton, Nechako, Middle Shuswap, Alouette, Stave, Cheakamus, Cheekeye, Nanaimo, Quinsam, Campbell, Salmon, Theodosia, Capilano, Seymour, Coquitlam, Cowichan and Big Qualicum rivers.
With the caveat that their study did not consider changes from historical levels (for example, in the Canadian portion of the Columbia River), Riddell et al. (2013) concluded that there were no obvious freshwater environmental drivers that could explain recent trends in southern BC Chinook Salmon spawner abundance. No correlations were found between recent trends in escapement and human-induced changes characterized by a set of pressure indicators (see Stahlberg et al. 2009). Human-induced watershed changes that have been considered include urbanization, forestry, Mountain Pine Beetle (Dendroctonus ponderosae), changes in land cover, mining development (Kjelson et al. 1981, 1982), agricultural/rural development, road development (Bradford and Irvine 2000), stream crossing density, riparian disturbance, permitted waste water discharge, and water allocation (Kjelson et al. 1981, 1982). In recent years, climate-related observations for freshwater habitat include warmer river temperatures and earlier freshets (DFO 2017), which can influence upstream migration conditions, temperatures on the spawning grounds, and smolt outmigration conditions.
Despite the findings of Riddell et al., watershed changes were quantified for this report based on available data from Porter et al. (2013) and were included in the IUCN threats calculator assessment (see Format of Designatable Unit-specific Chapters section below).
While migration patterns and other aspects of marine ecology continue to be investigated, ocean residence is recognized as a very important component of the life cycle of all Pacific salmon (Brown et al. 2013a). Riddell et al. (2013) concluded that marine habitat conditions during the first year of marine residency were very likely a key driver of recent trends in survival and productivity. All DUs share the marine environment, but stocks that start their marine life in the Strait of Georgia, for example, encounter a different marine environment than those that start off the west coast of Vancouver Island. The fish in each DU may also have different marine migration patterns as described in the Dispersal and Migration section of this report. Regardless, because of the broader shared environment, large-scale changes are likely to affect all DUs in a coherent way. This is supported by the fact that an overall declining trend is observed across DUs of southern BC Chinook Salmon (Riddell et al. 2013).
Chinook Salmon spend much of their lifetime in the marine environment, where they are exposed to a wide array of factors that can affect growth and survival (for example, Peterman 1987; Beamish and Mahnken 2001; Pearcy and McKinnell 2007; Farley et al. 2007). The importance of the early marine period to survival of Pacific salmon is supported by studies of spatial scales of covariation in recruits per spawner and studies that demonstrate greater correlation between coastal ocean conditions during early sea life and marine survival than survival at other life-history stages (Magnusson 2001; Wertheimer et al. 2004; Mueter et al. 2005; Pyper et al. 2005). Local marine conditions (that is, the marine environment near the point of ocean entry) are also demonstrated to have a high correlation with Chinook Salmon marine survival (Riddell et al. 2013). Stocks that enter the ocean near each other have more similar trends in survival than those entering the ocean further apart (Figure 21), although a recent study suggests a more complex story, with Salish Sea stocks exhibiting weaker coherence than those located outside the Salish Sea (Ruff et al. 2017).
In both direct and indirect ways, marine water temperature poses a number of challenges to all species of Pacific salmon (Meuter et al. 2005; Richter and Kolmes 2005). In 2012, Irvine and Crawford summarized Pacific Ocean oxygen, salinity and temperature conditions to examine their potential food web effects (phytoplankton, zooplankton, invertebrate, piscine, and avian populations). Their data at the time showed that the Northeast (NE) Pacific Ocean was cooler than average, continuing a trend beginning in 2005. More recent data indicate ocean and land temperatures in the NE Pacific and British Columbia/Yukon were above average from 2013 through to 2016 (DFO 2017), in part due to large scale climate anomalies (the 'warm blob' – see Bond et al. 2015; Di Lorenzo and Mantua 2016).
Large scale climate anomalies can share patterns with salmon marine survival rates. For example, the North Pacific Gyre Oscillation (NPGO) has exhibited a pattern since 1995 similar to the widely shared trend in marine Chinook Salmon survival (Riddell et al. 2013) (Figure 22, Figure 23) (excluding Salish Sea stocks – see Ruff et al. 2017). In the latter half of 2013 a warm sea-surface-temperature (SST) anomaly, referred to as the 'warm blob', developed in the NE Pacific, extending to depths of 100m. By 2014 the blob had moved into BC and Yukon coastal waters and, in 2015, combined with the effects of an El Niño event. Together, these events further increased local land and ocean temperatures. In the Northeast Pacific shifts were observed from cool to warm water copepod species (DFO 2016; DFO 2017). Unusual fish species from southern latitudes were also observed, which may have affected predator-prey dynamics within the salmon ecosystem (DFO 2016; DFO 2017). Beginning in late 2016 and through to 2017, the blob and El Niño were no longer present in the NE Pacific. In 2017, adult returns to many southern BC Chinook Salmon DUs (for example, in the Fraser and West Coast Vancouver Island systems) were extremely poor across most salmon species and populations. More variable responses occurred in the more northern Pacific salmon populations.
With the exception of Fraser River stocks, most southern BC Coded Wire Tag (CWT) indicator stocks are released at or near the ocean, and therefore provide excellent indicators of marine survival conditions from release to ocean age two (Riddell et al. 2013). Fraser River CWT stocks are released roughly more than 100km from the ocean, so data from these stocks represent the combined effect of survival after the fish were released until the first age of maturity. Based on CWT data, reduced survival rates have been observed in recent years for most southern BC Chinook Salmon indicator stocks, although some Chinook Salmon DUs did not show a decline (Figure 24, Figure 25). These latter populations entered the Strait of Georgia either early or late, as opposed to the more common May/June timing, suggesting that temporal differences in early marine conditions can have strong effects on early marine survival (Beamish et al. 2010).
Of the 31 species of marine mammals that occur in waters off the Pacific coast of Canada, seven are known to prey on salmonids (Brown et al. 2013a). Although rates of predation specifically for Chinook Salmon are in many cases unknown, it is generally understood that in some cases marine mammal predation can play a significant role in mortality rates for certain Chinook Salmon stocks.
