Appendix A. Tables showing (A) climatic factors influencing the INDVI in May, the maximum NDVI increase and the average slope of NDVI between early May and early July in three study sites in Alberta, (B) for the GPNP (Italy), climatic factors influencing the INDVI in May, the maximum NDVI increase, and the average slope in NDVI between early May and early July, and correlation coefficients between climatic variables, and (C) and (D) model selection procedures, and figures showing interannual variations in NDVI during the study periods in the four study sites and maximum increase in NDVI from 1982 to 2004 in four study sites in Alberta and Italy.
TABLE A1. Climatic factors influencing (a) the INDVI in May, (b) the maximum NDVI increase, and (c) the average slope of NDVI between early May and early July in three study sites in Alberta. Correlation coefficients (d) between variables in Alberta (* indicates significant correlations, P < 0.05). Legend: Temp. April: Average temperature in April; Temp. May: Average temperature in May; Prec. May: Precipitations in May.
(a) INDVI in May (R2 = 0.85)
Parameters |
L.S. means |
SE |
T |
P |
Intercept |
0.33 |
0.08 |
3.96 |
<0.001 |
(Ram-Caw Ridge) |
-0.38 |
0.04 |
-10.20 |
<0.001 |
(Sheep River-Caw Ridge) |
0.18 |
0.04 |
4.60 |
<0.001 |
Temp. April |
0.014 |
0.009 |
1.57 |
0.12 |
Log(Prec. May + 4) |
0.14 |
0.04 |
3.26 |
0.002 |
Res(Log(Snow April+4)) |
-0.11 |
0.06 |
-1.96 |
0.05 |
(b) Maximum Increase in NDVI (R2 = 0.49)
Parameters |
L.S. means |
SE |
T |
P |
Intercept |
0.08 |
0.03 |
2.54 |
0.01 |
(Ram-Caw Ridge) |
0.05 |
0.01 |
4.02 |
<0.001 |
(Sheep River-Caw Ridge) |
-0.05 |
0.01 |
-3.00 |
<0.001 |
Temp. April |
0.006 |
0.003 |
1.77 |
0.08 |
Log(Prec. May + 4) |
0.03 |
0.02 |
1.90 |
0.06 |
Res(Log(Snow April+4)) |
0.03 |
0.02 |
1.48 |
0.14 |
(c) Average slope between early May and early July (R2 = 0.30)
Parameters |
L.S. means |
SE |
T |
P |
Intercept |
0.10 |
0.02 |
6.60 |
<0.001 |
(Ram-Caw Ridge) |
-0.009 |
0.007 |
-1.26 |
0.21 |
(Sheep River-Caw Ridge) |
-0.03 |
0.007 |
-3.97 |
<0.001 |
Temp. April |
-0.0002 |
0.002 |
-0.11 |
0.91 |
Log(Prec. May + 4) |
-0.004 |
0.008 |
-0.46 |
0.65 |
Res(Log(Snow April+4)) |
0.02 |
0.01 |
1.53 |
0.13 |
(d) Correlation coefficients
|
Log(Prec. May + 4) |
Temp. April |
Temp. May |
Log(Snow April + 4) |
Log(Snow May + 4) |
Temp. April |
-0.1 |
1 |
|
|
|
Temp. May |
-0.39* |
0.62* |
1 |
|
|
Log(Snow April + 4) |
0.03 |
-0.33* |
-0.21 |
1 |
|
Log(Snow May + 4) |
0.46* |
-0.36* |
-0.59* |
0.10 |
1 |
TABLE A2. For the GPNP (Italy), climatic factors influencing (a) the INDVI in May, (b) the maximum NDVI increase, (c) the average slope in NDVI between early May and early July, and (d) correlation coefficients between climatic variables (* indicates significant correlations, P < 0.05). Legend: Temp. April: Average temperature in April; Temp. May: Average temperature in May; Prec. May: Precipitations in May.
