Response Of Greater Sagegrouse to nIcdSOm mt83
â–¡ .30
â–¡ .75
â–¡ .70
1
!â– I
i 1
â–¡ .65
â–¡ SO
OSS
50
46




â€”
â€”


40
â–
1
â–
3S
â– 1
II
â– 1
1
â–
83 B3 04 85 31 92
Re
sponse
of Greater Sagegrouse to
statsgoSOm rnt03
clip
, Jil

1 1 ' J[ 1 1
1
1
1

153 199 244
332 37B 422 466 510 555 599
Jo90m_mtB3_Cl[)
Respo
nse of Greater
Saqs
grouse to curve p
anSO
m mtS3
Response of Greater_Sagegrouse to elev90m_mt83cllp
Response
of Greater Sagegrouse to precip annSO
m mtS3
0.80
1^â€” ' 1
"
nj5
So.TO
la
â– gn.en
1.0.55
^050
40
1
lirecip
Response of Greater
_8nn90m_mt83
Sagegrouse to tmax90m mt83
1 â€”
_
^
\
/
'
500 500
Response of Greater_Sage grouse to slope90m_mtS3clJp
:
Response
of Greater
pe90in_mt83clip
Sag e grouse to tminSOm mtS3c
p
/
^
/
\
S
/
/
\
/
\
/
\
\
,
^ â€”
Response
ofG
re ate r
1.00
â–¡,95
â–¡ .9Q
â–¡,B5
o.isa
â–¡ .75
â–¡ .?â–¡
â–¡,65
â–¡.â‚¬â–¡
â–¡ .55
0.50
0.45
1

1
1
1
1
1
1
1
Respon
0.90 
â–¡.8S
iOHO
3 75
5 â–¡ 70
i 65
se of Greater
Sagegrouse to soil tmpSDm mt83
mm
ttttttt
I_tmp9am_mt83
tiTisx9am_m183
tmin9[)m_mta3clip
Figure C5. Response curves for individual environmental variables showing how the logistic prediction changes as each environmental variable is varied while all other envi
ronmental variables are held constant at their average sample values. The value on the yaxis is predicted probability of suitable conditions as given by the logistic formula P(x)
= exp(cl * fl(x) + c2 *f2(x) + c3 *f3(x)...) / Z. Note that if any of the environmental variables are correlated, the marginal response curves can be misleading (e.g., two highly
correlated variables with opposite response curves could effectively cancel each other out). Value definitions and/or links to metadata containing these definitions can be found in
the Descriptions of Environmental Input Layers section of the appendix above.
Jackknife of regularized training gain for GreaterSagegrous
aspect90m_mtS3clip
cuive_plan90m_mt83
elev90m_mtS3clip
geombmg90mt83elip
nlcd90m_mt83
piecip_ann90m_mt83
slope9Qm_mtS3clip
soil_tmp90m_mt83
statsgo90m_mt83_clip
tma5{90m_mt83
tmin90m_mtS3clip
i
Without variable
Witli only variable
With all variables
Q.Q Q.2 0.4 0.6 0.8 1.0
regularized training gain
1.2
1.4
Figure C6. Jackknife chart showing the relative importance of environmental variables as a function of the change in "gain "
(the log of the number of grid cells minus the average of the negative log probabilities of the sample locations) resulting from the
exclusion or sole inclusion of the environmental variable in the model Variables with the highest training gain resulting from
sole inclusion of those variables (dark blue bars) are the best individual variables at describing suitable habitat for the species.
Jariables with the greatest reduction in training gain resulting from their exclusion (light blue bars) contain information on the
species habitat use that is not present in other variables. The red bar indicates the maximum gain achieved with inclusion of all
variables.
Appendix C  13
Mountain Plover {Charadrius montanus)
TO
s
X
Figure C7. The hottocold color map indicates the suitability of each grid cell as a function of the environmental variables at that grid cell. Hotter colors indicate areas that are
predicted to have more suitable habitat for the species. Black dots are positive data used to build the model A shaded relief map, BLM Field Office boundaries, and county lines
are included for reference.
mission vs.
Predicted Area for Mountain,
Plover
1.D
<^
D.9
â–¡.e
â–¡ 7
15
0.5
(J
CIS
D.3
â–¡ .2
â–¡ .1
y
J
/
7/
/
A
y
/
^
/
/
i.
7
/
/
^
/^
^
y
I r
^
V
rJ
J^
^
D.G
Fraction of background predicted
Omission on training sampies
Omission on test sampies
Predicted omission
10
?0
30
70
SO
90
100
40 50 60
Cumuiative tiiresiioid
Figure C8. An evaluation of omission error rates for training (dark blue line) and test (light blue line) data as a function of the
cumulative threshold and overall predicted area. The red line indicates the overall fraction of the map area fitting each value of
the cumulative threshold. The black line is the predicted omission rate for each cumulative threshold.
Appendix C  15
Sens
tivity
vs. Sp
ecificity for MountainPlover
1.D
/ ,
/
â–¡ .9
0.8
I0.7
Â£Z
SO.6
E
;_0.5
;go.4
I0.3
â–¡ .2
D.l
â–¡ .â–¡
(r^
f
/
I
/

