Skip to main content

Table 1 Univariate models of log(CORT) within G32, G38 and G42 stages

From: Context dependent variation in corticosterone and phenotypic divergence of Rana arvalis populations along an acidification gradient

Factors

Developmental stage

G32

G38

G42

df

F

p

df

F

p

df

F

p

Population

2

5.71

0.005

2

2.60

0.081

2

1.55

0.218

pH treatment

1

0.02

0.880

1

0.06

0.800

1

1.45

0.233

Block

1

15.15

 < 0.001

1

152.51

 < 0.001

1

2.10

0.151

Population × pH

2

0.34

0.712

2

1.36

0.262

2

0.52

0.596

pH × Block

1

4.84

0.031

–

–

–

–

–

–

Residual SE and (df)

0.46 (85)

  

0.22 (83)

  

0.32 (84)

  

Posthoc (Tukey)

LS mean ± SE

Population:

TT 1.21 ± 0.09 b

BS 1.63 ± 0.08 a

RD 1.72 ± 0.08 a

Block:

A 1.89 ± 0.07 a B 1.15 ± 0.07 b

pH x Block:

4A 1.83 ± 0.09 a 7A 1.95 ± 0.10 a 4B 1.30 ± 0.10 b 7B 1.00 ± 0.10 b

LS mean ± SE

Block:

A 2.09 ± 0.03 a B 1.50 ± 0.03 b

       
  1. Results of univariate linear models on log(CORT) for G32, G38 and G42, respectively. Results are presented for Rana arvalis tadpoles from the Acid (TT), Neutral (RD) and the Intermediate (BS) origin population, reared in Acid (4) or Neutral (7) pH treatment and two Blocks (A: morning/warmer, B: afternoon/colder). These are final models following removal of non-significant three- or two-way interactions. Statistically significant effects (p < 0.05) are shown in bold. Different letters in posthoc tests denote significantly different LS means (Tukey tests), indicating that 4A and 4B, and 7A and 7B, differ from each other (i.e. letters do not overlap), whereas there are no differences between the pH treatments within each block