Skip booklet index and go to page content

COSEWIC assessment and update status report on the white sturgeon Acipenser transmontanus in Canada 2003



The white sturgeon is a long-lived species. Growth, age of maturity, and spawning intervals vary with location and sex. Sexually mature adults gather in aggregations to broadcast spawn during spring and early summer. Fertilized eggs fall to the bottom and adhere quickly to the substrate. Embryonic development under controlled conditions has been studied for culture purposes (Wang (1985) and Beer (1981), cited in Conte et al. 1988). At 15o C eggs hatch in 6.5 days into larvae. Yolk-sac larvae go through a dispersal period where they move vertically and swim or drift with the current for about five or six days in the lab (Brannon et al. 1986 and Conte et al. 1988). The dispersal period may be considerably shorter under natural conditions as larvae may actively select particular velocities (Perrin et al. 1999). Larvae are most vulnerable to predation during dispersal (Hildebrand et al. 1999). After the dispersal phase larvae become negatively phototaxic and hide in the substrate during daylight. At 15o C the yolk sac is absorbed around 12 to 14 days of age and the larvae begin to actively feed during daylight (Conte et al. 1988). About 20 to 30 days post-hatch, larvae metamorphose into fry (young-of-the-year). Around 55 days they orient to the current and swim freely exhibiting diurnal movement toward the surface at sunset and descending to the bottom at sunrise.

In the following sections, reported data on age characteristics are from age analyses using a small cross-section of the leading right pectoral fin ray. This is the most practical and reliable ageing method studied (Brennan and Cailliet 1989). Age estimates are less accurate for slow growing fish in colder systems or in regulated rivers with a less defined growing period. In these cases, true age can be underestimated because the growth rings form close together. This is a known problem for white sturgeon in the Kootenay and Columbia rivers (UCRRP 2002, Paragamian and Beamesderfer 2003). Under-aging may be a problem in all white sturgeon populations as the bias increases with age (Paragamian and Beamesderfer. 2003). Under estimating age in demographics such as age of sexual maturity and generation time would increase the risk of extirpation.


Length-frequency distributions for populations and age analyses indicate regular recruitment is occurring in the mainstem Fraser River, SP1-3 (RL&L 2000 and Yarmish and Toth 2002). Populations in all regulated systems (Nechako, Columbia and Kootenay rivers, SP4-6) are undergoing an extended recruitment failure (RL&L 2000, Hildebrand et al. 1999, Duke et al. 1999). The Nechako River population is aging with few fish younger than 30 years of age (Fig. 6). Recruitment reconstruction (Korman and Walters 2001) indicates recruitment declined slowly as the Nechako Reservoir filled; it then dropped suddenly in 1964 and has failed since 1967 as flow regulation impacts took effect (S. McAdam pers. comm.). Regular spawning events continue to be documented in the Canadian portion of the Columbia River; however, age structure analysis shows that recruitment began to decline in 1969 and has failed completely since 1985 (RL&L 1994). Maximum operating level for HLK was achieved in 1969 and Revelstoke Dam was completed in 1984. Studies on the Kootenay River indicate recruitment has been severely constrained since 1972, the year Libby Dam became partially operational (Duke et al. 1999).

In the Fraser River, age of sexual maturity increases from the lower river to the upper river (RL&L 2000). Below Hell’s Gate some females may spawn as young as 18 and males at 14 years of age (Semakula and Larkin 1968). The preliminary age estimate for first spawning in females in the mid-river population is late 20s; males could be younger than 20 years of age. As yet data are limited on the upper Fraser population but J. Yarmish (Lheidli T’enneh First Nation, pers. comm.) believes they are late maturing and may not reach maturity until about 30 years of age. In the Nechako River system, females may not reach maturity until their late 40s, or older, and males may not spawn until their early 30s (although this is a slow growing subpopulation, this estimate may be skewed by the lack of younger fish in the population). In the Columbia River, females may mature as early as 27 years of age and males may start to mature at 16 (RL&L 1995). In Kootenay River females, the earliest spawning may occur is at age 22 and for males at 16 (Duke et al. 1999).

Figure 6: Comparison of Age Structure in Surveys from Mainstem Fraser River and the Nechako River (RL&L 2000)

Figure 6: Comparison of age structure in surveys from mainstem Fraser River and the Nechako River (RL&L 2000).

A similar age shift has occurred in the upper Columbia and Kootenay groups.

