# fdschneider/Schneider_et_al_2012_Corrigendum

Corrigendum to Schneider, Scheu and Brose 2012 Ecology Letters
Switch branches/tags
Nothing to show
Fetching latest commit…
Cannot retrieve the latest commit at this time.
 Failed to load latest commit information. code data documentation manuscript .gitignore README.md

# Corrigendum to Schneider, Scheu & Brose (2012)

## "Body mass constraints on feeding rates determine the consequences of predator loss"

original paper published in Ecology Letters, 15:436-443, Mai 2012 doi: 10.1111/j.1461-0248.2012.01750.x

#### Florian Dirk Schneider1*, Stefan Scheu2 , and Ulrich Brose2

1 Institut des Sciences de l’Evolution, CNRS, Université Montpellier 2 - CC065, Montpellier Cedex 05, 34095, France

2 Georg August University Göttingen, J.F. Blumenbach Institute of Zoology and Anthropology, Berliner Str. 28, 37073 Göttingen, Germany

* Correspondence E-mail: florian.schneider@univ-montp2.fr

In the simulations for our paper "Body mass constraints on feeding rates determine the consequences of predator loss" in Ecology Letters we discovered a technical flaw which led to wrong predictions of interaction strengths. We apologize for this confusion. Here, we provide a corrected set of parameters which produces similar results with regard to the major interpretations of the paper. Additionally, we correct an erroneous count of springtails in one replicate of the experiment, resulting in higher experimental interaction strength of spiders on springtails. However, after these corrections, all interpretations of the paper remain valid.

As a consequence of these mistakes, we commit ourselves to transparent and reproducible computational research by providing raw data and the full simulation code as an open access repository on GitHub (https://github.com/fdschneider/Schneider_et_al_2012_Corrigendum).

### Parameter selection

All equations and allometric exponents given in the paper remain valid. We refined few parameters to reproduce interaction strengths matching the experiment similarly well as in the original publication: The parameters of the Ricker function (eq 2 in the paper, success rate) $R_\mathrm{opt} = 200$, $γ = 1$; the normalisation constants of attack rate $a_0 = 0.15$ and handling time $h_0 = 8.0$. All parameters are still in the range of previously documented values (references in original paper). Both, the original and corrected simulations were initialised with densities $N_1 = 912$, $N_2 = 350$, $N_3 = 4$, $N_4 = 2$ and body masses $M_1 = 0.0104$, $M_2 = 0.157$, $M_3 = 25.69$ and $M_4 = 125.3$ [mg], for springtails, predatory mites, spiders and centipedes, respectively (the body masses reported in the paper were rounded to one digit less). In the simulations we did not assume a delayed restocking of mites as it was the case in the experiment. Simulations ran over a time of 1152 hours.

### Results

The statical estimates of feeding rates reported in the paper are correct. The experimental and simulated $IS_{ij}$ were positively correlated at the population level ($r = 0.91$; $P < 0.001$; Fig. 4b), whereas the positive correlation at the per capita level was only marginally significant ($r = 0.49$; $P = 0.093$).

Fig 4 (b) All population ISij of simulation (corrected) vs. experiment; dark green: effects of centipedes; green: effects of spider; light green: effects of mites.

### Acknowledgements

We are grateful to Tomas Jonsson who brought the mistake to our attention and to Kevin Cazelles for a thorough review of the code.