Ecological Archives A016-025-A3

John F. Lehmkuhl, Keith D. Kistler, James S. Begley, and John Boulanger. 2006. Demography of northern flying squirrels informs ecosystem management of western interior forests. Ecological Applications 16:584–600.

Appendix C. AIC model selection results for Huggins mark-recapture estimation of northern flying squirrel abundance in dry forests of eastern Washington. Models are listed in order of complexity. All models were estimated separately for each cover type; but, only the top 10 models based on AIC values were listed. Model rank, ΔAICc values, and AICc weights compare models within each cover type only. Models with substantial support from the data (ΔAICc values ≤2) are in bold font. See Appendices A and B for covariate definitions.

Modela

Open ponderosa pine

 

Young mixed conifer

 

Mature mixed conifer

Rankb

ΔAICcc

wid

Rank

ΔAICc

wi

Rank

ΔAICc

wi

Mo (stand*year)

       

8

15.2

0.000

       

Mt (snow max)

       

10

16.1

0.000

       

Mt (stand*year)

               

1

0.0

0.339

Mt (stand+snowmax)

10

15.4

0.000

               

Mt (year)

8

6.9

0.012

 

7

14.7

0.000

 

8

14.1

0.000

Mt (year+ccov)

               

10

14.3

0.000

Mt (year+ucov)

5

4.8

0.035

         

5

3.2

0.069

Mt (year+ucov+ccov)

               

3

1.5

0.163

Mth (year+wt)

9

8.9

0.005

 

1

0.0

0.498

       

Mth (year+wt+stand)

       

3

2.6

0.133

 

6

3.6

0.058

Mth (year+wt+ucov)

6

6.2

0.017

 

2

0.7

0.344

       

Mth (year+wt+ucov+ccov)

               

4

1.8

0.135

Mth2 (θ1(.) π1, π 2)+snowmax

4

4.0

0.053

 

4

8.0

0.009

       

Mth2 (θ1(.)π 1, π 2)+year

1

0.0

0.385

 

5

8.1

0.009

 

9

14.1

0.000

Mth2 (θ1(.)π 1, π 2)+year+ucov

2

0.3

0.330

               

Mth2 (θ1(stand) π 1, π 2)+year

3

1.9

0.147

 

6

8.7

0.006

 

2

1.3

0.181

Mth2 (θ1(stand) π 1, π 2 (stand))+year

7

6.4

0.016

 

9

15.6

0.000

       

Mth2 (θ1(stand) π 1(stand) π 2(stand))

               

7

3.7

0.054

a Model subscript definitions: Mo – no behavior, time, or heterogeneity effects on capture probability; Mt - capture probabilities vary with time; Mth - capture probabilities vary by time and individual; Mth2 –capture probabilities vary by time and individual animals in a bimodal distribution such that pi = probability the animal has capture probability (θ1).

b Rank among top 10 models for that cover type.

c The difference between the given model and the minimum AIC model for that cover type. Models with values ≤2 (bold) have substantial support from the data, values from 3-7 have considerably less support, and differences ≥10 have essentially no support (Burnham and Anderson 2002: p. 70).

d Akaike weight of evidence, or probability, that model is the actual “best” among models within the cover type.

LITERATURE CITED

Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: A practical information-theoretic approach, Second edition. Springer, New York, New York, USA.



[Back to A016-025]