Generation time (average age of parents of the current cohort) in the lower Fraser River appears to be around 30-35 years; in the mainstem downstream of the Nechako River it seems to be in the range of 35 to 40 years of age, while in the Nechako River system the natural generation time could be 40 years. J. Yarmish (pers. comm.) estimates that it is in the range of 40 years in the upper Fraser River, but data are still limited. In the Columbia River, generation time is estimated at 30-35 years (C. Spence, B.C. Ministry of Water, Land and Air Protection, pers. comm.). The Kootenay fish appear to have a generation time of around 30 years (C. Spence pers. comm.). Generation times are roughly estimated based on age analyses conducted to date. Another factor that may affect estimates is the possibility of reproductive senescence (S. McAdam pers. comm.). Currently, the average age of adults in the three regulated systems lacking recruitment is greater than would be expected in a natural situation; this also affects the accuracy of the estimate for these populations.

Studies by Perrin et al. (1999 and 2000) in the lower Fraser River indicate spawning occurs in the spring or early summer (May to July) at water temperatures between 11.3o and 18.4o C, either at the peak of the freshet or as flows decline. In the upper Columbia River spawning events occur at temperatures 14o to 21o C (Hildebrand et al. 1999). Spawning in the Kootenay River population has occurred between 7o and 17o C. Sticky eggs are dispersed in fast flowing water, which prevents the eggs from clumping together and smothering (Perrin et al. 2000), and disperses them downstream. Scott and Crossman (1973) reported that fecundity varies with female size and can be as high as 4 million eggs in the largest spawners. Less than 0.1% will survive their first year (Columbia River Investigations webpage).

Data from Fraser River fish examined by RL&L (2000) showed the adult sex ratio was skewed to males and varied from 4 - 4.5:1 below Hell’s Gate, to 7.6:1 in the middle Fraser, and 2:1 in the Nechako. In the Columbia River, current sampling (B.C. Ministry of Water, Land and Air Protection, unpubl. data) indicates the sex ratio is 1.35 males per female (n=94). In the Kootenay River the sex ratio was assumed to be equal, as reliable data are not available (Paragamian et al. in prep). There are no sex ratio data for upper Fraser River sturgeon. Adults survive spawning but do not spawn every year. Limited data from an earlier study indicated intervals between spawning in the lower Fraser River vary from 4 to 9 years (Semakula and Larkin 1968). In the Fraser River study, fewer than 10% of the females and 12% of the males examined in the adult size class were in the late reproductive stages (n=117), indicating that the proportion of reproductively mature individuals in a given year is quite low. The number of female spawners each year in the lower Fraser River is probably less than 100 (M. Rosenau, B.C. Ministry of Water, Land and Air Protection, pers. comm.). In the Nechako population the situation is dire; there are likely fewer than 15 female spawners per year at the present time. Sex maturity data on the Columbia River indicate that 4% of females would be capable of spawning in a particular year; in theory 14 -20 females may spawn each year, but the actual number is probably less (Hildebrand et al. 1999). The rate of spawning maturity is lower than seen in the mainstem Fraser River and may make recovery of Columbia River white sturgeon more difficult.


In impounded systems, the decrease in spring flows due to flow regulation is considered the major issue (Duke et al. 1999, Korman and Walters 2001). Kohlhorst et al. (1991, as cited in UCRRP 2002) reported that the strength of sturgeon year classes was positively correlated with discharge volume, and Anders and Beckman (1993, as cited in UCRRP 2002) related high spring flows to the amount of available spawning habitat, spawning success and rate of recruitment. Additionally, a lack of flushing flows can cause fine sediments to build up and decrease egg survival (Beamesderfer and Farr 1997, Duke et al. 1999, Korman and Walters 2001), food production (Duke et al. 1999) and juvenile habitat (Beamesderfer and Farr 1997, Duke et al. 1999, Korman and Walters 2001). The compounded effects of incremental increases in several mortality factors during early development may be enough to explain recruitment failures in the upper Columbia River (UCRRP 2002).

Flow diversion and impoundment have reduced turbidity, particularly in the Columbia and Kootenay rivers (Duke et al. 1999, Hildebrand et al. 1999, RL&L 2000, UCRRP 2002). Studies have predicted that increased water clarity may reduce the effectiveness of foraging by adults, change spawning behaviour and increase predation on sturgeon eggs and juveniles (Hildebrand et al. 1999, RL&L 2000, Gadomski et al. 2001, UCRRP 2002, Perrin et al. 2003).

