1
2
Appendix A. Temporal sequence of stream temperature records from the Boise River basin used
to parameterize temperature models (n = 780).
Stream temperature records
200
160
120
80
40
0
3
4
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
5
6
Appendix B. Changes in riparian vegetation determined from Thematic Mapper satellite imagery
relative to wildfire perimeters within the Boise River basin between 1989 and 2002.
TM Classification
Fire
Cells with
Cells with
Perimeter 1989
2002
Cell count Cell changes
vegetation loss
vegetation gain
Inside
Open
Open
130636
Open
Shrub
9602
9602
9602 (60.2%)
Open
Tree
4227
4227
4227 (26.5%)
Shrub
Open
8173
8173
8173 (8.10%)
Shrub
Shrub
17628
Shrub
Tree
2116
2116
2116 (13.3%)
Tree
Open
71483
71483
71483 (70.6%)
Tree
Shrub
21558
21558
21558 (21.3%)
Tree
Tree
43417
Water
Water
2125
Total =
310965
117159 (37.7%)
101214 (32.6%) 15945 (5.13%)
Outside
7
Open
Open
Open
Shrub
Shrub
Shrub
Tree
Tree
Tree
Water
Open
Shrub
Tree
Open
Shrub
Tree
Open
Shrub
Tree
Water
Total =
278725
46585
55188
14708
102933
24300
42203
56176
363934
20742
1005494
46585
55188
14708
46585 (37.0%)
55188 (43.8%)
14708 (13.0%)
24300
42203
56176
42203 (37.3%)
56176 (49.7%)
239160 (23.8%)
113087 (11.3%)
24300 (19.3%)
126073 (12.5%)
8
9
10
11
12
Appendix C. An example of riparian vegetation classifications derived from Thematic Mapper
satellite imagery before a wildfire in 1989 (a) and after fire in 2002 (b). Classifications mapped
as vegetative gains and losses (c).
13
14
15
16
17
18
19
20
21
Appendix D. Relationships between radiation, watershed area, and vegetation class used to
predict radiation values for the stream network in the BRB. Vegetation classifications were
derived from Thematic Mapper imagery and radiation was measured at 181 field sites using
hemispherical photography. Because the vegetation classification did not account for variation in
vegetative height, density, or species composition within or among individual pixels,
considerable variation occurred in the power-law relationships we developed. Despite these
omissions, however, the approach did capture predictable distinctions between vegetation types
and proved adequate for describing dramatic changes in vegetative structure and radiation inputs
that occurred after fires (Appendix C).
y = 105x0.283
r2 = 0.64
2/year)
(µJ/m
Radiation
(Mj/m2/year)
Radiation
1000
800
y = 92.0x0.283
r2 = 0.25
600
y = 71.5x0.283
r2 = 0.31
Open
400
Shrub
Conifer
200
Open - observed
Shrub - observed
Conifer - observed
0
0
22
23
1000
2000
3000
Watershed contributing area (ha)
4000
Appendix E. Stream temperature thresholds used to delineate habitat quality for bull trout (left
panels) and rainbow trout (right panels). Thresholds were based on observed densities of bull
trout < 150 mm and rainbow trout collected during electrofishing surveys of 249 sites on 20
streams in or near the Boise River basin in 2007. Temperatures in several of the warmest sites
where bull trout occurred were affected by fires after surveys were complete. Rainbow trout
probably occurred in streams warmer than those we sampled, as Dunham et al. (2007) observed
rainbow trout in Boise River basin sites with MWMTs exceeding 25.5°C and there are several
published accounts of rainbow trout in streams as warm as 27°C - 28°C (McCullough et al.
2001).
16
High
Quality
12
8
4
0
15
20
Summer MWMT (C)
High
Quality
15
10
5
16
15
20
25
Summer MWMT (C)
25
Suitable
Suitable
2
m)
(#/100 m2)
trout(#/100
Rainbowtrout
Rainbow
High
Quality
12
10
25
8
4
20
High
Quality
15
10
5
0
0
7
33
34
Suitable
20
0
10
2 m 2)
Juvenile
(#/100
trout(#/100
bulltrout
Juvenilebull
) m
25
Suitable
2)
m2)
Rainbow
Rainbow trout (#/100 m
2
(#/100
trout(#/100
bulltrout
Juvenilebull
m2m
Juvenile
) )
24
25
26
27
28
29
30
31
32
9
11
Summer Mean (C)
13
15
7
9
11
Summer mean (C)
13
15
Appendix F. Correlations among variables at 780 sites used in stream temperature models for the Boise River basin.