Simulation modelling indicates that mortality rates of Chinook Salmon from marine mammal predation increased in the 1990s relative to levels during the preceding 30 years (Figure 26). These results may be partially explained because predators were eating a higher proportion of fish due to declining Chinook Salmon populations (Bailey pers. comm. 2018), although other models suggest this interpretation may not be supported by the evidence (Preikshot et al. 2013). Nevertheless, populations of Sea Lions, Harbour Seals, White-sided Dolphins, and Humpback Whales dramatically increased since the 1970s and may have led to higher consumption (Riddell et al. 2013; Chasco et al. 2017). Between 1970 and 2015, for example, the annual biomass of Chinook Salmon consumed by pinnipeds in Puget Sound (Washington State, USA) rose from 68 to 625 tons (Chasco et al. 2017).
Northern and southern resident killer whales – which in 2013 totalled approximately 350 animals in BC waters – are considered salmonid specialists (Brown et al. 2013a). These whales congregate in groups during summer and fall in specific areas to intercept salmon migrating to natal spawning rivers. Extensive field studies of foraging behaviour indicate that resident killer whales forage selectively for Chinook Salmon and, to a lesser extent, Chum Salmon (Ford and Ellis 2006; Hanson et al. 2010). The whales appear to target large fish, with most being four years of age or older. Riddell et al. (2013) discuss workshop findings that identified the South Thompson Chinook Salmon population (DU12, DU13 and DU14) as the dominant stock in the diet of southern resident killer whales. Other Fraser River stocks of Chinook Salmon, some of which are declining, also figure prominently in the diet of resident killer whales. While only assessed during a single year and not considered in relation to relative DU abundance for that year, Hanson et al. (2010) ranked each DU in terms of inferred importance as follows: Upper Fraser (DU11), Middle Fraser (DU7, DU8, DU9, DU10), South Thompson River (DU12, DU13, DU14), and Lower Fraser stocks (DU2, DU3, DU4, DU5, DU6).
Harbour Seal abundance along the Pacific coast has increased dramatically since harvests ended in the late 1960s (Brown et al. 2013a). Consistent with trends south of the border, Harbour Seal abundance increased in the Strait of Georgia at a rate of 11.5 percent per year after the mid-1970s before stabilizing in the mid-1990s at about 40,000 animals. This trend is typical of the BC coast generally, with current total abundance estimated at 105,000 animals (Olesiuk 2010). Extensive scat collections during the 1980s indicated that Harbour Seals in the Strait of Georgia consumed a wide variety of prey species, but their diet was dominated by Herring and Hake. Overall, salmonids represented only about 4-7 percent of their diet, with salmonid consumption concentrated on pre-spawning adult salmon in estuaries and rivers (Olesiuk 1993; Thomas et al. 2016). Such predation can potentially be a major source of mortality for returning adult Chinook Salmon in cases where run size is small and habitat modification increases vulnerability to predation (for example, channelization of lower Puntledge River). Juvenile salmon, including Chinook Salmon, are also preyed upon by Harbour Seals (Thomas et al. 2016). Predation of juveniles can occur in marine areas as well as in rivers. Predation rates of downstream migrating juveniles can be significant in areas that are artificially illuminated at night such as bridge crossings (for example, Puntledge River, Olesiuk et al. 1996). The constrained morphology of a river can increase vulnerability to highly mobile and agile predators such as seals. The extent of predation on juvenile Chinook Salmon by Harbour Seals in natural settings is currently unknown. Chasco et al. (2017) estimate that between 1970 and 2015 the annual biomass of Chinook Salmon consumed by pinnipeds (Harbour Seals [Phoca vitulina], California Sea Lions [Zalophus Californianus], Steller Sea Lions [Eumetopias jubatus]) in Puget Sound increased from 68 to 625 metric tons. By 2015, pinnipeds consumed double that of resident killer whales and six times the combined commercial and recreational catches.
Steller Sea Lion abundance in British Columbia has increased approximately three-fold in BC since harvesting ended in the late 1960s (Brown et al. 2013a). Prior to 2013, abundance was increasing at 5 percent per annum and, based on pup production, current abundance in BC and adjacent waters of Southeast Alaska is approximately 60,000 animals, which is considerably greater than the estimated abundance for the early 1900s. Steller Sea Lions range widely in coastal waters, but during summer the majority congregate at traditional breeding rookeries, the largest of which are found in the Scott Islands, off the north end of Vancouver Island, and at Forrester Island, Alaska just north of the Haida Gwaii (Queen Charlotte Islands). Diet studies using prey remains found in scats collected at these rookeries and other haul-out sites indicate that Steller Sea Lions feed on a variety of fish and cephalopods, and that salmon constitutes a significant portion of their diet particularly in summer and fall. Salmonids have been estimated to represent about 10 percent of their overall diet (Olesiuk et al. 2010). On average, Steller Sea Lions eat about 18 kg of prey per capita per day, which may translate to about 17,200 tonnes per year for the population that uses Canadian waters. Preliminary studies on the salmonid species composition of Steller Sea Lion diets indicates that Chinook Salmon may represent a significant component of salmonids consumed (Olesiuk et al. 2010).
Other species of marine mammals in the region that are known to consume salmonids include Dall's porpoise, Pacific White-sided Dolphin, California Sea Lion, and Northern Fur Seal. The extent of predation by these species on different species of salmonids is poorly known (Ford and Olesiuk 2010).
There is no consistent association between harmful algal blooms and Chinook Salmon marine survival (Riddell et al. 2013). However, Cowichan River (DU21) and Dome Creek (DU11) stocks did have lower survival during major bloom years, as shown by a large negative mean survival anomaly compared to a mean survival rate, suggesting that the location of blooms might affect some populations more than others (Figure 27) (Riddell et al. 2013).
There is no evidence of a correlation between southern BC Chinook Salmon marine survival and competition with Pink Salmon juveniles, as no consistent patterns were observed in years exhibiting contrasting abundance of juvenile Pink Salmon in the Strait of Georgia (high abundance of juvenile Pink Salmon in even years, and almost absent in odd years) (Riddell et al. 2013) (Figure 28).
Chart illustrating bivariate Pearson correlation coefficients of marine survival rates and distance between marine entry points for BC Chinook Salmon stocks.
Chart showing the common time trend in the CWT-based age-2 cohort marine survival rate, for ocean-entry years 1995 to 2009.
Multi-panel chart illustrating three climate indices (Southern Oscillation, Pacific Decadal Oscillation, North Pacific Gyre Oscillation) reflecting temperature and oceanographic conditions in the North Pacific Ocean.
Multi-panel chart illustrating marine survival anomalies in hatchery indicator stocks for six Strait of Georgia fall Chinook Salmon populations.
