Table 1 Environmental variables are stronger predictors of Cliona orientalis distribution and abundance than biotic variables.

From: A decadal analysis of bioeroding sponge cover on the inshore Great Barrier Reef

Response

Predictors

R2

AIC

∆AIC

Category

n

Description

Presence-absence

Environmental

3

Chlorophyll a, fine sediment*, total carbon in sediment

0.29

39.5

3.4

  

1

Chlorophyll a

0.01

47.8

11.7

  

1

Total carbon in sediment

0.12

43.1

10.0

  

1

Fine sediment*

0.28

36.1

0

 

Biotic

4

Coral, macroalgae, sponge, and abiotic percent cover

0.11

44.2

7.9

 

Geography

1

Latitude

0

52.1

16.0

Percent cover

Environmental

3

Chlorophyll a, fine sediment*, total carbon in sediment

0.40

260.0

36.0

  

1

Fine sediment*

0.38

256.8

32.8

  

1

Total carbon in sediment*

0.19

266.6

42.6

  

1

Chlorophyll a

0

275.7

51.7

 

Biotic

4

Coral, macroalgae, sponge, and abiotic percent cover

0.10

224.0

0

 

Geography

1

Latitude

0.05

272.9

48.9

Change in percent cover

Environmental

3

Chlorophyll a, fine sediment, total carbon in sediment

0.33

−32.6

5.4

  

1

Chlorophyll a*

0.30

−35.9

2.1

  

1

Total carbon in sediment

0.12

−32.3

5.7

  

1

Fine sediment

0.03

−30.7

7.3

 

Biotic

4

Coral, macroalgae*, sponge, and abiotic percent cover

0.42

−33.0

5.0

  

1

Coral

0.01

−30.4

7.6

  

1

Macroalgae*

0.39

−38.0

0

  

1

Sponge

0.05

−31.0

7.0

  

1

Abiotic

0.11

−32.1

5.9

 

Geography

1

Latitude

0.10

−31.8

6.2

  1. The table contains a comparison of models with three categories of predictors, representing the hypotheses that the C. orientalis response was influenced by the percent cover of other taxa, environmental conditions, or latitude. The table includes the C. orientalis response variable; the category, number and description of predictors; the proportion of deviance explained by the predictors (R2); and the Aikaike Information criterion score (AIC). An * indicates predictors which were statistically significant and statistics are reported in figure legends. The most parsimonious model, in terms of R2 and AIC, is indicated in bold.