Ecological Archives E094175A3
Amanda J. Klemmer, John S. Richardson. 2013. Quantitative gradient of subsidies reveals a threshold in communitylevel trophic cascades. Ecology 94:1920–1926. http://dx.doi.org/10.1890/121444.1
Appendix C. Analysis of sticklebacks found in cages at the end of the experiment.
The effect of subsidies on stickleback abundance in the cages for predator (trout) and no predator (no trout) treatments was analyzed using a linear model that included predator presence/absence, detritus subsidy rate as a continuous variable, and the predator by subsidy interaction. There was no significant predator effect or predator by subsidy interaction, and therefore data were subsequently analyzed by linear regression for the effect of increasing subsidies on stickleback density. To meet all assumptions of normality and equal variance data were lntransformed. All statistical analyses were analyzed using R (R Development Core Team 2011).
Fig. C1. The effects of increasing subsidies on stickleback abundance per cage. Regression line is the effect of increasing subsidy on stickleback abundance across both predator treatments.
Table C1. Linear model results for stickleback abundance in cage experiment. P values that are bold are significant (P < 0.05).
Variable 
F1,26 
P 
Stickleback abundance* 


Trout presence 
2.17 
0.15 
Mass of Alder 
4.32 
0.04 
Trout x Mass 
0.67 
0.42 
* ln (x+1)transformed data 


Table C2. Regression equations for effects of subsidy on stickleback abundance in cage experiment. Bold P values are significant (P < 0.05).
Variable 
intercept 
slope 
F1,28 
P 
r2 
Stickleback abundance* 
1.57 
0.01 
4.20 
0.04 
0.13 
* ln (x +1)transformed data 





Literature cited
R Development Core Team. 2011. R: A language and environment for statistical computing R Foundation for Statistical Computing, Vienna, Austria.