P.(Paul) Hendricks. # Surveys for grassland birds of the Malta Field Office-BLM, including a seven-year study in north Valley County (Volume 2008) online

. **(page 8 of 10)**

Online Library → P.(Paul) Hendricks → Surveys for grassland birds of the Malta Field Office-BLM, including a seven-year study in north Valley County (Volume 2008) → online text (page 8 of 10)

Font size

Figure C30. 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 y-axis 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.

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

I

0.0

Jackknife of Training gain for LarkBunting

"T 1 1 1 1 1 1 1 1 1 r

_] L

Without variable

With only variable

With all variables

0.2

0.4

o.e

0.8 1.0 1.2

Trainintj cjain

1.4

1.6 1.8

2.0

Figure C31. 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 - 38

Grasshopper Sparrow {Ammodramus savannarum)

TO

s

X

Figure C32. The hot-to-cold 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 Grasslioppei

'_Sp arrow

1.D

â–¡ .S

0.8

â–¡ 7

>

15

|0.5

Z'

/

/^

r

J

r

I

/

J

/

\

*^ â€”

r

/

0.4

â–¡.3

â–¡ .2

â–¡ .1

\

/

f^

r^

\

V ,

A

<>

y

>^

/

^

^

P^

â– â– â€”-

D.D

Fraction of background predicted

Omission on training samples

Omission on test samples

Predicted omission

10

20

30

70

en

90

100

40 50 60

CurTuilative tlireshold

Figure C33. 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 - 40

Sensitivity vs. Specificity for Grassliopper_Sparrow

I.D

â–¡ .9

0.3

d;0.7

U.d

E

o

^0.5

:eo.4

0.3

0.2

0.1

0.0

(^^^^" ^^^^^^ ^ .Jf^

Training data (AUC = 0.963)

Test data [AUG = 0.91 9)

Random Prediction (AUG = 0.5)

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

Specificity (1 - Fractionai Predicted Area)

0.9

1.0

Figure C34. Receiver Operating Characteristic (ROC) curve evaluating the overall predictive power of the model with the Area

Under the Curve (AUC). TheAUC 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 y-axis) 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 pseudo-absences are used instead of true negative loca-

tions) that were correctly classified by the model as negative. One minus the Specificity (plotted on the x-axis) is known as the

false positive rate and represents the commission error rate.

Appendix C - 41

Log response of Grasshopper_Sp arrow to aspect90m_mt83cljp

TO

s

n

â–¡ 20

i

1 â€ž,â€ž

1

ll

1

1

i

1 1

-'

i

t

"

aspects Dm,

Log response of Grasshopper

mtB3cli|i

Sparrow to n

cd90m

mtS3

1

1

III

1

ll

T

1

Log response

of Grasshopper Sparrow to sUtsgodOm mtS3

lip

|35

1

1

01 r

â–¡ 5

â–¡ â–¡

1 1

510 554 599 e43 I

Log response of Grassh

spper

Sparrow to

curve

planSOm mtS3

â–¡ ,5D

i|ll4[l

In...

."-

D05

â–¡ a:

Log response of Grasshopper Sp

arrow

to precip ann90m mtS3

Log

response

of Grasshopper Sparrow to tmaxSOm mt83

-â–¡,1

\

1'

|-a5

E-or,

-D7

\

\

V-

Log response of Grasshopper Sparrow to elevSOm mtS3cllp

L

^

-500 son

Log response of Grasshopper_Sparrowto slope90m_mtS3clip

\

Log response of Grasshopper Sparrow to tmin90m mt83cljp

'

\

\

\_

Lob

respon

se of Grass

hopper Sparrow to geo

nt

m

390mtS3c

P

â–

Log response

of Grassh

pper

Sparrow

to

soil

mp90

m mtS3

1

^

/

/

/

msi<9Qm_mt83

in90m_m1S3cli|i

Figure C35. 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 y-axis 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 GrassliopperSparrow

tmin90m_mt83clip

tma){90m_mt83

stats g 9 m_mtS 3_c I i p

soil_tmp90m_mt83

slope90m_mt83clip

piecip_ann90m_mt83

nlcd90m_mt83

geombmg90nnt83clip

elev90m_mt83clip

CLiiYe_plan90m_mt83

aspect9Qm_mta3clip

Without variable

With only variable

With all variables

0.0 0.2 0.4 0.6

0.8 1.0 1.2 1.4

Training cj;^in

1.6 1.S 2.0 2.2

Figure C36. 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 - 43

Baird's Sparrow (Ammodramus bairdii)

TO

s

X

Figure C37. The hot-to-cold 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 BairdsSp arrow

1.D

0.9

D.e

â–¡ 7

>

li

|0.5

â–¡ .3

â–¡ .2

â–¡ .1

/I

/

/

J

I

/

/

/

_i

/"

r

/

^

:^

/

A

r

I

/<

^

&^

V

A

tr^

A

r^

Xâ€”T

D.D

Fraction of background predicted

Omission entraining samples

Omission on test samples

Predicted omission

10 20 30 40 50 60 70

Cumulative tliresh old

SO

90

10Q

Figure C38. 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 - 45

Sens

tivity

vs. Sp

ecificityfor Baird

sSparrow

1.0

D.g

0.8

I0.7

Â£=

tn 0.6

E

^0.5

I0.3

0.2

0.1

0.0

â€” -

/

/

f

/

/

/

/

/

//

/^

/

y

/

Tmining data (AUC= 0.984)

Test data [AUG = 0.979)

Random Prediction tAUC = 0.5)

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

S|]ecifici1y (1 - Fractional Predicted Area)

0.9

1.0

Figure C39. Receiver Operating Characteristic (ROC) curve evaluating the overall predictive power of the model with the Area

Under the Curve (AUC). TheAUC 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 y-axis) 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 pseudo-absences are used instead of true negative loca-

tions) that were correctly classified by the model as negative. One minus the Specificity (plotted on the x-axis) is known as the

false positive rate and represents the commission error rate.

Appendix C - 46

TO

â– fcv

LO!

respon

rd_s_Sp

arrow to

aspects

1

clip

1

11

â– 1 1

fj

^

rJ

h

-\

ri

H\

1

1

1-

5spect9[]m_mta3cl[p

Log response of Balrd_s_SparrDwtD nlcd90m_mtS3

1

J

Log response

of Baird

s Sparrow

â–¡

statsgoSOm

Tit83

clip

1

30 E7 109 154 200 245 2B9 334 379 423 46B 513 557 ED2 647 E91

sta1sgo9aiTi_m183_cli|i

Lofl

response of Baird

s Sparrow

tocu

rve planSOm mtS3

1"

s Â°-Â°

1 [1.3

i"

S D,1

3-nn

-â–¡2

Log

response

ofBa

cuive_

rd s

lan90ni_mtB3

Sparrow to precl

annSOm mtS3

s

1

^ [1.05

B

1-ii.an

Â°-ll05

Log response of Baird s Sparrow to tmaKSOm mtS3

1

\

\

y

Loqre

ponse

>f Baird

s Sparrow to elevSOm

mtSSclip

/I

r

/

/

/

/

350[] iDDD

.og response of Bairc

s Sparrow to slopeSDm mtSScll

Log respo

slope9Dm_m103clip

nse of Baird s Sparrow to tmin90m mtS3ci

P

â–¡

-â–¡.5

\

s

^ 1

' â€” 1

1-30

1-3,5

^

'"

â– 5.0

-5 5

~~ ..^

" - -â– ^^

Log respon

se

of Baird

.s_Sparrowt

get

mbmg90mt83Glip

33 72 12[] â–

Lofl

esponse of Bs

ird

5 Sparrow to sc

II tmpSDm

mt83

1mj*90m_mt83

tmln90m_m183clip

Figure C40. 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 y-axis 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 BairdsSp arrow

tmin90m_mt83:lip

tmax90m_mtS3

stats g 9 m_mt8 3_c I i p

soil_tmp90m_mtS3

slope90m_mt83:lip

precip_ann90m_mtS3

nlcd90m_rrit83

geombmg90mt83:lip

elev90m_mt83:lip

cuiYe_plan90m_mtS3

aspect90m_mt83:lip

Q.Q

Without variable

With only variable

With all variables

0.5

1.0

1.S

Training gain

2.0

2.5

3.0

Figure C41. 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 - 48

McCown's Longspur {Calcarius mccownii)

TO

s

X

Figure C42. The hot-to-cold 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.

Om

ssion vs. Predicted Are

a for McCown_s_

Longs

pur

1.D

D.9

â–¡ .8

â–¡ 7

15

|0.5

(J

CIS

â–¡ .3

â–¡ .2

â–¡ .1

^

X

7

/

J

A

y

y

/j

-^

r^

7

/,

A

y

/,

fO

^

-i

^

i'

/

^^

^

r

K-

D.G

Fraction of background predicted

Omission on training sampies

Omission on test sampies

Predicted omission

10 20 30 40 50 60

Cumuiative tiiresiioii

70

SO

90

100

Figure C43. 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 - 50

Sensitivity vs

. Spec

if i city

for IVIcCown

_s_Lq

ngspur

1.D

D.9

O.S

I0.7

Â£=

SO.6

E

;go.4

I0.3

0.2

â–¡.1

a.u

[^

^

/

1

/

f

/

/

/

/

/

/

/

y

Training data (AUC = 0.993)

Test data (AUG =0.984)

Random Prediction (AUC = 0.5)

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

Specificity (1 - Fractionai Predicted Area)

0.9

1.0

Figure C44. 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 y-axis) 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 pseudo-absences are used instead of true negative loca-

ronmental variables are held constant at their average sample values. The value on the y-axis 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.

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

I

0.0

Jackknife of Training gain for LarkBunting

"T 1 1 1 1 1 1 1 1 1 r

_] L

Without variable

With only variable

With all variables

0.2

0.4

o.e

0.8 1.0 1.2

Trainintj cjain

1.4

1.6 1.8

2.0

Figure C31. 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 - 38

Grasshopper Sparrow {Ammodramus savannarum)

TO

s

X

Figure C32. The hot-to-cold 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 Grasslioppei

'_Sp arrow

1.D

â–¡ .S

0.8

â–¡ 7

>

15

|0.5

Z'

/

/^

r

J

r

I

/

J

/

\

*^ â€”

r

/

0.4

â–¡.3

â–¡ .2

â–¡ .1

\

/

f^

r^

\

V ,

A

<>

y

>^

/

^

^

P^

â– â– â€”-

D.D

Fraction of background predicted

Omission on training samples

Omission on test samples

Predicted omission

10

20

30

70

en

90

100

40 50 60

CurTuilative tlireshold

Figure C33. 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 - 40

Sensitivity vs. Specificity for Grassliopper_Sparrow

I.D

â–¡ .9

0.3

d;0.7

U.d

E

o

^0.5

:eo.4

0.3

0.2

0.1

0.0

(^^^^" ^^^^^^ ^ .Jf^

Training data (AUC = 0.963)

Test data [AUG = 0.91 9)

Random Prediction (AUG = 0.5)

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

Specificity (1 - Fractionai Predicted Area)

0.9

1.0

Figure C34. Receiver Operating Characteristic (ROC) curve evaluating the overall predictive power of the model with the Area

Under the Curve (AUC). TheAUC 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 y-axis) 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 pseudo-absences are used instead of true negative loca-

tions) that were correctly classified by the model as negative. One minus the Specificity (plotted on the x-axis) is known as the

false positive rate and represents the commission error rate.

Appendix C - 41

Log response of Grasshopper_Sp arrow to aspect90m_mt83cljp

TO

s

n

â–¡ 20

i

1 â€ž,â€ž

1

ll

1

1

i

1 1

-'

i

t

"

aspects Dm,

Log response of Grasshopper

mtB3cli|i

Sparrow to n

cd90m

mtS3

1

1

III

1

ll

T

1

Log response

of Grasshopper Sparrow to sUtsgodOm mtS3

lip

|35

1

1

01 r

â–¡ 5

â–¡ â–¡

1 1

510 554 599 e43 I

Log response of Grassh

spper

Sparrow to

curve

planSOm mtS3

â–¡ ,5D

i|ll4[l

In...

."-

D05

â–¡ a:

Log response of Grasshopper Sp

arrow

to precip ann90m mtS3

Log

response

of Grasshopper Sparrow to tmaxSOm mt83

-â–¡,1

\

1'

|-a5

E-or,

-D7

\

\

V-

Log response of Grasshopper Sparrow to elevSOm mtS3cllp

L

^

-500 son

Log response of Grasshopper_Sparrowto slope90m_mtS3clip

\

Log response of Grasshopper Sparrow to tmin90m mt83cljp

'

\

\

\_

Lob

respon

se of Grass

hopper Sparrow to geo

nt

m

390mtS3c

P

â–

Log response

of Grassh

pper

Sparrow

to

soil

mp90

m mtS3

1

^

/

/

/

msi<9Qm_mt83

in90m_m1S3cli|i

Figure C35. 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 y-axis 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 GrassliopperSparrow

tmin90m_mt83clip

tma){90m_mt83

stats g 9 m_mtS 3_c I i p

soil_tmp90m_mt83

slope90m_mt83clip

piecip_ann90m_mt83

nlcd90m_mt83

geombmg90nnt83clip

elev90m_mt83clip

CLiiYe_plan90m_mt83

aspect9Qm_mta3clip

Without variable

With only variable

With all variables

0.0 0.2 0.4 0.6

0.8 1.0 1.2 1.4

Training cj;^in

1.6 1.S 2.0 2.2

Figure C36. 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 - 43

Baird's Sparrow (Ammodramus bairdii)

TO

s

X

Figure C37. The hot-to-cold 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 BairdsSp arrow

1.D

0.9

D.e

â–¡ 7

>

li

|0.5

â–¡ .3

â–¡ .2

â–¡ .1

/I

/

/

J

I

/

/

/

_i

/"

r

/

^

:^

/

A

r

I

/<

^

&^

V

A

tr^

A

r^

Xâ€”T

D.D

Fraction of background predicted

Omission entraining samples

Omission on test samples

Predicted omission

10 20 30 40 50 60 70

Cumulative tliresh old

SO

90

10Q

Figure C38. 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 - 45

Sens

tivity

vs. Sp

ecificityfor Baird

sSparrow

1.0

D.g

0.8

I0.7

Â£=

tn 0.6

E

^0.5

I0.3

0.2

0.1

0.0

â€” -

/

/

f

/

/

/

/

/

//

/^

/

y

/

Tmining data (AUC= 0.984)

Test data [AUG = 0.979)

Random Prediction tAUC = 0.5)

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

S|]ecifici1y (1 - Fractional Predicted Area)

0.9

1.0

Figure C39. Receiver Operating Characteristic (ROC) curve evaluating the overall predictive power of the model with the Area

Under the Curve (AUC). TheAUC 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 y-axis) 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 pseudo-absences are used instead of true negative loca-

tions) that were correctly classified by the model as negative. One minus the Specificity (plotted on the x-axis) is known as the

false positive rate and represents the commission error rate.

Appendix C - 46

TO

â– fcv

LO!

respon

rd_s_Sp

arrow to

aspects

1

clip

1

11

â– 1 1

fj

^

rJ

h

-\

ri

H\

1

1

1-

5spect9[]m_mta3cl[p

Log response of Balrd_s_SparrDwtD nlcd90m_mtS3

1

J

Log response

of Baird

s Sparrow

â–¡

statsgoSOm

Tit83

clip

1

30 E7 109 154 200 245 2B9 334 379 423 46B 513 557 ED2 647 E91

sta1sgo9aiTi_m183_cli|i

Lofl

response of Baird

s Sparrow

tocu

rve planSOm mtS3

1"

s Â°-Â°

1 [1.3

i"

S D,1

3-nn

-â–¡2

Log

response

ofBa

cuive_

rd s

lan90ni_mtB3

Sparrow to precl

annSOm mtS3

s

1

^ [1.05

B

1-ii.an

Â°-ll05

Log response of Baird s Sparrow to tmaKSOm mtS3

1

\

\

y

Loqre

ponse

>f Baird

s Sparrow to elevSOm

mtSSclip

/I

r

/

/

/

/

350[] iDDD

.og response of Bairc

s Sparrow to slopeSDm mtSScll

Log respo

slope9Dm_m103clip

nse of Baird s Sparrow to tmin90m mtS3ci

P

â–¡

-â–¡.5

\

s

^ 1

' â€” 1

1-30

1-3,5

^

'"

â– 5.0

-5 5

~~ ..^

" - -â– ^^

Log respon

se

of Baird

.s_Sparrowt

get

mbmg90mt83Glip

33 72 12[] â–

Lofl

esponse of Bs

ird

5 Sparrow to sc

II tmpSDm

mt83

1mj*90m_mt83

tmln90m_m183clip

Figure C40. 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 y-axis 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 BairdsSp arrow

tmin90m_mt83:lip

tmax90m_mtS3

stats g 9 m_mt8 3_c I i p

soil_tmp90m_mtS3

slope90m_mt83:lip

precip_ann90m_mtS3

nlcd90m_rrit83

geombmg90mt83:lip

elev90m_mt83:lip

cuiYe_plan90m_mtS3

aspect90m_mt83:lip

Q.Q

Without variable

With only variable

With all variables

0.5

1.0

1.S

Training gain

2.0

2.5

3.0

Figure C41. 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 - 48

McCown's Longspur {Calcarius mccownii)

TO

s

X

Figure C42. The hot-to-cold 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.

Om

ssion vs. Predicted Are

a for McCown_s_

Longs

pur

1.D

D.9

â–¡ .8

â–¡ 7

15

|0.5

(J

CIS

â–¡ .3

â–¡ .2

â–¡ .1

^

X

7

/

J

A

y

y

/j

-^

r^

7

/,

A

y

/,

fO

^

-i

^

i'

/

^^

^

r

K-

D.G

Fraction of background predicted

Omission on training sampies

Omission on test sampies

Predicted omission

10 20 30 40 50 60

Cumuiative tiiresiioii

70

SO

90

100

Figure C43. 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 - 50

Sensitivity vs

. Spec

if i city

for IVIcCown

_s_Lq

ngspur

1.D

D.9

O.S

I0.7

Â£=

SO.6

E

;go.4

I0.3

0.2

â–¡.1

a.u

[^

^

/

1

/

f

/

/

/

/

/

/

/

y

Training data (AUC = 0.993)

Test data (AUG =0.984)

Random Prediction (AUC = 0.5)

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

Specificity (1 - Fractionai Predicted Area)

0.9

1.0

Figure C44. 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 y-axis) 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 pseudo-absences are used instead of true negative loca-