Ecological Archives E096-221-A3

Karen A. Stahlheber, Kimberly L. Crispin, Cassidy Anton, and Carla M. D'Antonio. 2015. The ghosts of trees past: savanna trees create enduring legacies in plant species composition. Ecology 96:25102522. http://dx.doi.org/10.1890/14-2035.1

Appendix C. Additional results describing environmental attributes affected by the tree crown and relationships with plant community composition.

Table C1. Linear correlation of environmental vectors with Bray-Curtis dissimilarity in plant composition (2009 plant data). Axis 1 and 2 values indicate the endpoint of the vector with an origin of (0,0). See Fig. C1 for overlap with NMDS ordination plot.

Variable

Axis 1

Axis 2

r²

p

Bare ground (%)

0.9578

0.2874

0.1206

0.0042

Thatch cover (%)

0.1360

-0.9907

0.1604

0.0006

Litter cover (%)

-0.7334

0.6798

0.6407

0.0001

Dead wood cover (%)

-0.5759

0.8176

0.4319

0.0001

Gopher disturbance (%)

0.4866

-0.8737

0.1638

0.0010

Cowpie cover (%)

-0.6965

-0.7176

0.0354

0.2341

Thatch/Litter depth (cm)

-0.5411

-0.8409

0.1621

0.0008

PAR (% of ambient)

-0.8903

0.4554

0.3658

0.0001

Soil Moisture (%)

-0.5676

0.8233

0.5540

0.0001

Bulk density (g/cm3)

0.8960

-0.4440

0.2349

0.0001

Coarse Organic Matter (%)

-0.9601

0.2797

0.3871

0.0001

Rockiness (%)

-0.9153

-0.4029

0.0272

0.3386

pH

-0.9192

0.3938

0.0107

0.6504

Electroconductivity (μS/mL)

-0.9987

0.0501

0.4647

0.0001

Total carbon (%)

-0.6142

0.7892

0.4894

0.0001

Total nitrogen (%)

-0.7943

0.6076

0.4285

0.0001

K+ (µg/100g)

-0.9359

0.3522

0.5011

0.0001

Na+ (µg/100g)

0.0490

0.9988

0.1456

0.0018

Ca2+ (µg/100g)

-0.9905

-0.1373

0.2591

0.0001

Mg2+ (µg/100g)

0.1146

0.9934

0.2035

0.0001

Ca:Mg ratio

-0.1854

-0.9827

0.3053

0.0001

 

Table C2. Soil and light availability measurements that were significantly different around Q. agrifolia and Q. lobata when averaging all locations relative to the tree crown. Data were collected at the Mesa site. F tests and p values are from the species effect of fixed-factor linear models. For C:N Ratio, July pH and Mg, maximum likelihood tests of deviance were used in place of F tests due to non-normal model residuals.

 

Q. agrifolia

Q. lobata

F

p

% Rocks

36.31

33.13

5.77

0.019

% Carbon

3.07

2.70

5.23

0.025

C:N Ratio

9.56

9.02

9.30

0.003

% Soil Moisture

(May-July 2008)

5.49

4.25

19.20

< 0.001

% Soil Moisture

(March 2010)

10.49

12.21

4.81

0.038

pH (March 2010)

4.56

4.83

6.19

0.020

Mg2+ (meq/100g)

4.58

3.30

5.23

0.025

PAR, % of ambient (Apr 2009)

53.74

64.71

16.53

< 0.001

 

Table C3. Soil properties around living and dead trees at the Mesa site. The three locations were always significantly different, but the location effect rarely differed between living and dead trees. F statistics and p values are from the Tree Type * Location interaction term of a fixed-factor least squares linear model. Lower case letters indicate means that are significantly different according to a Tukey HSD post-hoc test.

 

 

Under

Edge

Open

F

p

% Coarse OM

Living

1.42a

0.79a

0.12b

10.73

< 0.002

Dead

0.93a

0.22b

0.21b

% Carbon

Living

3.66a

2.92ab

2.00b

1.13

0.33

Dead

3.40a

2.37b

2.17b

% Nitrogen

Living

0.38a

0.32ab

0.26b

2.01

0.15

Dead

0.31ab

0.26b

0.24b

% Soil Moisture

Living

6.79a

5.57a

3.33b

2.33

0.10

Dead

6.53a

5.15ab

4.94ab

Electroconductivity (μS/mL) (May–Jul)

Living

56.80a

38.00b

24.09c

1.02

0.36

Dead

64.87a

38.47b

33.30bc

Bulk Density (g/cm3)

Living

0.96ab

0.92b

1.06ab

1.46

0.24

Dead

0.94ab

1.06ab

1.12a

 

FigC1

Fig. C1. NMDS of Bray-Curtis dissimilarity matrix of survey plots, illustrating relationships between plant community composition (2009 data) and environmental variables at the Mesa site. Location types are shown by the different styles of points. Final stress of four-dimensional, convergent solution = 0.111 after 8 iterations. Blue vectors indicate environmental vectors with a significant linear relationship with NMDS scores of the plots generated by the envfit function in vegan (p > 0.05); the length of the arrow is proportional to the strength of this effect (r²). See Table C1 for details on vectors.


[Back to E096-221]