Multi-panel chart illustrating marine survival anomalies in hatchery indicator stocks for four Strait of Georgia spring and summer Chinook Salmon populations and three outer coast Vancouver Island fall and summer populations.
Chart illustrating a co-path model simulation of adult Chinook Salmon mortality due to marine mammal predation and fishing.
Chart illustrating the marine survival of BC Chinook Salmon (by indicator stock) in relation to harmful algal blooms.
Multi-panel chart illustrating the effect of Pink Salmon on the survival rates of BC Chinook Salmon in three regions of the Strait of Georgia.
Since the beginning of the Salmonid Enhancement Program Chinook Salmon populations have been supplemented continuously, particularly those from the coastal areas of Vancouver Island and the Lower Fraser River (MacKinlay et. al. 2004).
COSEWIC considers hatchery fish as part of the population provided they supplement wild populations. The resultant naturally produced offspring are included in the application of quantitative criteria provided these individuals are predicted to provide a net positive impact on the wildlife species assessed and do not decrease the average fitness of individuals in the population (Guideline #7 in Appendix E7: Guidelines on Manipulated Populations, COSEWIC 2010).
Withler et al. (2018) recently summarized the current knowledge of the impacts of hatchery operations on wild stocks of Chinook Salmon. They conclude that hatcheries represent a risk factor to wild genetic diversity that requires management and mitigation to safeguard Pacific salmon biodiversity in Canada.
Based on data through 2013, the impact of hatcheries varies among DUs. The Middle and Upper Fraser DUs have had a low level of hatchery production over the last three generations, and therefore hatchery programs are unlikely to have influenced population viability or trends in abundance in natural populations in recent years (Riddell et al. 2013). Hatchery production was deemed unlikely to have a direct effect on the Thompson River DUs or Lower Fraser DUs. Several lines of evidence indicate that the hatchery program has reduced productivity and caused declines in abundance in Georgia Strait DUs and West Vancouver Island DUs, but data are limited. In these DUs, there are indications of significant genetic change and homogenization (Riddell et al. 2013).
It is important to note that widespread efforts to enhance Chinook Salmon populations started in the 1970s and hatcheries were constructed in a number of sites in the Fraser River watershed (Bailey pers. comm. 2018). At that time, the biology of many Fraser River stocks was poorly understood, and many projects attempted to enhance returns by producing age 0+ smolts in stocks that naturally smolted as yearlings. Success was rare, and many projects were discontinued, including hatcheries at Fort St. James, Quesnel River, Eagle River and Clearwater River (Bailey pers. comm. 2018). While these early attempts have been discontinued, it is unknown whether the attempted life-history manipulations resulted in any long-term negative impacts on the stocks. Enhancement continues in some DUs within the Fraser River system, and is strictly regulated and licensed. Most projects now feature native broodstocks, and produce juveniles released at similar timing and sizes to the natural populations. Previously, some intentional extra-DU crosses were carried out, and releases in non-natal DUs still occur. Additionally, some juveniles are released at much larger sizes than naturally produced juveniles which has resulted in greater survivals and altered maturation schedules. The long-term impact of enhancement on Southern BC stocks is not well understood. Overfishing of natural populations may occur in mixed-stock fisheries, but other impacts may be more subtle and difficult to detect. Recent advances in genetic and epi-genetic science suggest that the impact of hatchery enhancement may not be positive, and that inappropriate strategies combined with inter-basin transfers may compound those impacts.
Hatchery production most strongly affects variation in population trends (that is, annual proportional changes in abundance) (Hoekstra et al. 2007). There is evidence that enhancement may pose a risk to natural populations. Myers et al. (1998) argue that artificial propagation poses a number of genetic risks for natural salmon and Steelhead Salmon populations in addition to the complications it brings to evaluation of natural replacement rates. Interbreeding of hatchery and natural fish can lead to loss of fitness in local populations, for example, loss of local adaptations, loss of genetic diversity among populations. There is also some evidence that the survival rate of hatchery fish is lower than wild fish during times of low ocean productivity (Nickelson 1986; Zimmerman et al. 2015 – Coho Salmon, Beamish et al. 2012 – Chinook Salmon). However, there is little correlation between the scale of hatchery releases and marine survival of hatchery origin fish except for east coast of Vancouver Island stocks (Riddell et al. 2013) (Figure 29). Another critical potential effect is reduced reproductive capacity of natural origin fish (see Nickelson 2003; Buhle et al. 2009; Chilcote et al. 2011; Ford et al. 2012).
Gardner et al. (2004) prepared a comprehensive review of the information available on interactions between hatchery origin and wild origin salmon for the Pacific Fisheries Resource Conservation Council. The positive and/or negative effect of each interaction type was explained, with examples. In order of risk, Gardner et al. (2004) drew the following conclusions:
- mixed stock fishing: Research has identified situations where wild salmon have been negatively affected on a large scale, with the worst impacts related to the Georgia BasinFootnote 1 hatcheries. Fishery management strategies and fishing techniques are being directed towards lessening mixed stock harvesting of wild salmon
- genetic interactions: Although evidence of actual impacts is scarce, current theory indicates that enhancement could significantly reduce the genetic diversity and fitness of wild salmon. Genetic changes to hatchery fish may well be inevitable; the uncertainty relates to how extensive these changes are and how strongly they affect wild salmon. Hatchery practices implemented in recent years have alleviated some risks of genetic impacts, but the risks are still significant
- competition: Some studies have pointed to negative impacts on wild salmon as enhanced salmon consume food supplies that would otherwise be available to wild salmon. The increasing concerns about limited carrying capacity relate to both freshwater and at-sea habitats
- predation interactions: Shown to have a negative impact on wild salmon in some cases but other studies indicate that the presence of enhanced salmon can have a neutral or even positive effect on predation on wild salmon
- no studies illustrating negative fish health interactions between enhanced and wild salmon could be found through this research. Nevertheless, the potential for disease transfer between enhanced and wild fish does exist, because conditions in hatcheries can promote the spread of disease, which in theory can then be transferred to wild fish through water or fish-to-fish
- negative impacts of enhancement facilities on local fish habitat are possible, but research has not identified these as being significant. These facilities are the most localized and the easiest of the potential interactions to mitigate or avoid
Weber and Fausch (2003) also present information on the negative effects of hatchery production in the presence of wild Chinook Salmon. They found that despite the intent to reduce pressure on wild salmon stocks, the presence of hatchery-origin fish can have a suppressing effect on non-enhanced populations. Among a number of potential effects, artificially increasing the abundance of a few stocks in the presence of wild stocks can allow for higher fishing pressure by recreational and commercial fishing groups because of the greater overall abundance of fish (unless hatchery fish are marked and selectively retained), and high release mortality of wild unmarked fish. Although the propagation of hatchery-based stocks can continue with lower abundances, wild stocks at similarly low productivity levels are unable to withstand this level of exploitation.
