Table 1. Variables used to model detection and occupancy of semi-aquatic furbearers in Lake County, Illinois, USA 2018-2020.
Probability of: Variable Class Source
Detection Observer: 1 Observer Observer
Observer: 2 Observer Observer
Bank height (m) Soil Observer
Bank width (m) Soil Observer
Bank angle (%) Soil Observer
Maximum temperature (°C) Temperature Weather Undergrounda
Average temperature (°C)Temperature Weather Undergrounda
Minimum temperature (°C) Temperature Weather Undergrounda
Last precipitation type Rainfall Weather Undergrounda
Rain
Snow
Last precipitation amount (cm)Rainfall Weather Undergrounda
Last precipitation date Rainfall Weather Undergrounda
2-week precipitation amount (cm) Rainfall Weather Undergrounda
2-week precipitation type Rainfall Weather Undergrounda
Rain
Snow
Mix
Survey replicate Day Observer
Occupancy Restoration practices Management GIS; Modified from LCFPD Restoration Practicesb
Ground cover (%) Local Observer
Herbaceous
Grass
Exposed roots
Shrubs
Woody debris and boulders
Bare ground
Shoreline
Number of saplings Local Observer
Dominant landcover Landscape GIS; Modified from Lake County Landcover Datab
Grassland
Urban
Forest
Cropland
Prairie
Shrubland
Wetland
Percent forest (%) Landscape GIS; Modified from Lake County Landcover Datab
Wetland type Landscape GIS Interpolationb
Channel width (m) Landscape Observer
Area of lake (acres) Landscape GIS Interpolationb
Stream density (km of streams/km2) Landscape GIS; % Streamsb
Aquatic vegetation (%) Local Observer
Emergent
Submerged
No plants
Developed area (%) Landscape GIS; Modified from Lake County Landcover Datab
Road density (km of roads/km2) Landscape GIS; % Roadb
Distance to urban area (km) Landscape GIS; Lake County Population Data >2,500b
a Weather Underground 2019
b ESRI 2011
Table 3. Results of final models for multi-season occupancy of semi-aquatic mammals in Lake County, Illinois, USA, 2018-2020.
CI does not overlap 0
Species Probability of Variable β SE 95% 80%
Muskrats Detection Survey replicate 0.568 0.093 *
Last precipitation: snow -0.670 0.286 *
Bank angle -0.252 0.115 *
Initial occupancy Null 2.180 0.526 *
Colonization Null 1.280 1.170
Extinction Null -2.530 0.620 *
Beavers Detection Survey replicate 0.326 0.110 *
Bank angle -0.375 0.156 *
2-week precipitation amount 0.364 0.160 *
Initial occupancy Stream density 1.893 0.664 *
Saplings 0.723 0.453 *
Colonization Null -1.070 0.537 *
Extinction Null -2.450 0.738 *
Mink Detection Survey replicate 0.374 0.109 *
Last precipitation amount -0.421 0.171 *
2-week precipitation amount -0.325 0.164 *
Initial occupancy Stream density 4.560 1.910 *
Herbaceous 1.100 1.160
Colonization Null -1.270 0.896 *
Extinction Null -3.390 3.840
River otters Detection Null -0.966 0.910
Initial occupancy Null -2.160 0.707 *
Table 4. Top competitive (ΔAIC ≤ 2) model results for probability of detection (p), initial occupancy (Ψ), colonization (γ), and extinction (ε) of muskrats in Lake County, Illinois, USA, 2018-2020. Models were ranked based on the lowest Akaike’s Information Criterion (qAIC) where ΔqAIC =qAICi – minimum qAIC, K= number of parameters, w=qAIC weight, and LL= log likelihood. The variables in the top model for each parameter were represented as (top) in the next modeling step.