Korman and Walters (2001) used an instantaneous natural mortality rate of 0.1 for the Nechako River in their model, based on Fraser River sampling data. The authors indicate that the estimated mortality is high for such a long living animal,unless mortality rate drops off very sharply for older, extremely rare fish. Korman and Walters (2001) gave three additional possibilities for the apparent high mortality: recruitment in the lower Fraser may be increasing (older fish may be rare due to poor recruitment following the peak of the historical fishery); larger, older fish may be less vulnerable to sampling; and the age of older fish may be underestimated. Another explanation for the apparent high mortality may be that relatively few fish actually achieve great age and size (Sulak and Randall 2002). Semakula and Larkin’s (1968) age composition analysis for Fraser River white sturgeon also indicated a high total mortality rate (> 0.1). Most of their sample came from commercial salmon by-catch, and they indicate that gillnet mesh sizes likely biased the size and age of sturgeon captured. None of the fish in their study were aged at > 71 years of age, but they suggest that the presence of very large, old fish in the Fraser River could indicate a mortality rate around 0.05. Duke et al. (1999) estimated a mortality rate of 0.37 for Kootenay River sturgeon. RL&L (1994) used a natural mortality rate of 0.06 in calculations to determine the rate of decline for the upper Columbia River.

Environmental Effects on Physiology

Preliminary work on the lower Columbia River (Foster et al. 2001) indicates certain chemicals, such as organochlorides, chlorinated pesticides and PCBs may be contributing to population declines in areas subject to this type of contamination. Early data from this study indicate potential contaminant effects on plasma androgen concentrations and the induction of liver enzymes, and that contaminants may also be involved with reduced condition factor and altered gonad development.

Temperature effects may play a role in recruitment problems experienced in impounded rivers. Larval survival is optimal at 14-16o C and decreases below 10o and above 20o C (Wang 1985). At the only known spawning site for upper Columbia River fish (the Pend d’ Oreille River confluence; UCRRP 2002), water temperatures often reach 21o C or higher during the spawning season, well above 18o C where mortalities increase and abnormalities occur (Conte et al. 1988; Hildebrand et al.1999). Temperatures below HLK Dam tend to be warmer in winter and cooler in summer than before dam construction (UCRRP 2002). Changes to the natural temperature regime may affect juvenile survival by altering bioenergetic requirements (for example, by increasing metabolism at a time when food resources are limited).

Another water quality parameter that may impact larval survival, particularly in impounded rivers, is total dissolved gas pressure (TGP) (Counihan et al. 1998). Spills from dams can result in TGP levels >125% (Hildebrand et al. 1999). Conte et al. (1988) recommended a nitrogen gas pressure for cultured fish at <110%, which is now the standard. The swim-up stage following hatch may be the most sensitive period to gas bubble trauma (GBT) as larvae may be present near the surface where hydrostatic pressure can not compensate for excess TGP (Hildebrand et al. 1999). Apart from direct mortality, there is a potential for sublethal effects as the presence of gas bubbles in the head results in positive buoyancy and behavioural changes, which may increase vulnerability to predation (Counihan et al. 1998).


In the Fraser River, movements related to spawning typically occur in the fall or in the spring to areas at, or near, suitable spawning habitat (RL&L 2000). Fall migrations to overwintering areas were sustained unidirectional movements, followed by a period of low activity. This period typically lasted from October to March. RL&L (2000) also noted a defined movement in spring to feeding areas throughout the Fraser River. Direction, distance and timing of this movement varied with food availability. In the upper Fraser River some individuals moved more than 30 km to suspected overwintering and feeding areas (Yarmish and Toth 2002). Kenney Dam is thought to be located at the upstream range of the historical distribution of sturgeon in the Nechako River and is not believed to be a migration barrier, although drastically lower flows in the section between Kenney Dam and the Nautley River (the outflow of Fraser Lake) may impact sturgeon use (S. McAdam pers comm.) and possibly food supply.

In the Columbia River, defined seasonal movements have not been identified; localized movements occur between adjacent high-use areas (Hildebrand et al. 1999). Movements to nearby shallower areas, associated with feeding, occur during the spring and summer. Limited movements downstream into the U.S. do occur, but have not been observed beyond the town of Kettle Falls, which is located just south of the border (Hildebrand et al. 1999; RL&L 1995). Similarly, a limited number of fish tagged in Lake Roosevelt have been recaptured upstream during sampling operations in Canada (B.C. Ministry of Water, Land and Air Protection, unpubl. data). Long distance movements, in the late fall to overwintering habitat, have been monitored in the Kootenay River system (Apperson and Anders 1991). In this system white sturgeon regularly move back-and-forth across the British Columbia-Idaho border, to take up positions in deep holes in the mainstem or in Kootenay Lake.

Dams have formed migration barriers in the Columbia River (Fig. 3), blocking movements of individuals between habitats necessary for various life history stages (UCRRP 2002), such as migrations to spawning or feeding habitats (Rochard et al. 1990, Beamesderfer and Farr 1997). Using modeling, Jager et al. (2001) hypothesized that habitat fragmentation could significantly increase the extinction risk of sturgeon populations compared to non-fragmented populations.