In their pre-release life stage, hatchery-reared Chinook Salmon are not exposed to predators so normal aggressive behaviours result in reward (more food resources) rather than exposure to potential predation. These fish can also show rapid local adaptation to hatchery environments (Christie et al. 2012). After release into the natural environment, the generally larger, more aggressive hatchery smolt can out-compete their wild origin counterparts for food resources (Weber and Fausch 2003). This competitive advantage, in the situation of limited food, can be detrimental to wild fish, decreasing their initial survival.
Another application of hatchery practices not considered by Gardner et al. (2004) includes the use of seapen releases. This strategy has been used as a technique for increasing recreational fishing opportunities in marine areas, particularly off the West Coast of Vancouver Island (also in Vancouver harbour and Burrard inlet). Some seapen release sites (for example, Discovery Pass, Maclean Bay, Poett Nook) do not have local Chinook Salmon populations associated with the site. As a result, returning seapen Chinook Salmon have difficulty finding creeks and rivers to spawn in. In the Campbell River area, Quinsam Hatchery Chinook Salmon were released from a seapen site in Discovery Pass, and were subsequently found in small creeks on Quadra Island (Drew, Granite Bay and McKercher) that do not have the right spawning habitat for Chinook Salmon (DFO NuSEDs data). Additionally, returning seapen releases could be straying into rivers and breeding with established Chinook Salmon populations that are distinct from the seapen population, with unknown effects on the genetics of the locally adapted Chinook Salmon population (Brown et al. 2013a). CWT marking of seapen releases is infrequent, precluding characterization of movement patterns using existing study designs; however, recent research by DFO suggests higher 'stray' rates than previously considered (Bailey pers. comm. 2018).
The hatchery programs in the Middle-Upper Fraser River, Thompson River, and Lower Fraser River DU groups have been reduced to levels where the risk to wild Chinook Salmon is small and additional hatchery monitoring would yield little contribution to understanding hatchery impacts. However, hatchery programs in the Strait of Georgia and the WVI DU groups influence many aspects of Chinook Salmon natural population ecology and dynamics.
Multi-panel chart illustrating the marine survival of east coast Vancouver Island Chinook Salmon stocks in relation to the number of hatchery fish released. Panels are provided for four rivers.
Pathogens and aquaculture
The risk of population level impacts of pathogens on southern BC Chinook Salmon is inconclusive because monitoring of wild salmon populations for disease is largely non-existent in BC (Riddell et al. 2013). However, laboratory studies and observations from captive (farmed or hatchery) Chinook Salmon have shown that pathogens and disease can cause mortality. Three pathogens were identified as potential risks to Chinook Salmon productivity – Renbacterium salmoninarum, Aeromonas salmonicida, and Vibrio anguillarum (Riddell et al. 2013).
The risks of open net-pen salmon aquaculture on wild Chinook Salmon is considered low but data are limited (Riddell et al. 2013). The risk of transmission from farmed Atlantic salmon to wild Chinook Salmon is thought to be low because of differences in susceptibility to various diseases.
Global climate change
Globally, land and air temperatures have been increasing (NOAA 2017). The effects of climate change could produce coherent trends over a broad spatial scale, including within the freshwater environment. These changes are expected to manifest as increased stream temperatures to critical levels (>18°C) (Morrison et al. 2002; Ferrari et al. 2007), changes in stream flow rate and seasonality (Dery et al. 2012), increased glacier melt (Schiefer et al. 2007; Stahl et al. 2008), increased contaminants (Harvell et al. 2002; Noyes et al. 2009, Sanderson et al. 2009; Walker and Winton 2010), and changes in the marine environment (Moore et al. 2008; Mooney et al. 2009; Rensel et al. 2010). The possible impacts of such changes for salmon are summarized for different life stages below (Puget Sound Partnership (2017) and IUCN (2009)).
Freshwater eggs, fry and juveniles
Increased winter floods may increase scour of eggs, or increase mortality of rearing juveniles where flood refugia are not available. Rearing juveniles could also be displaced to less desirable habitats. A reduction in summer flow levels will serve to increase water temperatures further and is likely to reduce the overall habitat available to salmon. Increased summer temperature may decrease growth or kill juvenile salmon. Increased winter flows are likely to scour the river beds, disturbing nests and causing physical damage to both salmon eggs and juveniles.
Marine sub-adults and adults
Many of the food webs that include salmon are expected to be disrupted by climate change. For example, the timing of planktonic blooms required by young salmon is governed by climatic factors. Changes in the timing of these blooms could cause a scarcity of food at a critical life stage (IUCN 2009). Also, oceans help absorb increased concentrations of atmospheric carbon dioxide but this has caused a 30 percent increase in ocean acidity since 1750 (Orr et al. 2005). Consequences are expected to impact ocean food webs, including increased predation impacts on salmon (Fabry et al. 2008).
As freshwater temperatures increase, a number of negative effects on adult salmon may arise. Direct biological impacts on salmon include physiological stress, increased depletion of energy reserves, increased susceptibility and exposure to disease and disruptions to breeding efforts (IUCN 2009). Temperature-related barriers to migration and possible decreased access to or availability of spawning habitat may also occur. However, there is currently no evidence that increased summer temperatures will decrease spawning fecundity for salmon.
For many Chinook Salmon DUs, snow is predicted to be replaced by rain. This shift will lead to a reduction in summer flows for many rivers, coupled with an increase in freshwater inputs during the winter. At a regional scale, an ensemble of 30 projections to 2070 show that projected warming will be greater in the Interior portions of southern BC compared with the coastal region (Pike et al. 2010). Nelitz and Porter (2009) described the projected changes to the Thompson and Chilcotin watersheds as a result of climate change. The general patterns of this analysis suggest that regional climate change impacts on Chinook Salmon may be mixed. In some areas, there may be benefits of habitat changes, while in others there may be constraints on production. For instance, stream habitats with temperatures optimal for Chinook Salmon rearing are predicted to decrease in northern areas of the study area and increase in southern areas. However, reductions in late summer/early fall flows will create challenges for rearing juveniles and for spring and summer spawners. As a result, Chinook Salmon populations are predicted to decrease more markedly in the north than in the south. In some of the more northern streams summer/fall flows are predicted to decline to such an extent that minimum flows to support successful spawning and rearing may not be reached consistently in the future.
