Impact of Climate Variation and Change Events on Threatened or Vulnerable Salmonids in British Columbia: Implications for Conservation and

Restoration Planning

Kim Hyatt, Margot Stockwell & Paul RankinFisheries & Oceans Canada, Salmon in Regional Ecosystems Program,

Pacific Biological Station, Nanaimo, B.C.

Feb 3, 2005: CIG Work Seattle

Sockeye Salmon Index Stocks on the Eastern Rim of the Pacific

• approximately 30% of all BC sockeye populations originate from watersheds near the southern end of the species range in North America

• these populations are likely to be highly sensitive to climate change

Subject Stocks and Study Area

Trends in Adult Sockeye Returns to the Columbia River(1890 - 1999)

• Okanagan sockeye are the last anadromous salmon stock of dozens that formerly returned to Canada through the Columbia River.

• The stock is currently depressed and fluctuates with a time weighted average to decline.

Study ThemeS

T R

E S

SClimate change will challenge the scope for life history adaptation by southern sockeye populations to variations in seasonal thermal regimes and associated changes to aquatic ecosystems.

• many elements of behaviour, physiology, & ecology of sockeye are controlled by temperature

• physiological performance is optimal at 15 C

• sockeye are increasingly stressed between 17 and 24 C

• 25 C is lethal

Results presented here first for Okanagan sockeye are part of a general approach to developing: (1) time series estimates of annual and seasonal changes in aquatic thermal regimes, (2) analysis of sockeye population responses to historic variations in climate, (3) inferences about global warming impacts on sockeye populations and (4) identification of potential options for effective stock management, conservation and/or restoration.

Changes to Aquatic Thermal Regimes in Southern Watersheds Pose Threats to Salmon at Several Life History Stages

General procedure developed on Okanagan sockeye is to:

(1) identify a biophysical model or “set of rules” for a given life history response and then (2) predict consequences of climate change

Analysis of sockeye responses to seasonal & annual variations in thermal regimes are now complete for life history stages involving:

(1) adult migration

(2) adult spawn timing

(3) egg-fry incubation

(4) juvenile rearing

• Sockeye migration through Zosel Dam on the Okanagan River is often “polymodal” ( )

• Sockeye migration through Wells Dam on the Columbia mainstem is generally “unimodal” ( )

Derivation of Temperature “Rules” for Adult Migration(Hyatt et al. 2003. Can. Wat. Res. J. 28: 689-713)

true water temp

spawners

0

5

10

15

20

25

30

1-J

an

1-F

eb

1-M

ar

1-A

pr

1-M

ay

1-J

un

1-J

ul

1-A

ug

1-S

ep

1-O

ct

1-N

ov

1-D

ec

wa

ter

tem

pe

ratu

re

0

2

4

6

8

10

ad

ult

sock

eye

as

% t

ota

l pa

ssa

ge

0

5

10

15

20

25

1-J

an

1-F

eb

1-M

ar

1-A

pr

1-M

ay

1-J

un

1-J

ul

1-A

ug

1-S

ep

1-O

ct

1-N

ov

1-D

ec

wa

ter

tem

pe

ratu

re

0

2

4

6

8

10

ad

ult

sock

eye

as

% t

ota

l pa

ssa

ge

0

5

10

15

20

25

30

1-J

an

1-F

eb

1-M

ar

1-A

pr

1-M

ay

1-J

un

1-J

ul

1-A

ug

1-S

ep

1-O

ct

1-N

ov

1-D

ec

wa

ter

tem

pe

ratu

re

0

2

4

6

8

10

12

14

ad

ult

sock

eye

as

% t

ota

l pa

ssa

ge

• Disruptions & delays to entry of Okanagan coincide with seasonal temperature variations. ( )

• Adult migrations stop at 21 C when temperature is increasing• Migration resumes at 21 - 22 C when seasonal temperatures decrease

Predicted versus observed delays for Okanagan sockeye adults based on temperature stop & start migration model

• Applying temperature-migration rules suggest delays may range from 0 - 54 days & an average of 30 days per annum. The magnitude of delays alternate with PDO cycles & have been increasing steadily in association with climate warming during 1985 - 2000 compared with lesser delays in the “cool” 1947-1985 interval.

