LETTER TO THE EDITOR Dog attacks on adders; a comment on Worthington-Hill & Gill (2019) T. Madsen 1,2 , L. Luiselli 3,4 & P. Janssen 5 1 School of Biological Sciences, University of Wollongong, Wollongong, NSW, Australia 2 Centre for Integrative Ecology, School of Life & Environmental Sciences, Deakin University, Geelong, VIC, Australia 3 Institute for Development, Ecology, Conservation and Cooperation, Rome, Italy 4 Department of Applied and Environmental Biology, Rivers State University of Science and Technology, Port Harcourt, Nigeria 5 Working Group Adder Research Netherlands (WAN) RAVON. Natuurplaza, Nijmegen, The Netherlands Correspondence Thomas Madsen, School of Biological Sciences, University of Wollongong, Wollongong, NSW 2522, Australia. Email: madsen@uow.edu.au doi: 10.1111/acv.12535 Due to its impact on organismal behaviour, physiology, demography and its effect on conservation management, quantification of predation in the wild constitutes a crucial component in ecological studies (e.g. Magnhagen, 1991; Creel et al., 2007). In particular, quantification of predation risk on snakes is challenging and therefore researchers have often relied on using plasticine models which enable the identification of the predator attacking the models (e.g. Mad- sen, 1987; W€ uster et al., 2004; Niskanen & Mappes, 2005; Valkonen et al., 2011a,b). In a recent study entitled ‘Effects of large-scale heathland management on thermal regimes and predation on adders Vipera berus’ by J. O. Worthington-Hill & J. A. Gill, the authors quantified predation on adders by using 770 plasticine adder models in 10 heath land study sites. Of the 770 models, 26.4% were damaged but the proportion of damaged models differed among the sites, ranging from 6.3% to 68.8%. In figure 5 of the above-mentioned paper, the authors pre- sented data on the cause of the damage to the models. It is clear from the figure that damage caused by dogs constituted the major cause of damage across the 10 sites. This led the authors to suggest that dogs would be one of the main predators of adders across the 10 study sites. We have been conducting long-term studies of adders in Sweden (39 years), The Netherlands (42 years) and in Italy on asp vipers Vipera aspis for 32 years. In spite of the sometimes abundance of dogs, we have never recorded any predation on adders by dogs (T. Madsen unpublished data, P. Janssen unpublished data, L. Luiselli unpublished data). We have, however, recorded many dogs getting bitten. For example, data from the ‘Working group Adder research Netherlands’ show that each year at least 10 dogs are bitten by adders. These bites occur not because of aggressive beha- viour by the dogs towards adders, but more due to curiosity. Similarly to Sweden, in the Netherlands, there are only a few cases recorded where the dog actually died from the bite, which is supported by a study on adder bites on dogs conducted in the UK (Sutton et al., 2011). When one of us (TM) used plasticine models to investi- gate predation on juvenile grass snakes, the models were fre- quently nibbled by small rodents (Madsen, 1987). Madsen did not regard this as predation but rather that the small rodents nibbled on the plasticine because of its olfactory cues. This was further supported by a study conducted in Italy by (LL) using ‘plasticine balls’ and ‘plasticine strips’, where the frequencies of attacks by rodents on both models were very similar to those recorded in the Swedish study (L. Luiselli et al., unpublished data). However, in their study on fire salamanders Vela-Ant on & Cordeo-Rivera (2011) did not observe any rodent attack on ‘plasticine balls’, but since no data were provided on how many of the salamander mod- els were attacked, it is difficult to make an accurate compar- ison of attack rates on the two types of models. In conclusion, we believe that the frequent dog attacks recorded by Worthington-Hill & Gill (2019) should not be regarded as predation. Instead, we suggest that, similarly to what was recorded in rodents in by Madsen (1987) and L. Luiselli (unpublished data), the dogs were attracted to the models by olfactory cues, resulting in that the dogs nibbled on the models to investigate whether they were edible. Using plasticine models, in our view, constitutes an ade- quate way to investigate predation caused by visually hunt- ing snake predators, but not of predation caused by predators primarily relying on olfactory cues. Acknowledgements We are grateful for the comments provided by three referees and Elina Rantanen. References Creel, S., Christianson, D., Liley, S. & Winnie, J.A. (2007). Predation risk affects reproductive physiology and demography of elk. Science 315, 960–960. Animal Conservation 23 (2020) 119–120 ª 2019 The Zoological Society of London 119 Animal Conservation. Print ISSN 1367-9430