The scope of change in southern BC watersheds is likely to be widespread. The severity of such impacts is unknown at present, but could be serious, particularly in those DUs already stressed by other anthropogenic activities. In a recent review of southern BC Chinook Salmon threats, climate change was identified as likely already affecting southern BC Chinook Salmon, but no definitive conclusions could be drawn from current evidence (Riddell et al. 2013).
Format of designatable unit-specific chapters
In the following DU-specific chapters, the information covered for each DU will include:
- Names, life-history type, run-timing and generation time
- Extent of occurrence and area of occupancy
- Habitat trends
- Fluctuations and trends
- Threats and limiting factors
Names, life-history type, run-timing and generation time
Each DU chapter begins by listing the full DU name, the DU short name, the Joint Adaptive Zone (JAZ) short name, the life-history type (Ocean or Stream), the run-timing type (Fall, Spring, Summer), and generation time. Generation time is estimated as the average age of spawners in the absence of fishing mortality. These figures are based on CWT indicator stocks (listed below in Table 12). Where indicator stocks are not available within a DU, proxy indicator stocks are used (shown in Table 18). For southern BC Chinook Salmon DUs, all the CWT indicator stocks are integrated hatchery stocks. Since both natural and hatchery origin fish are used as brood stock and CWTs are applied to their progeny, it is assumed that other natural origin fish in the DU are represented reasonably by the indicator stocks. This assumption is often made with southern BC Chinook Salmon, but it is well known that these indicator stocks were chosen by convenience, and not by random selection or any other manner intended to accurately represent the characteristics of the conservation unit. These are currently the best data available for the purpose of estimating generation time (G. Brown, pers. comm.).
Extent of occurrence and area of occupancy
For each DU, extent of occurrence and area of occupancy data are reported at the Designatable Unit (DU) level of analysis (see the Designatable Unit Delineation section of this report). The spatial extent of all DU boundaries is shown in Figure 30, this coverage represents the terrestrial (that is, freshwater) extent of occurrence for the southern BC Chinook Salmon assessed in this report.
DU boundary delineations were adapted from CU Report Cards developed by Porter et al. (2013) which used third-order plus watersheds from the 1:50,000 British Columbia Watershed Atlas (www.env.gov.bc.ca/fish/watershed_atlas_maps/, presently not an active link) as a base spatial scale of analysis. Prior to release of the Porter et al. report, some of these CU boundaries were modified to allow for DFO-defined changes; as a result, associated metrics were recalculated. Generally, DU boundaries used in this report correspond to CU boundaries. In the cases of DU12 and DU21, multiple CUs comprise the DU. For DU-specific chapters, individual DU areal extents are estimated in GIS software using geospatial shapefiles. The DU map in Figure 30 is confirmed as up-to-date and accurate as of 2012. However, after the report's release, the spatial extent of the CU areas were again redefined – in all cases they were expanded. At the time of writing, data were unavailable for these revised boundaries.
The marine extent of Chinook Salmon cannot be precisely defined geospatially due to lack of available data, but the extent of occurrence for all southern BC Chinook Salmon is known to be >20,000 km2. According to harvest statistics, Chinook Salmon ocean ranges extend northward to Southeast Alaska (Riddell et al. 2013). Ranges specific to southern BC Chinook Salmon vary depending on life-history strategy with 'local' stocks moving as far north as central Queen Charlotte Islands and as far south as the Columbia River mouth (Bailey pers. comm. 2018). 'Offshore' stocks are believed to range as far north as the Bering Sea and into the North Pacific Gyre (Bailey pers. comm. 2018).
Following methods used for the COSEWIC Fraser Sockeye Salmon Status Report, the area of occupancy of each DU is calculated as two times the spawning length, and is reported in square-kilometres. This method is equivalent to overlaying a 2×2 km2 grid over the stream, and adding up the total area. To assist in comparison across DUs, each DU description also states the proportion of spawning habitat within each DU relative to the total across all DUs. Chinook Salmon spawning extents were provided by the Province's Fisheries Information Summary System (FISS), and are meant to cover the total linear length of known Chinook Salmon spawning habitat within each DU. FISS presently represents the best available data in GIS format; however. the database is known to be incomplete due to a lack of comprehensive source information for southern BC Chinook Salmon distributions (Porter et al. 2013).
Map illustrating the areal extent of the freshwater area for all southern BC Chinook Salmon DUs.
Habitat trends reported for each DU's freshwater-based area describe some of the known indicators adapted from the Porter et al. (2013) DFO Report Card data. Reported trends include land-based habitat alteration, urban development, rural development, mining, road density, the number of stream crossings, riparian habitat disturbance, forest disturbance and Mountain Pine Beetle-affected pine stands.
Sampling effort and methods
While total abundance is the most desirable metric for this category, such data are unavailable for many DUs so this report relies on escapement data. Escapement data quality and quantity varies across DUs and over time. Escapement, defined as the number of fish arriving at a natal stream or river to spawn (also termed 'spawner abundance') can be assessed by presence/absence, relative abundance, or total (“true”) abundance. The New Salmon Escapement Database System (NuSEDS) is a centralized database that holds adult salmon escapement data used by Fisheries and Oceans Canada (DFO). Escapement data used for the status metrics in this report originated from NuSEDS, with the understanding that not all escapement data from NuSEDS represent absolute abundances.
In 2013, DFO undertook a process to determine thresholds for data quality of escapement data that included a three-day workshop in February 2013. The NuSEDS Estimate Classification scheme (Table 11) was central to selecting data considered to be sufficient in quality and completeness to be used for calculation of status metrics. The process is described in greater detail in (Brown et al. 2013b), but it is also described briefly here. Data considered suitable for use were Type-1 through Type-4 estimates only ('true abundance' and 'relative abundance'). When using the NuSEDS Estimate Classification scheme, over 61 percent of escapement records between 1953 and 1995 were excluded due to missing Estimate Classification information, and therefore marked as 'unknown'. DFO identified the missing data as a high priority for 'data rescue'. The NuSEDS Estimate Classifications were further grouped into high, moderate, low and unknown categories: H (High) = True Abundance (Type 1 or 2), M (Mod) = Relative Abundance (Type 3 or 4), L (Low) = Relative Abundance (Type 5) or Presence/Absence (Type 6), and ? (Unknown) = Type Unknown is reported in the database or is blank.