(a) INDVI in May (R2 = 0.27)
Parameters |
L.S. means |
SE |
T |
P |
Intercept |
0.72 |
0.24 |
3.07 |
0.006 |
Temp. May |
0.008 |
0.006 |
1.26 |
0.22 |
Log(Snow in April+4) |
-0.16 |
0.09 |
-1.78 |
0.09 |
Log(Rain in May + 4) |
-0.06 |
0.05 |
-1.15 |
0.26 |
(b) Maximum Increase in NDVI (R2 = 0.22)
Parameters |
L.S. means |
SE |
T |
P |
Intercept |
0.19 |
0.09 |
2.06 |
0.05 |
Temp. May |
0.004 |
0.002 |
1.57 |
0.13 |
Log(Snow in April+4) |
-0.04 |
0.03 |
-1.19 |
0.25 |
Log(Rain in May + 4) |
0.009 |
0.02 |
0.47 |
0.64 |
(c) Average slope between early May and early July (R2 = 0.09)
Parameters |
L.S. means |
SE |
T |
P |
Intercept |
0.04 |
0.03 |
1.05 |
0.31 |
Temp. May |
0.001 |
0.001 |
1.28 |
0.21 |
Log(Snow in April+4) |
0.01 |
0.01 |
0.75 |
0.46 |
Log(Rain in May + 4) |
0.001 |
0.008 |
0.20 |
0.84 |
(d) Correlation coefficients
|
Log(Rain in May + 4) |
Temp. April |
Temp. May |
Log(Snow in April + 4) |
Log(Snow in May + 4) |
Temp. April |
-0.10 |
1 |
|
|
|
Temp. May |
-0.09 |
0.62* |
1 |
|
|
Log(Snow in April + 4) |
-0.14 |
-0.22 |
-0.20 |
1 |
|
Log(Snow in May + 4) |
0.32 |
-0.24 |
-0.26 |
0.71* |
1 |
TABLE A3. Model selection from linear models weighed for sample sizes comparing various measures of NDVI to the mean chest girth of bighorn lambs at Sheep River, and the body mass of bighorn lambs at Ram Mountain and mountain goat kids at Caw Ridge. For each model, the number of parameters (p), the percentage of variance explained (R2) and the AICc are provided. Legend: Res(BMY): residuals from the linear relationship of the maximum increase in NDVI and the average body mass of yearling females, indexing resource availability; Max. Inc.: Maximum increase in NDVI during green-up; Average slope: NDVI slope between early May and early July; St Max. Inc.: standardized annual maximum increase in NDVI.
(a) Ram Mountain
Models |
p |
R2 |
AICc |
Sex |
1 |
0.17 |
-63.27 |
Sex + Res(BMY) |
2 |
0.43 |
-77.82 |
Sex * Res(BMY) |
3 |
0.44 |
-76.3 |
Sex + Res(BMY) + Max. Inc. |
3 |
0.52 |
-83.62 |
Sex + Res(BMY) + INDVI in May |
3 |
0.47 |
-78.80 |
Sex + Res(BMY) + Average slope |
3 |
0.44 |
-76.22 |
Sex + Res(BMY) + Max. Inc. + Max. Inc.2 |
4 |
0.55 |
-83.60 |
Sex * Max. Inc. + Res(BMY) |
4 |
0.54 |
-82.51 |
Sex+ Res(BMY) * St Max. Inc. |
4 |
0.52 |
-81.36 |
(b) Sheep River
Models |
p |
R2 |
AICc |
Sex |
1 |
0.33 |
-191.00 |
Sex + Max. Inc. |
2 |
0.48 |
-200.25 |
Sex + INDVI in May |
2 |
0.33 |
-188.72 |
Sex + Average slope |
2 |
0.34 |
-189.24 |
Sex + Max. Inc. + Max. Inc.2 |
3 |
0.49 |
-199.09 |
Sex * Max. Inc. |
3 |
0.48 |
-197.91 |
(c) Caw Ridge
Models |
p |
R2 |
AICc |
Intercept |
0 |
- |
-27.80 |