Temperature-based Stop & Start “Rule” Predictions of Adult Sockeye Migration Delay

Adult sockeye migration delay based on seasonal changes in water temperature of the Okanagan River

Derivation of Temperature “Rules” for Adult Spawn Timing

0

5

10

15

20

25

1-Ja

n

1-F

eb

1-M

ar

1-A

pr

1-M

ay

1-Ju

n

1-Ju

l

1-A

ug

1-S

ep

1-O

ct

1-N

ov

1-D

ec

wat

er t

empe

ratu

re

0

10

20

30

40

50

spaw

ners

(as

% t

otal

)

1947

0

5

10

15

20

25

1-Ja

n

1-F

eb

1-M

ar

1-A

pr

1-M

ay

1-Ju

n

1-Ju

l

1-A

ug

1-S

ep

1-O

ct

1-N

ov

1-D

ec

wat

er t

empe

ratu

re

0

10

20

30

40

50

spaw

ners

(as

% t

otal

)

1951

0

5

10

15

20

25

1-Ja

n

1-F

eb

1-M

ar

1-A

pr

1-M

ay

1-Ju

n

1-Ju

l

1-A

ug

1-S

ep

1-O

ct

1-N

ov

1-D

ec

wat

er t

empe

ratu

re

0

10

20

30

40

50

spaw

ners

(as

% t

otal

)

1952

0

5

10

15

20

25

1-Ja

n

1-F

eb

1-M

ar

1-A

pr

1-M

ay

1-Ju

n

1-Ju

l

1-A

ug

1-S

ep

1-O

ct

1-N

ov

1-D

ec

wat

er t

empe

ratu

re

0

10

20

30

40

50

spaw

ners

(as

% t

otal

)

1957

0

5

10

15

20

25

1-Ja

n

1-F

eb

1-M

ar

1-A

pr

1-M

ay

1-Ju

n

1-Ju

l

1-A

ug

1-S

ep

1-O

ct

1-N

ov

1-D

ec

wat

er t

empe

ratu

re

0

10

20

30

40

50

spaw

ners

(as

% t

otal

)

1962

0

5

10

15

20

25

1-Ja

n

1-F

eb

1-M

ar

1-A

pr

1-M

ay

1-Ju

n

1-Ju

l

1-A

ug

1-S

ep

1-O

ct

1-N

ov

1-D

ec

wat

er t

empe

ratu

re

0

10

20

30

40

50

spaw

ners

(as

% t

otal

)

1971

0

5

10

15

20

25

1-Ja

n

1-F

eb

1-M

ar

1-A

pr

1-M

ay

1-Ju

n

1-Ju

l

1-A

ug

1-S

ep

1-O

ct

1-N

ov

1-D

ec

0

10

20

30

40

50

spaw

ners

(as

% t

otal

)

0

5

10

15

20

25

1-Ja

n

1-F

eb

1-M

ar

1-A

pr

1-M

ay

1-Ju

n

1-Ju

l

1-A

ug

1-S

ep

1-O

ct

1-N

ov

1-D

ec

0

10

20

30

40

50

spaw

ners

(as

% t

otal

)

0

5

10

15

20

25

1-Ja

n

1-F

eb

1-M

ar

1-A

pr

1-M

ay

1-Ju

n

1-Ju

l

1-A

ug

1-S

ep

1-O

ct

1-N

ov

1-D

ec

0

10

20

30

40

50

60

spaw

ners

(as

% t

otal

)

water temperature (actual)

sockeye spawningwater temperature (interpolated)

water temperature 12º C

• peak spawn timing occurs at 12 °C on average (range 8 - 15 °C)

Okanagan Adult Sockeye Average Peak Spawn Date

“Cool” Period1947 - 1985

“Warm” Period1985 - 2000

October 11th

(Oct. 3 - Oct. 19)n = 11

October 19th

(Oct. 10 - Oct. 24)n = 12

• peak spawning by adult sockeye has been delayed, on average, by 9 days during recent “warm” interval

ATU Model “Rules” Used to Estimate Hatch & Emergence Dates for Okanagan Sockeye Eggs and Fry

“Cool Period”

1947-1985

“Warm” Period

1985-2001

Peak Hatching Dates

Feb. 16Dec. 03 – Apr. 09

n = 10*

Mar. 03Feb. 05 – Mar. 26

n = 11*

On average, 100% hatch date is delayed by 15 days during warm intervals (i.e. 1985-2001) relative to cooler intervals (i.e. 1947-1985).

Eggs Hatching

Fry Recruitment

OKANAGANRIVER

SPAWNINGAREA

OSOYOOS LAKE

Derivation of Temperature “Rules” for Habitat Use by Juvenile Sockeye in Osoyoos Lake

November June

Winter – juvenile sockeye use most of the water column

Summer – sockeye concentrate near the thermocline

• Juvenile sockeye rear in Osoyoos Lake for 1 year.

• Acoustic and trawl surveys are used to follow seasonal changes in habitat use by juvenile sockeye rearing in Osoyoos Lake.

Day versus Night Distribution Changes Reveal “Rules” for Seasonal Habitat Utilization by Juvenile Sockeye in Lakes

• “17 degree rule”: night ascent limited by 17 C water

• “4 mg rule”: daytime descent limited by 4 mg/l O2 level

Seasonal Restriction of Vertical Distribution

Temperature and Oxygen extremes operate togetherto restrict the useable rearing volume of Osoyoos Lake.

South - unsuitable for rearing through summer & fall. Sockeye avoid this basin or die.Central - unsuitable for summer rearing although some sockeye may survive.North - suitable for rearing but still poses a challenge for sockeye in late summer and fall in some years.

North Central South

Seasonal & Annual Variations in Physical Conditions May Cause Mortality Events

5.9% loss per week = 95 % loss over 4 months

• temperature induced exclusions of juvenile sockeye from surface waters of Osoyoos Lake range from 40 - 120 days per year and average 78 days per year.

• the temperature-oxygen “squeeze” has likely been more severe during the current “warm” period (1985 - 2000) compared to the earlier “cool” period (1945 - 1985).

• these effects are especially clear in Osoyoos Lake but are similar in distant locations suggesting a regional climate “driver” associated with PDO cycles.

17 ºC “Rule” Predictions of Number of Days Juvenile Sockeye are Excluded from Surface Waters of Osoyoos (B.C. Interior)

and Great Central (Vancouver Island) Lakes (1924 - 1998)

(a) Osoyoos (a) Great Central

20

40

60

80

100

120

140

1920 1940 1960 1980 2000

Year

Day

s o

ver

17 C

elci

us

20

40

60

80

100

120

140

1920 1940 1960 1980 2000

Year

Day

s o

ver

17 C

elci

us

all year mean = 77 days all year mean = 91 days

warm cool warm warm cool warm

x = 85 days x = 68 days x = 78 days x = 100 days x = 82 days x = 94 days

Climate Change Impacts on Sockeye

To date, our studies on Okanagan sockeye suggest:

• high temperatures delay adult migration by 30 days on average (range 0-54) and average delays double in “warm” (1985-2002) vs “cool” (1947-1985) intervals. (Migration delays induce more pre-spawn mortality & reduce gamete viability).

• spawn timing during “warm” vs “cool” intervals is delayed by an average of 9 days while 100% egg hatch may be delayed by 15 days on average. (Hatching delays expose eggs to additional risk of loss from redd scour events during spring “floods”).

• summer temperature and O2 extremes control the quantity and quality of juvenile rearing habitat in Osoyoos Lake. (Differences in days of summer exclusion from surface waters in warm and cool intervals will influence feeding energetics and production levels of juvenile sockeye).

• the temp-O2 “squeezes” may be severe enough to induce major mortality events (e.g. 1998 event). (Rearing habitats have been reduced to zero in all basins of Osoyoos Lake at least twice in the past 30 years).

Implications for “Threatened” Sakinaw and Cultus Lake Sockeye Salmon

Small populations of sockeye originating from Sakinaw and Cultus lakes have exhibited more than a 90% decline within a few generations prompting an “emergency recovery” designation by COSEWIC and a petition to list them under SARA legislation in Canada.

Sakinaw Cultus

1

10

100

1,000

10,000

100,000

1940 1950 1960 1970 1980 1990 2000Year

Spa

wne

r A

bund

ance

1

10

100

1,000

10,000

100,000

1950 1960 1970 1980 1990 2000Year

Spa

wne

r A

bund

ance

Climate Change Events Will Affect Recovery Prospects for Depressed or Threatened Sockeye Stocks Such as Sakinaw and Cultus

Climate variation & change events are “non-stationary” & have a predictable influence on salmon life history events during specific periods (e.g. “warm” versus “cold” PDO intervals that each last 20-30 years) .

-5

5

15

25

35

45

55

65

1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Da

ys O

ve

r 2

1 C

elc

ius

-5

5

15

25

35

45

55

65

1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Da

ys O

ve

r 2

1 C

elc

ius

60

80

100

120

140

160

180

1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Da

ys O

ve

r 1

7 C

elc

ius

40

60

80

100

120

140

160

1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Da

ys O

ve

r 1

7 C

elc

ius

(b) “Exclusion” of Sakinaw fry (d) “Exclusion” of Cultus fry

warm cool warm

warm cool warm

warm cool warmcool

warm cool warmcool

Mean = 10 days

Mean = 122 days

x = 5

% ND = 45.5

x = 9

% ND = 29.0

x = 16

% ND = 7.7

x = 119 x = 118 x = 127

x = 19 x = 29 x = 23 x = 34

% ND = 12.0 % ND = 0.0 % ND = 9.7 % ND = 0.0

x = 93 x = 109 x = 118x = 118

Mean = 26 days

Mean = 110 days

(a) “Delay” of Sakinaw adults (b) “Delay” of Cultus adults

CVC Effects on Resident Salmonids in Georgia Basin Lakes

200 m

N

groundwater spring

CVC Effects on Resident Salmonids in Marion Lake, B.C.

Kokanee

Rainbow Trout

Marion L. is a small (29 ha), shallow (6m maximum depth), sub-alpine lake in the UBC Research Forest on the north shore of the Fraser River

Rainbow Trout Surface Feeding Activity and Temperature

Trout feeding activity at the lake surface falls off rapidly at temperatures <60C or >180C.

Salmonid Aggregations in a Spring-fed Thermal Refuge

Kokanee and large trout routinely aggregate in cool water refuges when lake surface temperatures exceed 210C

Temperatures at Depth in Marion Lake and the Spring

When seasonal temperatures exceed 21oC at the surface and 17oC at 4m in the lake they remain below 110C at depths > 2m in the spring

Climate Change Impacts on Frequency and Duration of Thermal Refuging by Marion Lake Salmonids

cool warm cool warm

x = 10.7

nr = 8 yr nr = 4 yr nr = 7 yr nr = 2 yr

nr = no refuge (# days above 21°C = 0)

adr = 7 days adr = 7.8 days adr = 10.1 days adr = 17 days

Interim Conclusions

• Our impact assessment model of sockeye life history responses to climate variations developed for Okanagan sockeye suggests southern stocks will be highly sensitive to future climate change.

• Our life history “models” developed on Okanagan sockeye may be usefully extended to quantify impacts on sockeye populations throughout the Pacific region (e.g. stocks in Sakinaw, Cultus and Marion lakes) .

• Climate variation & change effects on salmon populations are “non-stationary” (e.g. PDO cycles & longer term climate warming), partially predictable & should be assessed for their importance for effective salmon conservation & restoration planning.

• The latter will challenge institutional capacity and the existing socioeconomic order (e.g. Columbia R. salmon conservation & restoration experience).

CVC Events & Barkley Sd. Sockeye Production

Barkley Sound Sockeye Return Variations

- returns exhibit two multi-decadal, production “regimes” (i.e. low & high average production)

- lots of within decade variation to complicate harvest management

19601900 200019401920 1980

1800

1600

2000

1000

800

600

400

200

0

1400

1200Sockeye returns decline 2 yrs after strong ENSO events

Sock

eye

(tho

usan

ds)

Long term (1903 – 1986) trends in sockeye returns exhibit stanzas of below & above average production that are not related

in any meaningful way to parent stock & recruitment

40

20

0

80

100

120

60

- 20

- 40- 60- 80

- 120

10 20 30 40 50 60 70 80

Above Average

Below Average

% o

f M

ean

Trends in marine temperature (Year N) & Barkley Sd sockeye returns (Year N+2) suggest production is depressed or enhanced by warm & cool coastal ocean

conditions respectively reflecting both ENSO & PDO cycles

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0 5 10 15 20 25 30 35 40 45 50

SN

D’s

1935YEAR

1960 1984

Warm PDO

Warm PDOCool PDO

Year NYear N + 2

Partitioning of freshwater & marine survival within & among sockeye index stocks confirmed declines in early 80’s were localized to SWVI stocks and had

marine, not freshwater, origins.