Within a DU or, in the case of the WSP, within a CU, a key challenge is how to combine data for 'true abundance' (Type 1 or 2) with data for 'relative abundance' (Type 3 or 4). In many cases where multiple spawning sites existed, relative abundance estimates were summed with true abundance estimates to arrive at total abundance within the DU. In these cases, the entire DU was considered a 'relative abundance index'. Under the WSP CUs, there were 4 CUs that were considered to provide actual abundance: CK-03, CK-15, CK-21, and CK-22. However, when combined into DUs, only DU2 (CK-03) was considered to provide actual abundance (CK-15 is combined into DU12 with CK-13; CK-21 and CK-22 are combined into DU 21 with CK-25 and CK-27). In both relative abundance index and actual abundance cases, all CUs/DUs had considerable past and current enhancement (Brown et al. 2013b).
|DFO ranking||Estimate type||Survey method(s)||Analytical method(s)||Reliability (within stock comparisons)||Units||Accuracy||Precision||Documentation|
|Type-1, True abundance, high resolution||total, seasonal counts through fence or fishway; virtually no bypass||simple, often single step||reliable resolution of between year differences >10 percent (in absolute units)||absolute abundance||actual, very high||infinite that is,+ or - zero percent||detailed SIL(s), SEN, field notes or diaries, published report on methods|
|Type-2, True abundance, medium resolution||high effort (5 or more trips), standard methods (for example, mark-recapture, serial counts for area under curve, etc...)||simple to complex multi-step, but always rigorous||reliable resolution of between year differences >25 percent (in absolute units)||absolute abundance||actual or assigned estimate and high||actual estimate, high to moderate||detailed SIL(s), SEN, field notes or diaries, published report on methods|
|Moderate (M)||Type-3, Relative abundance, high resolution||high effort (5 or more trips), standard methods (for example, equal effort surveys executed by walk, swim, overflight, etc.)||simple to complex multi-step, but always rigorous||reliable resolution of between year differences >25 percent (in absolute units)||relative abundance linked to method||assigned range and medium to high||assigned estimate, medium to high||detailed SIL(s), SEN, field notes or diaries, published report on methods|
|Moderate (M)||Type-4, Relative abundance, medium resolution||low to moderate effort (1-4 trips), known survey method||simple analysis by known methods||reliable resolution of between year differences >200 percent (in relative units)||relative abundance linked to method||unknown assumed fairly constant||unknown assumed fairly constant||complete SEN or equivalent with sufficient detail to verify both survey and analytical procedures|
|Type-5, Relative abundance, low resolution||low effort (for example, 1 trip), use of vaguely defined, inconsistent or poorly executed methods||unknown to ill defined; inconsistent or poorly executed||uncertain numeric comparisons, but high reliability for presence or absence||relative abundance, but vague or no identification of method||unknown assumed highly variable||unknown assumed highly variable||incomplete SEN, only reliable to confirm estimate is from an actual survey|
|Type-6, Presence or absence||any of above||not required||moderate to high reliability for presence/absence||(+) or (-)||medium to high||unknown||any of above sufficient to confirm survey and reliable species identification|
Sample sites within a DU were assessed based on the quality and completeness of their time series. The full list of sample sites is presented in Table 43 (Appendix 2) of this report along with start/end dates for each estimation method. Note that sites with very low contributions are not included in the figures reported in Panels c and d of the 'abundance, enhancement, and hatchery release' data graphics (Figure 31) of each DU chapter (for example, Wap Creek in DU12). The process is described in greater detail by Brown et al. (2013), and relied on the following criteria:
- sites must be 'persistent'. 'Persistent' sites ('P') were defined as those having more than 50 percent high quality observations (Type-1 to Type-4) during the period Start Year to 2012, with no more than one generation of years missing in sequence. For example, for CUs with a start year of 1995, this translates into at least 10 years of high quality data from the period that was part of the in-depth data review, and no more than 3, 4 or 5 years in a row missing (depending on the average generation time for the CU) for each persistent census site in the CU
- for sites with marginal numbers of high quality observations during the Start Year to 2012 period, the pattern of missing data was investigated to determine if it could be infilled to provide a sufficiently complete time series (that is, the pattern of missing observations for the census site did not include a full generation—based on the average generation time for the DU—at any point in the Start Year to 2012 period). Those that could meet the sufficiency criteria with infilling were identified as 'P', and the rest were classified as data deficient ('DD')
- if a site had 50 percent or fewer high quality observations during the Start Year to 2012 period and could not be infilled to achieve a 50 percent level, it was categorized as 'DD'
When combining data from more than one site within a CU that contained years with missing data, infilling was performed by DFO. This process is described in greater detail in Brown et al. (2013b). Infilling followed the procedure outlined in English et al. (2006), whereby the average proportion (across years) each census site contributed to the total was calculated, and used to infill years with no escapement data. When the time series of several CUs within a DU were combined (that is, DU12 and DU21), the same English et al. (2006) approach was adopted.
Within the DU-specific chapters, two pieces of information are presented when available:
- the proportion of spawners originating from census sites of varying data quality
- the number of spawners above and below threshold abundances
The calculation of the proportion of spawners originating from census sites of varying data quality combines the work on categorizing the data quality of abundance estimates with the work on categorizing the data quality from different census sites. The number of spawners above or below threshold abundance is based on COSEWIC quantitative criteria for the total number of mature individuals. The benchmarks are 10,000, 2,500, 1,000, and 250 individuals.
Coded-wire tags (CWT) are used as a source of detailed information for many populations of Chinook Salmon along the Pacific coast of North America (Hankin et al. 2005; Nandor et al. 2010). Chinook Salmon populations with consistent annual releases of CWTs are referred to as CWT indicator stocks and are used to represent naturally spawning wild stocks which exhibit the same adult and juvenile life-history patterns and are assumed to exhibit the same behavioural patterns within a similar geographic area. To produce sufficient CWTs for analysis, most of the CWT indicators are tied to hatchery programs, where fish are reared, tagged, and released. There are 11 Canadian CWT indicator stocks distributed among southern BC Chinook Salmon DUs (Table 12). Most of these stocks are from large-scale conventional hatchery facilities with five located within the Fraser River drainage (DU2, DU11, DU12, DU15, and the Chilliwack River) and six distributed around Vancouver Island (DU19, DU20, DU21, DU22, DU23, and DU24). Two hatcheries have been terminated in recent years (DU11 and DU21 – Nanaimo River) but funds administered by the Coded Wire Tag Improvement Team of the Pacific Salmon Commission (PSC) have been used recently to improve aspects of the others (PSC-CTC 2012a).