Parameter Model K qAIC ΔqAIC w LL
Detection Ψ(.) γ(.) ε(.) p(survey replicate + last snow +
bank angle) 8 368.44 0.00 0.28 -174.16
Ψ(.) γ(.) ε(.) p(survey replicate + last snow) 7 368.81 0.37 0.23 -175.85
Ψ(.) γ(.) ε(.) p(survey replicate + bank angle) 7 369.30 0.86 0.18 -176.09
Ψ(.) γ(.) ε(.) p(survey replicate) 6 370.26 1.82 0.11 -177.99
Initial Occupancy Ψ(.) γ(.) ε(.) p(top) 8 368.44 0.00 0.22 -174.16
Colonization Ψ(top) γ(.) ε(.) p(top) 8 368.44 0.00 0.22 -174.16
Extinction Ψ(top) γ(top) ε(.) p(top) 8 368.44 0.00 0.22 -174.16
Table 5. Top competitive (ΔAIC ≤ 2) model results for probability of detection (p), initial occupancy (Ψ), colonization (γ), and extinction (ε) of beavers in Lake County, Illinois, USA, 2018-2020. Models were ranked based on the lowest Akaike’s Information Criterion (qAIC) where ΔqAIC =qAICi – minimum qAIC, K= number of parameters, w=qAIC weight, and LL= log likelihood. The variables in the top model for each parameter were represented as (top) in the next modeling step.
Parameter Model K qAIC ΔqAIC w LL
Detection Ψ(.) γ(.) ε(.) p(survey replicate) 6 161.45 0.00 0.14 -73.59
Ψ(.) γ(.) ε(.) p(survey replicate + bank angle) 7 161.97 0.52 0.11 -72.43
Ψ(.) γ(.) ε(.) p(survey replicate + 2-week amount) 7 162.20 0.75 0.10 -72.54
Initial Occupancy Ψ(stream density) γ(.) ε(.) p(top) 7 157.48 0.00 0.33 -70.18
Ψ(stream density + saplings) γ(.) ε(.) p(top) 8 159.40 1.92 0.13 -69.64
Colonization Ψ(top) γ(.) ε(.) p(top) 7 157.48 0.00 0.33 -70.18
Extinction Ψ(top) γ(top) ε(.) p(top) 7 157.48 0.00 0.33 -70.18
Table 6. Top competitive (ΔAIC ≤ 2) model results for probability of detection (p), initial occupancy (Ψ), colonization (γ), and extinction (ε) of mink in Lake County, Illinois, USA, 2018-2020. Models were ranked based on the lowest Akaike’s Information Criterion (qAIC) where ΔqAIC =qAICi – minimum qAIC, K= number of parameters, w=qAIC weight, and LL= log likelihood. The variables in the top model for each parameter were represented as (top) in the next modeling step.
Parameter Model K qAIC ΔqAIC w LL
Detection Ψ(.) γ(.) ε(.) p(survey replicate +
precipitation amount) 7 218.04 0.00 0.36 -100.47
Ψ(.) γ(.) ε(.) p(survey replicate) 6 219.07 1.02 0.21 -102.40
Ψ(.) γ(.) ε(.) p(survey replicate + 2-week amount) 7 219.50 1.46 0.17 -101.19
Initial Occupancy Ψ(stream density) γ(.) ε(.) p(top) 8 212.78 0.00 0.34 -96.33
Ψ(stream density + herbaceous) γ(.) ε(.) p(top) 9 214.55 1.77 0.14 -95.63
Colonization Ψ(top) γ(top) ε(.) p(top) 8 212.78 0.00 0.34 -96.33
Extinction Ψ(top) γ(top) ε(.) p(top) 8 212.78 0.00 0.34 -96.33
Table 7. Top competitive (ΔAIC ≤ 2) model results for probability of detection (p) and initial occupancy (Ψ) of river otters in Lake County, Illinois, USA, 2018-2020. Models were ranked based on the lowest Akaike’s Information Criterion (qAIC) where ΔqAIC =qAICi – minimum qAIC, K= number of parameters, w=qAIC weight, and LL= log likelihood. The variables in the top model for each parameter were represented as (top) in the next modeling step.
Parameter Model K qAIC ΔqAIC w LL
Detection Ψ(.) p(.) 5 42.83 0.00 0.26 -15.63
Occupancy Ψ(.) p(top) 5 42.83 0.00 0.37 -15.63
Colonization Ψ(top) γ(top) ε(.) p(top) 5 42.83 0.00 0.37 -15.63
Extinction Ψ(top) γ(top) ε(.) p(top) 5 42.83 0.00 0.37 -15.63