Ecological Archives E096172A1
Allert I. Bijleveld, Sönke Twietmeyer, Julia Piechocki, Jan A. van Gils, and Theunis Piersma. 2015. Natural selection by pulsed predation: survival of the thickest. Ecology 96:1653–1666. http://dx.doi.org/10.1890/141845.1
Appendix A. Figures showing the study area and the allometric relations between cockle length and body composition, and a table showing the mixedmodeling results for the effects of predation on a cockle’s relative body composition.
Fig. A1. Map of the study area. Our study area was situated near the islet of Griend in the Dutch Wadden Sea (53°14.615'N, 5°15.219'E). The three study plots of 100 × 100 m are outlined in red.
Fig. A2. Allometric relations between cockle length and body composition. Normally, allometric relationships are analyzed with linear regression on a loglog scale (dashed lines). Because of remaining nonlinearity in these allometric relationships, we fitted nonlinear local regression models (LOESS, solid lines). We obtained an individual’s relative body composition by extracting its residual from these regressions for (A) dry mass of the shell (DM_{shell}), or (B) ashfree dry mass of the flesh (AFDM_{flesh}) with length on a loglog scale.
Table A1. Mixedmodeling results for the effects of predation on a cockle’s relative body composition.

response variables 
random effect 
predictors 
est. 
SE 
P 
(A) 
relative DM_{shell} 

intercept 
0.01 
0.01 
0.23 



final sampling 
0.01 
0.01 
0.47 



predated 
0.02 
0.01 
0.18 



final sampling × predated 
0.07 
0.02 
<0.01 


sampling station 
0.03 
0.00 
<0.01 



residual 
0.07 
0.00 
<0.01 

(B) 
relative AFDM_{flesh} 

intercept 
0.05 
0.02 
0.01 



final sampling 
0.05 
0.02 
0.01 



predated 
0.10 
0.02 
<0.01 



final sampling × predated 
0.03 
0.03 
0.42 


sampling station 
0.06 
0.01 
<0.01 



residual 
0.07 
0.00 
<0.01 
Note: The mixedmodelling results for the effects of predation on an individual’s relative (A) dry mass of the shell (DM_{shell}), and (B) ashfree dry mass of the flesh (AFDM_{flesh}). The predictor ‘intercept’ refers to the mean values at the initial sampling occasion in the reference plot. The other predictors give the difference of the intercept with the final sampling in the reference plot (‘final sampling’), the initial sampling in the predated plots (‘predated’), and the final sampling in the predated plots (‘final sampling × predated’). The random effect estimates refer to standard deviations. Note that we limited these analyzes to cockles that red knots are able to ingest (lengths < 16 mm).