Information provided by CWTs includes ocean distribution (via catch of tagged fish vulnerable to fishing gear), exploitation, smolt survival, and mean age at maturity. The Pacific Salmon Treaty (PST) between Canada and the United States supports annual sampling programs to collect information from CWT indicator stocks using a consistent and unbiased design (Brown et al. 2013b). Information from CWT indicator stocks is obtained from the cohort analysis output files, which extend to the end of 2012, and were used to produce the Chinook Salmon Technical Committee (CTC) 2013 annual report (Brown et al. 2013b). The details of the cohort analysis procedure are described in PSC-CTC (1987).
|Indicator stock site/name||Indicator stock acronym||DU number||Run type||Release information
mean associated non-CWT
|Estimated CWT information
|Estimated CWT information
|Estimated CWT information
|Estimated CWT information
|Estimated CWT information
|Estimated CWT information
|Chilliwack R||CHI||N/A||Fall||10||101,904||472,864||12||4153||9.2 percent||15.0 percent||7.4 percent||68.4 percent|
|Harrison R||HAR||DU2||Fall||9||149,096||804,461||12||1113||10.5 percent||20.9 percent||1.6 percent||66.9 percent|
|Dome Cr||DOM||DU11||Spring||3||83,602||3,718||8||155||1.8 percent||23.5 percent||50.1 percent||24.6 percent|
|Lower Shuswap R||SHU||DU12||Summer||10||186,708||370,005||12||1444||15.4 percent||26.8 percent||9.5 percent||48.3 percent|
|Nicola R||NIC||DU15||Spring||9||107,174||46,275||12||1089||1.2 percent||6.3 percent||10.3 percent||82.3 percent|
|Puntledge R||PPS||DU20||Summer||10||115,953||508,058||12||290||15.9 percent||23.4 percent||0.0 percent||60.7 percent|
|Cowichan R||COW||DU21||Fall||9||299,815||1,209,989||12||781||12.7 percent||48.0 percent||6.1 percent||33.2 percent|
|Nanaimo R||NAN||DU21||Fall||4||145,257||96,884||9||819||7.8 percent||33.7 percent||6.7 percent||51.8 percent|
|Big Qualicum R||BQR||DU21||Fall||10||235,183||3,388,613||12||501||15.1 percent||26.3 percent||2.2 percent||56.4 percent|
|Quinsam R||QUI||DU23||Fall||10||287,024||1,842,503||12||814||22.6 percent||20.3 percent||0.1 percent||57.1 percent|
|Robertson Cr||RBT||DU24||Fall||10||256,807||6,153,023||12||2360||20.2 percent||16.0 percent||27.1 percent||36.7 percent|
Wild-born fish cannot be distinguished from their hatchery counterparts with certainty unless mass marked. However, mass marking is not currently employed in Canada for Chinook Salmon (only hatchery Coho Salmon). Therefore the authors of the Pre-COSEWIC report adopted a higher-level approach based on categorizing sites by enhancement activity level (Brown et al. 2013b). Census sites within Designatable Units were assigned a level of enhancement based on a standardized procedure developed by DFO during the Pre-COSEWIC process (Brown et al. 2013b). The standardized rank classified the census sites as:
- category 1. Unknown (no evidence of recent active enhancement) - no release records, brood records or enhanced contribution estimates during the period 2000-2011
- category 2. Low enhancement activity level - release records, brood records or enhanced contribution estimates exist prior to 2000 but there were none from 2000-2011
- category 3. Moderate enhancement activity level, defined as:
- number of release records is less than or equal to 4 out of 12 years (≤25 percent or roughly 1 per generation)
- number of brood take records are less than or equal to 4 out of 12 years (≤25 percent or roughly 1 per generation)
- hatchery-origin contribution estimate is available via expanded CWT data and 12-year mean is <25 percent (assessing adult contribution only)
- category 4. High enhancement activity level, defined as:
- number of release records exceeds 4 out of 12 years (>25 percent or >1 per generation)
- number of brood take records exceeds 4 out of 12 years (>25 percent or >1 per generation)
- hatchery-origin contribution estimate is available via expanded CWT data and 12-year mean is ≥25 percent (assessing adult contribution only)
For each DU-specific chapter, a figure is presented showing the proportion of spawners originating from wild-born and enhanced census sites. These figures are updates to 2015 from the figures developed for the Pre-COSEWIC report (Brown et al. 2013b), and are adapted from CU-level time series of escapement for wild and enhanced sites. Where multiple CUs are combined within a DU (DU12, DU21), figures for each individual CU are included. In developing these figures, CU-level time series of escapement for the wild-born sites and the enhanced sites were created. Data were combined from sites with low or unknown levels of enhancement ('Low+Unk') and sites with moderate or high levels of enhancement ('Mod+High').
For each DU, the time series of hatchery releases from within the DU and/or from outside the DU are presented. These are reproductions of the 'dashboard' graphics found in Brown et al. (2013b). When a DU is a combination of several CUs (DU12, DU21), the time series for each individual CU is included.
Interpretation of abundance, enhancement and hatchery release data
Data permitting, abundance, enhancement and hatchery release data are presented graphically for each DU as a six-panel figure (see example for DU2 below). Table 13 describes how to interpret each panel:
Multi-panel figure containing examples of the charts and tables (a to f) used to illustrate abundance, enhancement, and hatchery release data for individual BC Chinook Salmon DUs.
