Ecological Archives E096-170-A2

Erin L. Kurten, S. Joseph Wright, Walter P. Carson. 2015. Hunting alters seedling functional trait composition in a Neotropical forest. Ecology 96:1633–1642. http://dx.doi.org/10.1890/14-1735.1

Appendix B. Tables of statistical results and figure of community mean wood density for tree species.

For details of statistical methods, see “Methods” in the main text. Tables and figures include tree and liana species in all analyses, unless otherwise noted. The “intercept” estimate parameter, where given, is the control treatment of the exclosure experiment and is the reference state for all other model parameters.

Table B1. The percentage of total individuals and species for which trait data was available for testing predictions.

 

Exclosure Trait Coverage

 

Hunting Comparison Trait Coverage

Trait

Individuals
(n = 12,075)

Species
(n = 249)

 

Individuals
(n = 20,231)

Species
(n = 279)

Seed dispersal agent

94.9%

88.0%

 

90.2%

91.4%

Dry seed mass

84.3%

61.8%

 

65.1%

59.9%

Leaf toughness

84.9%

67.9%

 

56.1%

65.6%

Wood density

90.1%

77.9%

 

79.8%

78.5%

Growth form

97.6%

74.8%

 

88.3%

87.8%

LMA

86.7%

67.9%

 

62.6%

66.3%

 

Table B2. GLM results for proportion of individuals with hunted dispersal agents.

Fixed effects

Estimate ± Std. Err.

z value

p value

Intercept (control)

-0.323 ± 0.031

-10.56

< 0.001

Vertebrate reduction

 0.256 ± 0.039

6.59

< 0.001

Driver

 0.580 ± 0.038

15.45

< 0.001

Vertebrate reduction × Driver

-1.194 ± 0.049

24.12

< 0.001

 

Table B3. GLM results for proportion of individuals with both hunted and non-hunted dispersal agents.

Fixed effects

Estimate ± Std. Err.

z value

p value

Intercept (control)

-1.57 ± 0.040

38.906

< 0.001

Vertebrate reduction

0.277 ± 0.050

5.591

< 0.001

Driver

0.209 ± 0.048

4.333

< 0.001

Vertebrate reduction × Driver

-0.503 ± 0.062

-8.042

< 0.001

 

Table B4. GLM results for proportion of individuals with non-hunted dispersal agents.

Fixed effects

Estimate ± Std. Err.

z value

p value

Intercept (control)

-0.370 ± 0.031

-11.94

< 0.001

Vertebrate reduction

-0.468 ± 0.041

-11.61

< 0.001

Driver

-0.825 ± 0.040

-20.64

< 0.001

Vertebrate reduction × Driver

1.64 ± 0.052

31.66

< 0.001

 

Table B5. GLM results for proportion of individuals with climbing growth forms (lianas).

Fixed effects

Estimate ± Std. Err.

z value

p value

Intercept (control)

-0.945 ± 0.035

-27.313

< 0.001

Vertebrate reduction

-0.344 ± 0.045

-7.608

< 0.001

Driver

-0.480 ± 0.044

-10.907

< 0.001

Vertebrate reduction × Driver

1.756 ± 0.057

30.983

< 0.001

 

Table B6. LM results for mean log(seed dry mass).

Fixed effects

Estimate ± Std. Err.

t value

p value

Intercept (control)

-2.486 ± 0.183

-13.603

< 0.001

Vertebrate reduction

 0.383 ± 0.258

1.481

0.148

Driver

 0.597 ± 0.251

2.377

0.023

Vertebrate reduction × Driver

-0.504 ± 0.348

-1.452

0.156

 

Table B7. LM results for mean wood density.

Fixed effects

Estimate ±Std. Err.

t value

p value

Intercept (control)

0.566 ± 0.008

73.559

< 0.001

Vertebrate reduction

0.007 ± 0.011

0.603

0.551

Driver

0.024 ± 0.011

2.317

0.027

Vertebrate reduction × Driver

-0.044 ± 0.015

-2.993

0.005

 

Table B8. LM results for mean LMA.

Fixed effects

Estimate ± Std. Err.

t value

p value

Intercept (control)

59.15 ± 1.71

34.632

< 0.001

Vertebrate reduction

 -0.95 ± 2.42

-0.394

0.696

Driver

  1.44 ± 2.35

0.612

0.545

Vertebrate reduction × Driver

  0.62 ± 3.25

0.190

0.851

 

Table B9. LM results for mean log(laminar toughness).

Fixed effects

Estimate ± Std. Err.

t value

p value

Intercept (control)

6.131 ± 0.032

191.2

< 0.001

Vertebrate reduction

 0.021 ± 0.045

0.473

0.640

Driver

 0.119 ± 0.044

2.706

0.011

Vertebrate reduction × Driver

-0.097 ± 0.061

-1.587

0.122

 

Table B10. Effects of vertebrate exclosure on functional traits and effects of hunting on functional traits, analyzed separately. Significant results are in bold text.

Response variable

Exclosure Experiment

Hunting Comparison

Proportion of individuals with

 

 

Hunted dispersal agents

P < 0.001

P < 0.001

Mixed dispersal agents

P < 0.001

P < 0.001

Non-hunted dispersal agents

P < 0.001

P < 0.001

Climbing growth forms

P < 0.001

P < 0.001

Community mean

 

 

Seed mass

P < 0.001

P = 0.685

Wood density

P = 0.308

P = 0.003

LMA

P = 0.622

P = 0.887

Leaf toughness

P = 0.448

P = 0.141

 

FigB1

Fig. B1. Responses of seedling community mean wood density to vertebrate reduction caused by (A) exclosures and (B) hunting, for tree species only. The heavy horizontal lines represent medians, boxes represent first and third quartiles, and whiskers represent extreme values. Lower-case letters distinguish groups that are significantly different from one another, where a two-way analysis of exclosure and hunting data together identified a significant interaction between main effects.


[Back to E096-170]