Sex |
1 |
0.10 |
-28.17 |
Max. Inc. |
1 |
0.16 |
-29.67 |
Sex + Max. Inc. |
2 |
0.25 |
-30.05 |
Sex + INDVI in May |
2 |
0.11 |
-25.71 |
Sex + Average slope |
2 |
0.19 |
-28.22 |
Sex * Max. Inc. |
3 |
0.26 |
-27.49 |
Sex * INDVI in May |
3 |
0.16 |
-24.47 |
TABLE A4. Model selection for overwinter survival of juvenile ungulates at (a) Ram Mountain, (b) Sheep River, (c) Caw Ridge and (d) the GPNP. For each model, the number of parameters (p), the percentage of variance explained (R2) and the AICc are provided. Legend: Predation: dummy variable (0: absence of heavy predation or pneumonia, 1: heavy predation or pneumonia); Res(BMY): residuals from the linear relationship linking the maximum increase in NDVI and the average body mass of yearling females; Max. Inc.: maximum Increase in NDVI during green-up; Average slope: NDVI slope between early May and early June; St Max. Inc.: standardized maximum increase in NDVI.
(a) Ram Mountain
Models |
p |
R2 |
AICc |
Res(BMY) |
1 |
0.25 |
2.07 |
Res(BMY) + Predation |
2 |
0.25 |
4.52 |
Res(BMY) + INDVI in May |
2 |
0.30 |
3.13 |
Res(BMY) + Average slope |
2 |
0.25 |
4.67 |
Res(BMY) + Max. Inc. |
2 |
0.43 |
-1.55 |
Res(BMY) + Max. Inc. + Max. Inc. 2 |
3 |
0.43 |
1.26 |
Res(BMY) * St Max. Inc. |
3 |
0.44 |
0.92 |
Res(BMY) + Max. Inc.+ Predation |
3 |
0.44 |
0.99 |
(b) Sheep River
Models |
p |
R2 |
AICc |
Predation |
1 |
0.34 |
-1.56 |
Predation + INDVI in May |
2 |
0.35 |
0.84 |
Predation + Average slope |
2 |
0.37 |
0.08 |
Predation + Max. Inc. |
2 |
0.44 |
-2.81 |
Predation + Max. Inc. + Max. Inc. 2 |
3 |
0.45 |
-0.28 |
(c) Caw Ridge
Models |
p |
R2 |
AICc |
Intercept |
0 |
- |
-10.12 |
Max. Inc. |
1 |
0.002 |
-7.52 |
INDVI in May |
1 |
0.006 |
-7.59 |
Average slope |
1 |
0.05 |
-8.26 |
(d) Gran Paradiso National Park
Models |
p |
R2 |
AICc |
Ln(density) |
1 |
0.0008 |
-30.4 |
INDVI in May |
1 |
0.08 |
-32.45 |
Average slope |
1 |
0.02 |
-30.77 |
Max. Inc. |
1 |
0.33 |
-39.48 |
Max. Inc. + Max. Inc. 2 |
2 |
0.33 |
-37.12 |
Ln(density) + Max. Inc. |
2 |
0.33 |
-36.87 |
Stln(density) * St Max. Inc. |
3 |
0.33 |
-34.01 |
|
| FIG. A1. Inter-annual variations in NDVI during the study periods in the four study sites. |
 |
| FIG. A2. Maximum increase in NDVI from 1982 to 2004 in four study sites in Alberta and Italy. For the GPNP (Italy), the maximum increase in NDVI increased over the study period (r = 0.45, n = 23, P = 0.03). At Ram Mountain (Alberta), we found a similar but not significant trend (r = 0.36, n = 23, P = 0.09). There were no relationships between the maximal increase in NDVI and year at Sheep River and Caw Ridge (both in Alberta, P’s > 0.10). |