/
/
/
/
/
/
/
Tmining data [AUC = 0.994)
Test data (AUG =0.979)
Random Prediction [AUG = 0.5)
0.0
0.1
0.9
1.0
0.2 0.3 0.4 0.5 0.6 0.7 0.8
Specificity (1  Fractional Predicted Area)
Figure C9. Receiver Operating Characteristic (ROC) curve evaluating the overall predictive power of the model with the Area
Under the Curve (AUC). The AUC value indicates that when two random locations are chosen the model has that probability
of assigning a higher cumulative threshold value to the location with more suitable habitat. The light blue line indicates how a
neutral or random model would perform (i.e., it only has a 50% probability of assigning a higher cumulative threshold value to
a random location with more suitable habitat than a random location with less suitable habitat). The further toward the top left
of the graph the training (red) and test (blue) data lines are, the better the model is at predicting the presences contained in the
training data. Sensitivity (plotted on the yaxis) is the proportion of positive locations that were correctly classified by the model.
Sensitivity is also known as the true positive rate and can be thought of as the degree of absence of omission errors. Specificity is
the proportion of random locations chosen from the background (these pseudoabsences are used instead of true negative loca
tions) that were correctly classified by the model as negative. One minus the Specificity (plotted on the xaxis) is known as the
false positive rate and represents the commission error rate.
Appendix C  16
TO
s
n
Log response of Mountain P
over to
aspectSOm in1S3
lip
^
1 Â°'^
DD
S
IC14
1
1
1
1
. Il
1 Ml
ll+l
1
â–
I
f
"
' 1
1
I.D
'
Log response of Mountain
Plover to curve p
an90m mt83
1.B
I
In.e
â–¡ a
Log response of MDuntain_Plover to elev90m_mt83clip
aspect9am_
Log response of Mountain
mtB3cllp
Plover
to nIcdSOm
mt83
1,3
1
%
B
g 04
io2
â–¡,2
0.4
II
1
1
, 1
Log response
of Mountain Plover to
precip
ann90m mt83
M