The following table explains how to interpret each panel and can be used as a guide while reviewing each DU.
percent Spn From Wild Sites
|Observed number of spawners in sites identified as wild as a proportion of spawners in all sites with data for this DU. When no data are available, no bars are present. If present, wild sites are represented in blue, and are defined as sites with unknown or low enhancement. Enhanced sites are represented in grey and are defined as sites with moderate to high enhancement. Note that this panel does not show annual estimates of enhanced contribution, it shows the proportion of spawner estimates for each year that come from sites CURRENTLY classified as either wild or enhanced. The plot is based on available site records, not on expanded estimates to account for non-surveyed populations|
percent Spn by Survey Quality
|Percent of observed spawners in sites with different levels of survey quality as coded in NuSEDS. H (High) = True Abundance (Type 1 or 2); M (Mod) = Relative Abundance (Type 3 or 4); L (Low) = Relative Abundance (Type 5) or Presence/Absence (Type 6); and ? (Unknown) = Type Unknown is reported in the database or is blank. When no data are available, no bars are present.|
Spn Range by Year
|Number of years where spawner abundance (generational average) for the whole DU falls into different abundance bins, adapted from COSEWIC's absolute abundance criterion. The first row shows the total number of years that have spawner estimates. Subsequent rows show how many observations fall above or below various cut-off points.|
Spn Range by Site
|Number of sites where spawner abundance (generational average) for the whole DU falls into different abundance bins, adapted from COSEWIC's absolute abundance criterion. The first row shows the total number of sites that have spawner estimates. Subsequent rows show how many observations fall below various cut-off points. Note that sites with very low contributions are not included in the figures reported in Panels c and d of the 'abundance, enhancement, and hatchery release' data graphics of each DU chapter (for example, Wap Creek in DU12).|
|Panels (e) and (f)
Hatchery releases from within and outside Unit (BY)
|The number of hatchery releases from within and outside the DU by brood year (BY). The left panel is the total number of hatchery-reared juveniles produced from broods collected from return sites within the DU and released at sites within the DU. The right panel is the total number of hatchery-reared juveniles produced from broods collected from return sites outside the DU and released at sites within the DU. When no data are available, no graph is present.|
Fluctuations and trends
For each DU, fluctuations and trends are presented in a summary table and a five-panel figure. The summary table provides two Bayesian estimates of changes in spawner abundance over the last three generations, one using the last three generations of data, and the other using entire time series of data. Probabilities of a 30 percent, 50 percent and 70 percent decline in spawner abundance over 3 generations are also presented for each of the two data samples. Data were available for most DUs up to the 2015 return year and were provided by DFO (Gayle Brown, pers. comm.).
The summary table for each DU has the following form (Table 14; categories described in Table 15):
|DU name||Generation length||Year range||Median percent change||95 percent CI||p|30 percent decline||p|50 percent decline||p|70 percent decline||Number of observations|
|Example DU||4||3gen||Not applicable||Not applicable||Not applicable||Not applicable||Not applicable||Not applicable|
|Example DU||4||All years||Not applicable||Not applicable||Not applicable||Not applicable||Not applicable||Not applicable|
|DU name||Full-name of each DU|
|Generation length||Average generation time estimated as the average age of spawners in the absence of fishing mortality|
|Year range||Beginning and ending year of the data set used|
|Median percent change||Median of the posterior distribution for the slope parameter outputs from Bayesian regression|
|95 percent CI||±95 percent credible interval of median percent change|
|p|30 percent decline||Probability of a 30 percent or greater decline in abundance|
|p|50 percent decline||Probability of a 50 percent or greater decline in abundance|
|p|70 percent decline||Probability of a 70 percent or greater decline in abundance|
|Number of observations||Number of observations in the data set|
Data permitting, trends in spawner abundance, exploitation rate and marine (smolt-to-adult) survival are presented graphically for each DU as a five-panel figure (see example for DU2 below). Table 16 describes how to interpret each panel:
Multi-panel figure containing examples of the graphs used to illustrate trends in spawner abundance, exploitation rate, and marine (smolt-to-adult) survival for individual Chinook Salmon DUs.
Trend in Spawner Abundance
|Trend in spawner abundance with two estimates of the linear rate of change in abundance through time: (1) rate of change over the last three generations based only on the last full three generations of data (that is, 13 years for a DU with a 4 year generation time); (2) rate of change over the last three generations based on all available data. The latter is shown because indicators of changes in abundance based on the rate of change over entire time series have been shown to be more reliable than shorter time series (Porszt et al. 2010; d'Eon-Eggerston et al. 2012). Data used for the last three generations were calculated as the generation time + 1 data point such that the selected data spanned the latest three generations. If the 3-generation time was not a round number, it was rounded up Rates of change were calculated using a Bayesian estimation framework. Doing so allowed the presentation of probabilities associated with estimated changes in abundance, which are more intuitive to interpret than frequentist confidence intervals. Bayesian modelling and parameter estimation was conducted in R using JAGS software (R Core Team 2017; Plummer 2011) with the package R2jags (Su and Yajima 2015). Uninformative priors were assumed for slope (), intercept ) and standard deviation . The linear model for a single chain used a burn-in of 5,000 observations, and retaining 100,000 samples after burn-in. Only every 5th observation was saved to reduce autocorrelation (thin=5). These settings are the same as those used for the COSEWIC Fraser Sockeye Salmon status report.|
percent change over last 3 gen. (last 3 gen. posterior distribution)
|Posterior distribution and median estimate (as vertical line) of estimated percent change over last three generations based on a linear rate of change of spawner abundances over the most recent three generations of data.|
percent change over last 3 gen. (all years posterior distribution)
|As for panel (b) but based on regression of data for entire time series.|
|Total of coded-wire tagged fish of any age from a brood (breeding stock) estimated in coast wide pre-terminal and terminal fishery catches divided by the same total plus the total estimated in the escapement. Fishery impacts include an estimate of the non-landed (incidental) mortalities, which occur when fish escape from or are released from fishing gear but later die anyway. Pre-terminal fishery mortalities have been adjusted by a brood- and age-specific adult equivalency factor which accounts for the fact that even if there were no fisheries, fish may still die before reaching the spawning grounds but the probability of surviving to spawn increases at each age (for example, a fish caught in the ocean at age 2 equates to a lower adult equivalent than a fish caught at age 4 because there is less of a chance of surviving and maturing at any possible future age compared to an older fish).|
|Estimated cohort size of fish alive at the start of the youngest possible age of mature fish divided by the number of smolts released from the parental brood year.|
Where available, stock productivity data (recruits per spawner) are also presented. Stock productivity was calculated as the total number of adults recruiting to the population (that is, spawners + catch) produced by the spawners from a given year (brood year). Only two time series of stock productivity data are available, for DU2 and DU22 (Brown et al. 2013b). The methods used to generate these productivity time series were based on Canadian Science Advisory Secretariat (CSAS) reports (Tompkins et al. 2005, G. Brown, DFO, unpublished data). Data used for these time series were provided by DFO (Cowichan River time series: M. Labelle, DFO, unpublished data; Harrison River time series: G. Brown, DFO, unpublished data).
- Date modified: