Treering techniques for Pacific rockfish otoliths:

1
Tree-ring techniques for Pacific rockfish
otoliths age validation, chronology
development, and effects of ocean variability
Bryan Black1, George Boehlert1, and Mary
Yoklavich2 1. Hatfield Marine Science
Center 2. NMFS Southwest Fisheries Science
Center Oregon State University Santa Cruz
Laboratory Newport, Oregon Santa Cruz, CA
NOAA fisheries FATE program
2
The Concept otoliths are environmental
chronometers
3
The Concept otoliths are environmental
chronometers
Growth increment width reflects effects of 1.
Physiological and developmental status (age,
vigor, sex) 2. Physical environment (temperatur
e, upwelling) 3. Biological environment (compet
ition, food availability) FATE goals effects of
environment on population Important information
for stock assessment
4
Original analytical approach (Boehlert, Yoklavich
Chelton. 1989, Fish. Bull.)
Otolith sections cut, animals aged to estimate
birthyear Measurement of first 6 years of growth
back-calculated New analysis time series through
1994 additional sample depth
5
Target Species
100
canary rockfish growth increment sample depth
splitnose rockfish growth increment sample depth
80
60
40
20
0
1880
1900
1920
1940
1960
1980
2000
6
Chronology development
Calculate z- score for each growth increment -
within each of the 6 age groups GI
GI
z
std dev (GI)
Calculate average z for each calendar year
yields 6 chronologies
7
Anomaly time series
canary rockfish
splitnose rockfish
1
1
1
0
0
-1
-1
1
1
2
0
0
-1
-1
1
1
3
0
0
-1
-1
1
1
4
0
0
-1
-1
1
1
5
0
0
-1
-1
1
1
6
0
0
-1
-1
1920
1930
1940
1950
1960
1970
1980
1990
2000
1940
1950
1960
1970
1980
1990
2000
8
Anomaly time series
0.6
0.4
0.2
0
-0.2
-0.4
canary rockfish
-0.6
1920
1940
1960
1980
2000
0.6
0.4
0.2
0
-0.2
-0.4
splitnose rockfish
-0.6
1920
1940
1960
1980
2000
9
Anomaly time series
1.5
yr1
yr2
yr3
yr4
yr5
yr6
1
0.5
0
-0.5
canary rockfish
-1
-1.5
1920
1940
1960
1980
2000
1.5
yr1
yr2
yr3
yr4
yr5
yr6
1
0.5
0
-0.5
splitnose rockfish
-1
-1.5
1920
1940
1960
1980
2000
10
Interchronology correlations
Average correlation in 10-year moving
windows High frequency variability splitnose
1960 - 1994 canary 1970 - 1994
11
Effects of environment
Correlate master chronology with coastal
variables (38 to 50 N) Colbert
Schirripa 1. Sea surface temperature (1967
1994) 2. Upwelling index (1946 1994) 3. Sea
level pressure (1967 1994) basin-wide
variables 3. Pacific Decadal Oscillation (PDO)
(1900 1994) 4. Northern Oscillation Index
(NOI) (1948 - 1994)
12
Effects of environment
Correlate rockfish chronologies with seasonal
averages of environmental variables
winter Dec Feb spring Mar May summer June
Aug fall Sep Nov
13
canary
splitnose
SST
PDO
SLP
UW
NOI
0.6
0.6
1
0
0
splitnose rockfish
-0.6
-0.6
0.6
0.6
2
0
0
-0.6
-0.6
0.6
0.6
3
0
0
-0.6
-0.6
0.6
0.6
4
0
0
-0.6
-0.6
0.8
0.6
5
0
0
-0.8
-0.6
0.8
0.6
6
0
0
-0.8
-0.6
summer (L)
fall (L)
winter
spring
summer
fall
summer (L)
fall (L)
winter
spring
summer
fall
14
Applications of Dendrochronology
Assign correct calendar year Quantify effects
of climate on adults population-wide
15
Applications of Dendrochronology
Dendrochronology applies to longer time
series (at least 15 or 20 years) Must
have more than just 6 increments per otolith!
many trees hundreds of years old some
thousands ex. Englemann spruce
photo H.D. Grissino-Mayer, Ultimate tree-ring
web pages
16
Extend otolith time series
New Axis of Measurement
almost all growth increments measured
old axis (6 years)
17
New axis of measurement
18
Dendrochronology Procedures

• Step 1 AGE VALIDATION
• climate affects radial growth
• -induces synchronous growth patterns
• -growth patterns should match among samples
• -if not error likely
• method crossdating

19
Crossdating
direction of growth
Photo credit H.D. Grissino-Mayer, The Ultimate
Tree-Ring Web Pages
20
Crossdating
Synchronous growth in ponderosa pine N 10 trees
21
Crossdating
Synchronous growth in ponderosa pine example N
10 trees
22
Crossdating
Thats pretty cool, but will it apply to
rockfish???
23
Crossdating
Synchronous growth in splitnose rockfish
24
Crossdating
Statistically, how do we prove it? -correlate
all otoliths with one another -low correlation
potential errors but we only want the climate
signal otoliths growth increments also
reflect -age, vigor, sex, artifacts of
preparation, error
25
Crossdating
In trees, and probably also in fish high
frequency (year-to-year) variation -due almost
exclusively to climate -statistically valid for
correlations independent Isolate
climate (high frequency) signal and reduce
autocorrelation via DETRENDING
26
Detrending
Fit a spline curve to the measurements Divide
the measurements by the fitted function -reduces
long-term trends (autocorrelation) -stabilizes
variance -all otolith time series have a mean of
1 Thus, all otolith time series are equally
weighted, also statistically valid for
correlation
27
Detrending
0.08
0.07
0.06
0.05
ring width (mm)
0.04
0.03
0.02
0.01
0
1930
1940
1950
1960
1970
1980
1990
year
28
Detrending
ring width measurements, splines
29
Detrending
all splitnose residual chronologies N 15
30
Crossdating
Correlate each otolith with sample-wide averages
31
Crossdating
Step 1 select the first otolith time series
selected series
32
Crossdating
Step 2 average remaining (14) samples
33
Crossdating
Step 3 correlate single series with average of
others
selected series
r 0.62 p lt 0.001
average of remaining 14 series
34
Crossdating
If we overlooked 1983 (combined 1984 and 1983)
selected series with error
r -0.11
average of remaining 14 series
35
Crossdating
Step 4 repeat for each of the series
selected series
r 0.58
36
Crossdating
Step 4 repeat for each of the series
selected series
r 0.64
37
Crossdating
Step 4 repeat for each of the series
selected series
r 0.45
38
Crossdating
39
Crossdating
40
Crossdating
N 15 N 15
Step 5 check then remeasure, drop, or keep
problematic series typical eastern
forests hemlock 0.3 0.4 oaks 0.5 0.6
41
Chronology development
Average of 15 time series
42
Full data set 50 otoliths
Splitnose captured in 1989 and 1995 ISC
0.512 Average of all 50 time series
splitnose master chronology
43
Effects of environment splitnose
0.8
SST
PDO

0.6


SLP
UW
NOI

0.4


0.2
0
correlation coefficient
-0.2


-0.4
-0.6
if p lt 0.05

-0.8
summer (L)
fall (L)
winter
spring
summer
fall
44
Crossdating in canary rockfish
N 43 canary otoliths measured Chronology
timespan 1945-1994 interseries correlation r
0.412 (compare to 0.512 for splitnose)
45
Canary chronology
Canary rockfish ISC 0.412 Average of all 43
time series
46
Effects of environment canary
0.6
SST

PDO
SLP

0.4
UW
NOI
0.2
0
-0.2
-0.4
if p lt 0.05
-0.6
summer (L)
fall (L)
winter
summer
fall
spring
47
Chronology comparison
48
Effects of environment comparison
splitnose rockfish
0.8
0.6
0.4
0.2
0
-0.2
-0.4
SST
-0.6
PDO
-0.8
SLP
canary rockfish
correlation coefficient
UW
NOI
49
Conclusions
6-year approach reveals age - specific
differences Dendrochronology improved accuracy
- will become more widely used in fisheries -
establish effects of environment - puts
contemporary growth in context of past
growth BA Black, GW Boehlert, MM Yoklavich.
2005. Using tree-ring crossdating techniques to
validate age in long-lived fishes. Canadian
Journal of Fisheries and Aquatic Sciences. In
Press
50
Future Directions
More rigorously explore environmental
effects, and competition Yelloweye ideal for
exploring spatial variability - Gulf of Alaska
to California Current - trees stand specific
chronologies, what about fish?
51
Contact Information
Bryan Black Hatfield Marine Science
Center Newport, OR 97365 (541)
867-0283 Bryan.Black_at_oregonstate.edu