f '"
s
i
â–¡.3
on
J
SOD n 50C
Log response of Mountaln_Plover to 5lope90m_mtS3cllp
Log response
of Mo
untain
Plo^
ip
2 '
r'
E"
3
05
0.0
1
^0,6
S'O,*
Loare
ponse
of Mountain
Plover
to soil
tmpSO
Ti mtsa
I
1
^
T
1
Logrespo
nse
of Mounts
n Plover to
stats go90
m
Tit83
clip
5.0
t
[35
ho
E
05
DO
,
j10 554 599 043 6B7
Log respon
se of Mountain Plover to tmax90m mtS3
â–¡ ,1S
2 0,14
1
,^012
Â§0 00
1
^0 04
0,03
0,00
Log response of Mountain Plover to tminSOm mISSclip
5.0
4.5
,Â°4
h
2,D
_____
~
"*
'
s
â–¡ ,5
â–¡ â–¡
tmax90m_m183
Figure CIO. Response curves for individual environmental variables showing how the logistic prediction changes as each environmental variable is varied while all other envi
ronmental variables are held constant at their average sample values. The value on the yaxis is predicted probability of suitable conditions as given by the logistic formula P(x)
= expfcl * fl(x) + c2 *f2(x) + c3 *f3(x)...) / Z. Note that if any of the environmental variables are correlated, the marginal response curves can be misleading (e.g., two highly
correlated variables with opposite response curves could effectively cancel each other out). Value definitions and/or links to metadata containing these definitions can be found in
the Descriptions of Environmental Input Layers section of the appendix above.
Jackknife of Training gain for MountainPlover
tmin9Dm_mte3clip
tmax90m_mt83
statsgo90m_mt83_clip
soil_tmp90m_mt83
slope9Dm_mt83clip
piecip_ann90m_mt83
nlcd90m_mt83
geombmg90mt83clip
elev9Dm_mt83clip
c u ive_p I a n 9 m_mt8 3
aspect90m_mt83clip
Without variable
With only variable
With all variables
0.5
1.0
2.5
3.0
3.5
1.5 2.0
Training cjain
Figure Cll. Jackknife chart showing the relative importance of environmental variables as a function of the change in "gain "
(the log of the number of grid cells minus the average of the negative log probabilities of the sample locations) resulting from the
exclusion or sole inclusion of the environmental variable in the model. Variables with the highest training gain resulting from
sole inclusion of those variables (dark blue bars) are the best individual variables at describing suitable habitat for the species.
Variables with the greatest reduction in training gain resulting from their exclusion (light blue bars) contain information on the
species habitat use that is not present in other variables. The red bar indicates the maximum gain achieved with inclusion of all
variables.
Appendix C  18
Longbilled Curlew {Numenius americanus)
TO
s
X
Figure CI 2. The hottocold color map indicates the suitability of each grid cell as a function of the environmental variables at that grid cell. Hotter colors indicate areas that
are predicted to have more suitable habitat for the species. Black dots are positive data used to build the model Gray dots are locations where a survey capable of detecting the
species has been performed. A shaded relief map, BLM Field Office boundaries, and county lines are included for reference.
Omission vs. Predicted Area for LongbilledCurlew
D.g
D.B
â–¡ .7
jÂ»
I 0.5
â– ^0.4
â–¡ .3
â–¡ .2
â–¡.1
â–¡ .â–¡
Fraction of background predicted
OrTiission on training samples
Omission on test samples
Predicted orTiission
10
30 40 50 60
Cumulative threslioli
70
SO
90
100
Figure C13. An evaluation of omission error rates for training (dark blue line) and test (light blue line) data as a function of the
cumulative threshold and overall predicted area. The red line indicates the overall fraction of the map area fitting each value of
the cumulative threshold. The black line is the predicted omission rate for each cumulative threshold.
Appendix C  20
i.a
D.9
o.s
â„¢ â„¢ .
q;0.7
Â£Z
O
E
o
â– 0.5
:eo.4
0.3
0.2
0.1
0.0
Sensitivity vs. Specificit
/ for Longbilled_Curlew
â€” 
/
, /
^
' "
/
h
f
/
1
/
I
/
\
/

/
/
/
y
y
Training data (AUC = 0.984)
Test data (AUG =0.934)
Random Prediction (AUG = 0.5)
0.0
0.1
0.:
0.9
1.0
0.3 0.4 0.5 0.6 0.7 0.8
G3ecifi:ity (1  Fractional Predicted Area)
Figure CI 4. Receiver Operating Characteristic (ROC) curve evaluating the overall predictive power of the model with the Area
Under the Curve (AUC). The AUC value indicates that when two random locations are chosen the model has that probability
of assigning a higher cumulative threshold value to the location with more suitable habitat. The light blue line indicates how a
neutral or random model would perform (i.e., it only has a 50% probability of assigning a higher cumulative threshold value to
a random location with more suitable habitat than a random location with less suitable habitat). The further toward the top left
of the graph the training (red) and test (blue) data lines are, the better the model is at predicting the presences contained in the
training data. Sensitivity (plotted on the yaxis) is the proportion of positive locations that were correctly classified by the model.
Sensitivity is also known as the true positive rate and can be thought of as the degree of absence of omission errors. Specificity is
the proportion of random locations chosen from the background (these pseudoabsences are used instead of true negative loca
tions) that were correctly classified by the model as negative. One minus the Specificity (plotted on the xaxis) is known as the
false positive rate and represents the commission error rate.
Appendix C  21
Log response of Longt>illed_Curlew to a5pect90m_mtS3clip
TO
s
n
^
aspect90m_mta3clip
Log response of LDngbJlled_CurlewtD nlcd90m_mtS3
JL _ ^^
t
Logrespo
3 31
seof
33 33 41
nif
Longbilled
42 43 51 T1 B1 B3 03 B4
9Om_ml03
Curlew to statsgoSOm mtS3
B5
Cl
91
P
2
6I]
1"
5:35
K3[]
s
1
i

