Ecological Archives E093-099-A3

Michael A. Sams and Michael J. Keough. 2012. Contrasting effects of variable species recruitment on marine sessile communities. Ecology 93:1153–1163.

Appendix C. Comparisons of community structure of experiments 1 and 2 Williamstown.

Methods

We ran an analysis to test whether community structure at Williamstown showed predictable community development through time, regardless of any differences in initial recruitment patterns. In order to do this we compared Williamstown experiments run at the same site but at different times using two factors: Experiment (fixed with two levels: Experiment 1 and Experiment 2) and Time (fixed with four levels: one month, two months, three months and six months). Principal co-ordinates (PCO) scaling plots using the Bray-Curtis dissimilarity matrix were done using both experiments to visualize differences in overall community structure between the different experiments through time. Again for clarity, PCO plots show a single object for each experiment at each time sampled rather than an object for every single plate. Each object represents the among group centroid based on a Bray-Curtis matrix of dissimilarities of the % cover taxa for each experiments at each time sampled. We also compared seasonal effects on community structure by running planned comparisons on the community structure of experiments 1 and 2 at Williamstown in January, February, and May 2009 when both experiments were sampled within at least one week of each other.

Results

Table C1. Results of PERMANOVA (a) comparing the overall community structure on settlement plates between Experiments 1 and 2 (Willamstown) at 1, 2, 3 and 6 months after experimental manipulations regardless of the time of year or (b) comparing Experiments 1 and 2 at times where both experiments were sampled within 1 week of each other. Numbers under factor headings are P values and numbers under MS(error) is the value of the mean square error.

(a) Experiment Time Treat × Time MS(error)
d.f. 1 4 4 320
Experiment 1 vs. 2 by age 0.000 0.000 0.000 653.6

(b) Experiment MS(error)
d.f. 1 64
January 0.000 728.13
February 0.000 697.91
April/May 0.000 806.22

Fig. C1. Principals co-ordinates (PCO) scaling plot of communities found colonizing settlement plates from Experiment 1 and 2 at Williamstown 2 weeks, 1 months, 2 months, 3 months and 6 months after experimental manipulations. Each object represents the among group centroid of a Bray-Curtis matrix of dissimilarities of the % cover of taxa for each experiment at each time sampled. For details of differences between treatments see pairwise comparisons for PERMANOVA in Table B1.


[Back to E093-099]