Ecological Archives E091-096-A5

Andrew Olaf Shelton. 2010. The origin of female-biased sex ratios in intertidal seagrasses (Phyllospadix spp.). Ecology 91:1380–1390.

Appendix E. Female Phyllospadix serrulatus spadix removal experiment.

Comparison of growth of plant within P. serrulatus plants in the spadix removal experiment.

I experimentally removed spadices from female P. serrulatus individuals to ask if removal of reproductive tissues affected plant growth rate. Twelve plants were identified within a 20 m2 area at SS and assigned haphazardly to removal or control treatments. The six removal plants had their reproductive shoots clipped with scissors in May 2006 and again in May 2007. Plant size, the length of a randomly selected rhizome within each plant, and number of spadices present were monitored twice each summer in 2006 and 2007 and in June 2008. Male plants were too rare and spatially dispersed to enable a parallel removal experiment. I calculated plant growth rates for four periods: summer 2006, winter 2006–2007, summer 2007, and winter 2007–2008. Due to irregular intervals between censuses, I standardized growth rates to a growth per day basis (cm2/day) and asked if growth rates were predicted by log(plant size), treatment, the four time periods, and all interactions as fixed effects and individual as a random effect. A parallel analysis on the effects of spadix removal on rhizome growth is presented in below.

Results

The experimental removal of spadices from female P. serrulatus plants produced no significant changes in female growth. While plant size (P = 0.008) and among time period differences in growth rate were observed (with a trend toward faster growth in summers and slower growth in the winter; P = 0.07; Fig. E1; Table E1), no effect of the spadix removal treatment was documented and all terms involved with the treatment term were dropped from the preferred model. Comparisons of rhizome growth produced similar results (see below).

TABLE E1. ANOVA table for effect of spadix removal on patch growth (AIC preferred model; see Fig. 4). Note that the treatment term has been dropped from the model entirely, indicating no effect of spadix removal on growth. The random effect associated with each individual was also dropped from the final model.

Source of variation

 

 

 

 

 

 

df

SSE

MSE

F

P

Log(patch size)

1

4.22

4.22

7.6

0.008

Interval

3

4.15

1.38

2.5

0.072

Error

43

23.79

0.55

 

 

 

FigE1 
 
   FIG. E1. Growth of P. serrulatus patches in the spadix removal experiment. Points indicate individual patches and horizontal lines designate treatment means for each time period. Growth did not differ significantly between treatments (see Appendix D for statistical tables).

 

Comparison of growth of individual rhizomes within P. serrulatus plants in the spadix removal experiment

In P. serrulatus individuals with experimentally removed spadices, I compared the growth of a single, haphazardly selected rhizome in each of the 12 individuals. After standardization to a daily growth rate (cm/day), the growth rate of each rhizome was predicted with a linear model using log(plant size), treatment (as a class variable; control or removal), the four time periods (as a class variable; summer 2006, winter 2006–2007, summer 2007, or winter 2007–2008), and all interactions as fixed effects and individual as a random effect. Due to rhizomes breaking and patches loosing tags, rhizome growth data was not available for all patches in all time intervals. A total of 38 growth observations were used for the 12 individuals.

Results demonstrate no differences in rhizome growth between control and spadix removal treatments. The AIC preferred model includes differences in growth among time intervals and no other effects (Table E2). Treatment, log(size), and all interactions were dropped from the model, indicating no effect of spadix removal or plant size on individual rhizome growth.

TABLE E2. ANOVA table for effect of spadix removal on rhizome growth (AIC preferred model). Note that the treatment term has been dropped from the model entirely, indicating no effect of spadix removal on growth. The random effect associated with each individual was also dropped from the final model.


Source of variation

 

 

 

 

 

 

df

SSE

MSE

F

P

Interval

3

0.0006

0.0002

3.24

0.036

Error

30

0.0019

0.00006

 

 


[Back to E091-096]