'^ID
Â°
l,U
U..
1
â–¡.D
1 1
Log response of
Long
}illed
Curlew to curve
plan90m mt83
Log response of Longbilled
Curlew to
elev90m mtSSclip
s
U
K.
E
1
o
Â°
â–¡
Log resp
cuive_ilan9[)m_mtB
3nse of Longbilled Curlew to
precip ann90m
mt83
1
'
Lo
S response of Long
billed Curlew to tma)(90m mt83
1
\
\
\
\
\
\
\
\
^
\
:^ J
Log response of Longbilled_Curlew to 5lDpe90m_mtS3cllp
.ogrespon
si
se of Long
pe9[]m_m183clip
billed Curlew to tmin90m mtS3
lip
1
^
 \
N
\
fs
Log resp
onse of Long
billed Curlew to georr
b
mg90mtS3cll
4.5
4.:
E
s ^
1
^'"
â–¡ 5
1
D â–¡
Log response
Curlew to soil tmpSOm
miss
"^
LI
1"
3
ins
â–¡ 3
â–¡ .1
â–¡ â–¡
Figure CIS. Response curves for individual environmental variables showing how the logistic prediction changes as each environmental variable is varied while all other envi
ronmental variables are held constant at their average sample values. The value on the yaxis is predicted probability of suitable conditions as given by the logistic formula P(x)
= exp(cl * fl(x) + c2 *f2(x) + c3 *f3(x)...) / Z. Note that if any of the environmental variables are correlated, the marginal response curves can be misleading (e.g., two highly
correlated variables with opposite response curves could effectively cancel each other out). Value definitions and/or links to metadata containing these definitions can be found in
the Descriptions of Environmental Input Layers section of the appendix above.
Jackknife of Training gain for LongbilledCurlew
tmin9Dm_mte3clip
tmax90m_mt83
statsgo90m_mt83_clip
soil_tmp90m_mt83
slope9Dm_mt83clip
piecip_ann90m_mt83
nlcd90m_mt83
geombmg90mt83clip
elev9Dm_mt83clip
c u ive_p I a n 9 m_mt8 3
aspect90m_mt83clip
"C
0.0
Without variable
With only variable
With all variables
0.5
1.0
2.0
2.5
1.5
Training gain
Figure C16. Jackknife chart showing the relative importance of environmental variables as a function of the change in "gain"
(the log of the number of grid cells minus the average of the negative log probabilities of the sample locations) resulting from the
exclusion or sole inclusion of the environmental variable in the model Variables with the highest training gain resulting from
sole inclusion of those variables (dark blue bars) are the best individual variables at describing suitable habitat for the species.
Variables with the greatest reduction in training gain resulting from their exclusion (light blue bars) contain information on the
species habitat use that is not present in other variables. The red bar indicates the maximum gain achieved with inclusion of all
variables.
Appendix C  23
Sprague's Pipit (Anthus spragueii)
TO
s
X
Figure CI 7. The hottocold color map indicates the suitability of each grid cell as a function of the environmental variables at that grid cell. Hotter colors indicate areas that
are predicted to have more suitable habitat for the species. Black dots are positive data used to build the model Gray dots are locations where a survey capable of detecting the
species has been performed. A shaded relief map, BLM Field Office boundaries, and county lines are included for reference.
Omission vs
Predicted Area for 5prague_
sPipIt
1.D
D.g
D.8
0.7
Hi
I0.6
>
li
/j
A
f
/
//
r
/
J
y
/
>
â€¢f
0.i5
^0.4
0.3
D.2
â–¡ .1
/
J
r
/
r/.
cy
1
A
^
^
\
A
^
f^
^
/^