Figure 1. Experimental setup. The present study is based on two switch experiments in which the larvae were moved to a different host plant. during their development. (a ) The first switch experiment (Switch A) included Urtica dioica (Ud ),Salix caprea (Sc ) and Ulmus glabra (Ug ). For the second experiment (Switch B) Urtica dioica , Salix caprea and Ribes uva-crispa (Ru ) were used. A summary of all host combinations are given for each switch experiment. Here rows correspond to the original host (Host 1) and columns to the host the larvae were transferred to. (b ) In both experiments larvae were first reared on one out of three established host plants (Host 1). When reaching the fourth larval instar, caterpillars were moved to a new host plant (Host 2). This new host could either be of the same (Control, e.gUdUd ) or one of the alternative plant species (Switch, eg.UdSc ). After the transfer larvae were allowed to feed for 2 respectively 17 hours on Host 2 before the midgut was collected for RNA-extraction and sequencing. A summary of all host combinations per sampling time point are given for each switch (d ande ). Here rows correspond to the original host, while columns represent the host the larvae are transferred to. Switches to the same host species (Control) are highlighted in grey.