Barkley Sound sockeye are especially sensitive because of theirtransitional zone location over which the subarctic domain has a varying influence (Fulton

and LeBrasseur 1985)

Strong El Niño- La Nina events induce major changes in continental shelf ecosystems.

Some of thesechanges alterecosystem capacityto support “robust”salmon populations.

Exotic Species Sightings off the BC Coast During 1983 (J. Fulton, P.B.S.)

(Brama japonica)

(Mola mola)

(Scomber japonicus)

(Sarda chiliensis)

(Pelicanus occidentalis)

Predator Abundance and Climate VariabilityIn the WCVI Study Area (Ware and McFarlane, 1995 )

0

100

200

300

400

500

600

11 12 13 14

Temperature

Bio

mas

s

Fig. 9. Relationship between average June-July sea surface temperature and swept-volume hake biomass for 13 survey years between 1968-91 (p = 0.003; Appendix 1).

Predator Abundance and Climate VariabilityIn the WCVI Study Area (Ware and McFarlane, 1995 )

0

100

200

300

400

500

600

11 12 13 14

Temperature

Bio

mas

s

Fig. 9. Relationship between average June-July sea surface temperature and swept-volume hake biomass for 13 survey years between 1968-91 (p = 0.003; Appendix 1).

Two-State Model of El Nino – La Nina Mediated Production Responses of Barkley Sd. Sockeye (Hyatt et al. 1989)

Slow growth by juvenile sockeye

Sockeye Fry Recruitment:increases

decreases

Low northward transport

Dominant piscivores coevolved with sockeye

Average to low mortality for juvenile sockeye

Average to above average adult returns

Zooplankton: high biomass and large size

Rapid growth by juvenile sockeye

High northward transport

Dominant piscivores “foreign” to sockeye

Above average mortality for juvenile sockeye

Below average adult returns

Zooplankton: low biomass, small size

Salmon have complex life histories that integrate events in freshwater & marine ecosystems. Simple associations don’t reveal freshwater or marine causation (i.e. mechanisms for regime changes in climate & salmon production identified in many recent studies remain untested for freshwater vs marine origins not to mention their specific causes)

Offshore

Marine

ZoneCoastal Marine Zone

Nearshore Freshwater Zone

Conclusions

• life history events & production variations of specific salmon populations do co-vary with short (ENSO) & longer term (PDO) climate regime changes.

• given complex life histories, generalizations about when, where and how salmon populations will respond to variations in climate regimes will be slow to emerge (e.g. conflicting inferences in existing studies).

• research into causal mechanisms will provide new information tools and models to improve fisheries management (e.g. Barkley Sd sockeye return forecasts, Okanagan sockeye migration delay prediction,

extinction prospects). Salmon are great indicators of many things but are not simple indicators of hardly anything !

Possible Adaptation Responses by Managers

• eliminate harvest to compensate for decreased stock productivity (most harvest already eliminated; requires renegotiation of Salmon Treaty provisions).

• alter water management rules to favour fish production benefits to compensate for decreased stock productivity (holding back scour inducing flows in Okanagan Lake may induce millions of $$ of flood losses in Kelowna; requires renegotiation of Canada-BC Okanagan Basin Water Agreement).

• engineer new water-control facilities to reduce water temp, adult migration and egg hatch delays (capital costs in excess of $100 million).

• allocate more stored water in summer-fall to control temp-O2 “squeeze” and compensate for mortality events in Osoyoos Lake (limited summer water worth millions of $$ to users e.g. wineries who won’t give up willingly).

• ignore stock declines and risk triggering SARA-COSEWIC listing in Canada and ESA listing in the U.S. (would trigger challenges in courts by water and fisheries resource stakeholders in both countries; could provoke constitutional challenge by First Nations groups in Canada).