Ecological Archives E093-084-A3

G. Darrel Jenerette and Amitava Chatterjee. 2012. Soil metabolic pulses: water, substrate, and biological regulation. Ecology 93:959–966.

Appendix C (Table C1). Statistical models for control, water pulse, and substrate addition response presented in Figs. 1–3.

Relationship
(corresponding figure)
Model AIC β1
(95% CI)
β2
(95% CI)
RMSE R2
Summer control
vs. elevation (1A)*
y = β1exβ2 16.7 0.11
(0.06–0.16)
8.26 × 10-4
(6.72–9.80 × 10-4)
0.1611 0.69
Summer water pulse
vs. elevation (1C)
y = β1 + xβ2 272.8 3.65
(2.02–3.14)
-1.37 × 10-3
(-1.96–-0.78 × 10-3)
1.94 0.18
Summer substrate pulse
vs. elevation (1E)
y = β1exβ2 269.9 3.54
(2.18–4.91)
-3.27 × 10-4
(-6.08–0.45 × 10-4)
1.90 0.07
Winter water pulse
vs. elevation (1D)*
y = β1exβ2 32.0 0.19
(0.03–0.35)
6.04 × 10-4
(1.80–10.28 × 10-4)
0.39 0.18
Summer control
vs. temp. (2A)*
y = β1exβ2 25.0 1.03
(0.88–1.31
-0.04
(-0.05–-0.03)
0.21 0.46
Summer control
vs. water content
y = β1 + xβ2 29.8 0.05
  (-0.22–0.12)**
0.16
(0.12–0.21)
0.24 0.31
Summer water pulse
vs. control
y = β1exβ2 272.4 5.31
(3.29–7.34)
-3.44
(-5.11–-1.77)
1.75 0.30
Summer substrate pulse
vs. control
y = β1exβ2 33.6 0.35
(0.15–0.54)
0.60
(0.01–1.19)
0.40 0.12
Winter substrate pulse
vs. control
y = β1 + xβ2 94.2 1.29
(0.51–2.08)
1.29
(0.22–2.34)
1.10 0.14

*The residuals of these models exhibited significant heteroskedasticity (P < 0.05 Breush-Pagan test) and errors in models were estimated using robust linear or nonlinear methods.
**This intercept coefficient was not significant, but the overall model was (P < 0.05).


[Back to E093-084]