Air
Air
Stream Stream
C_A
D_D
Ele
G_V
SL
V_B
Rad
MWMT mean Flow mean MWMT
C_A
1.00
D_D
0.14
1.00
Ele
-0.32
-0.54
1.00
G_V
-0.08
-0.23
0.52
1.00
SL
-0.15
-0.30
0.26
0.13
1.00
V_B
0.24
0.09
0.00
0.24
-0.44
1.00
Rad
0.33
0.11
-0.30
0.01
-0.40
0.34
1.00
Air MWMT
-0.02
0.01
0.00
-0.09
-0.08
-0.09
-0.13
1.00
Air mean
0.02
0.11
-0.12
-0.09
-0.11
0.02
0.09
0.66
1.00
Flow
-0.05
-0.01
0.02
0.21
0.09
0.02
0.18
-0.63
0.02
1.00
Stream mean
0.41
0.33
-0.72
-0.45
-0.25
0.11
0.46
0.12
0.25
-0.16
1.00
Stream MWMT
0.29
0.30
-0.60
-0.39
-0.29
0.17
0.46
0.14
0.19
-0.18
0.93
1.00
1
2
3
4
5
6
7
Appendix G. Semivariograms of the residuals from the final MWMT (left panels) and summer
mean (right panels) spatial stream temperature models, which included autocovariance structures
based on flow-connected, flow-unconnected, and Euclidean relationships. Semivariograms
quantify the average variability between pairwise combinations of model residuals for a series of
spatial lags and plot this variability as a function of the intervening distance. Semivariograms of
the temperature models suggested strong spatial trends in residuals based on Euclidean and flowconnected distances; with weaker trends among flow-unconnected sites.
Flow-connected
Flow-connected
Flow-connected
Flow-connected
20
20
20
30
30
30
40
40
40
20Distance
Distance (km)
(km)30
Distance (km)
40
20
20
30
30
20
30
20
Distance
Distance (km)
(km)30
Distance (km)
0
0
0
50
50
50
50
0
0
000
10
10
10
10
20
20
20
20
30
30
30
30
Distance
Distance (km)
(km)
Distance (km)
10
10
10
10
20
20
20
40
40
40
40
50
50
50
50
30
30
30
(km)
20Distance
30
Distance
Distance (km)
(km)
40
40
40
40
50
50
50
50
Flow-unconnected
Flow-unconnected
Flow-unconnected
10
10
10
10
20
20
20
30
30
30
(km)
20Distance
30
Distance (km)
Distance (km)
Flow-unconnected distance (km)
Gamma
Gamma
Gamma
Gamma
40
40
40
40
0 002 224 446 668 881010
10
Euclidean
Euclidean
Euclidean
Gamma
Gamma
Gamma
0
Gamma
Gamma
Gamma
10
10
10
10
Gamma
Gamma
Gamma
Gamma
Gamma
0 002 224 446 668 881010
10
00
8
9
0
0
0
50
50
50
50
Flow-connected distance (km)
Flow-unconnected
Flow-unconnected
Flow-unconnected
00
00 0 22 2 44 4 66 6 88 8
Gamma
Gamma
Gamma
10
10
10
10
00 0 22 2 44 4 66 6 88 8
00
0
0
Mean
40
40
40
40
50
50
50
50
Euclidean
Euclidean
Euclidean
00 0 22 2 44 4 66 6 88 8
Gamma
Gamma
Gamma
Gamma
0 002 224 446 668 881010
10
Flow-connected
Flow-connected
MWMT
0
000
Euclidean distance (km)
10
10
10
10
20
30
20
30
20
30
20
30
Distance (km)
Distance
Distance (km)
(km)
40
40
40
40
50
50
50
50
10
11
12
13
Appendix H. Percentage of the residual error structures in the final spatial stream temperature
models attributable to tail-up, tail-down, Euclidean, and nugget portions of the covariance
structure. The tail-up portion of the covariance structure explained the greatest residual variation,
which is generally expected for stream attributes with passive flow characteristics.
100%
Tail-up
Tail-down
80%
Euclidean
Nugget
60%
40%
20%
0%
14
15
MWMT
Mean