The potential for natural dispersal to re-establish extirpated populations or bolster declining populations isolated above or between dams is limited. Natural dispersal from the lower Fraser River upstream to the river system above the barrier at Hell’s Gate is unlikely. Tagging studies upstream of Hell’s Gate indicate that areas of rapids, fast canyon sections and long shallow glides probably inhibit movement between various sections of the river (RL&L 2000). Genetic information (Nelson et al. 1999, Pollard 2000) also supports the demographic separation of white sturgeon from these areas. Columbia River fish are known to use the marine environment extensively and a few tagged fish picked up in the lower Fraser River do not have numbers from the local tagging program (T. Nelson, LGL Ltd., pers. comm.). Lower Fraser River fish also make some use of the marine environment (Veinott et al. 1999). However, mitochondrial DNA comparisons of fish from the two rivers indicate that they are genetically different (Brown et al. 1992) and that straying or dispersal between systems is low. For instance, Brown et al. (1992) found heteroplasmy was higher in Fraser River fish (54%) than in Columbia River (25%) fish (c2 = 13.33, P <0.001). Restriction site diversities at the population level (Kc) were 0.687 for the Fraser and 0.362 for the Columbia. Mean diversity in mtDNA length variation within individuals (Kb) was significantly higher for the Fraser River (0.223 ±0.024) compared to the Columbia (0.127 ±0.27) population (Student’s t = 2.639; d.f. = 172, P <0.01).

Nutrition and Interspecific Interactions

White sturgeon eat a variety of organisms from benthic invertebrates like crayfish, shrimp and clams to fish such as lamprey, salmon, eulachon and smelt (Semakula and Larkin 1968; Lane and Rosenau 1995, Echols 1995). Smaller sturgeon tend to eat smaller invertebrates, while larger sturgeon consume mainly fish. White sturgeon will readily take live prey as well as carcasses (Lane and Rosenau 1995).

Prior to dam construction on the Columbia River, white sturgeon likely relied on runs of spawning salmon as an important seasonal food source (Hildebrand et al. 1999, UCRRP 2002). The elimination of this regular supply of marine derived nutrients has likely had an effect on ecosystem productivity as well. The loss of salmon runs may have implications for sturgeon spawning frequency and fecundity (Hildebrand et al. 1999). The construction of dams has also disrupted nutrient transport downstream further reducing productivity and altering the food web (Ashley et al. 1999, UCRRP 2002). The result is a decline in several species of native fishes and invertebrates affecting prey availability for all life-history stages of sturgeon (Duke et al. 1999). Limited food supply for juveniles has been predicted to increase foraging time to meet energetic requirements and result in increased vulnerability to predation (Korman and Walters 2001).

Food supply and, possibly, sturgeon reproductive capability, throughout the Fraser River may have been periodically affected by low salmon cycles. Productivity in the lower river has also been affected by the drastic decline in the spring eulachon run since 1994 (Eulachon Research Council 1998).

Humans are the only significant predator of sturgeon adults in riverine systems (UCRRP 2002), although various species of fish prey on sturgeon eggs and juveniles (Anders et al. 2001, UCRRP 2002). Following impoundment, predation may rise with increasing predator numbers due to more stable hydraulic, temperature and water quality conditions. Populations of native predators such as suckers (Catostomus sp.) and northern pikeminnow (Ptychocheilus oregonensis) increased in the Columbia River system following impoundment (UCRRP 2002). The illegal introduction of non-native walleye (Stizostedion vitreum) into Lake Roosevelt has also affected the Canadian portion of the Columbia River mainstem, as these fish make annual feeding migrations upstream during June to August (UCRRP 2002). Although this migration coincides with the downstream dispersal period for larval white sturgeon, it is not known whether walleye have an impact on larval abundance.


White sturgeon are specifically adapted to the large river systems of western Canada and the U.S. where they have evolved for millions of years. Their size and opportunistic behaviour have allowed them to take advantage of widely scattered seasonal resources (UCRRP 2002). Dam construction has blocked movement and restricted these fish to river fragments and reservoirs that in some cases no longer provide the full array of habitats or conditions necessary to complete their life cycle. Flow regulation has limited or changed cyclical hydrological and temperature patterns that are believed to have provided behavioral cues at appropriate spawning and rearing conditions (Beamesderfer and Farr 1997, RL&L 2000, Korman and Walters 2001, UCRRP 2002). The long lifespan of sturgeon also suggests that they have a limited ability to adapt to rapid environmental changes.