Skip to main content

Table 1 Comparison of linear models and piecewise models

From: Trophic structure of the pelagic food web in the East China Sea

Response

Predictor

AIC linear model

AIC piecewise model

Cut point

Larval fish

SSTa

452.1

440.5

25.3

 

SSS

460.6

462.1

32.2

 

Zooplankton biomass

446.6

447.0

−1.9

Zooplankton

SST

358.4

360.2

25.2

 

SSS

363.9

362.9

32.8

 

Phytoplankton biomass

354.2

358.0

8.9

Phytoplankton

SSTa

216.0

192.6

24.8

 

SSSa

216.6

204.7

32.0

 

DINa

211.4

198.9

10.4

 

PO4

219.4

223.4

0.2

 

SiO3

219.6

213.1

16.3

Zooplankton/phytoplankton ratio

SST

363.0

364.8

24.5

SSS

362.4

365.4

28.3

 

DIN

353.2

355.5

9.0

 

PO4

342.9

348.2

0.5

 

SiO3

353.8

356.8

4.7

 

Larval fish density

356.2

360.2

−1.3

  1. The relationships between the abundance/biomass of each trophic level and each predictor variable were analyzed using generalized linear mixed-effects models and mixed-effects piecewise regression models, and the AIC values of the two approaches were compared.
  2. aThe piecewise model had a much better fit compared to the linear model (ΔAIC ≥ 10), indicating a change of relationship at the cut point.