After plants it’s time think about what herbivorous insects would be useful for a tritrophic experiment.
Species: The plant species of the experiment has of course an effect on the possible herbivorous insect species. I have in my own experiments used generalist species, most often autumnal moths (Epirrita autumnata, photo on left). That is a common species throughout Finland, especially in north, and it eats almost anything deciduous. But I have also used twice sawfly (Arge fuscinervis, photo in middle). I don’t know what its English (or Finnish) name is and I think also its scientific name has changed but that was its scientific name in 2003-4 😉 . I have used autumnal moth caterpillars as herbivores of mountain birches (Betula pubescens ssp. czerepanovii; Mäntylä et al. 2008a, 2014), silver birches (B. pendula; Mäntylä et al. 2008b) and downy birches (B. pubescens; Mäntylä et al. 2010). Arge-sawfly larvae ate mountain birches in my experiments (Mäntylä et al. 2004). In my experiments with Scots pines (Pinus sylvestris) I used pine sawflies (Diprion pini, photo on right, and on the top of this posting), as they are herbivores of pines. That was an easy choice, since the research group where I was working, reared a laboratory population of the pine sawflies. Usually people think only caterpillars/larvae as herbivorous insects in experiments but depending on the species you can get tritrophic effects also with eggs (Hilker & Fatouros 2015) or adults (e.g. Kagata et al. 2005).
Whatever herbivorous insect species will be studied, its background information should be checked, such as its lifecycle, host plants, parasitoids and predators. Is it native or alien in the study area or for the study plant? Does it have outbreaks? Does it have toxins or bad-tasting chemicals? Is it aposematic?
Availability: It will be easy to use the herbivorous insect species that the university/research institute has a laboratory rearing or constant access. But if a specific species is needed, but not available, ask from researchers who use that. They may have extra individuals and/or want to use their insects in a new study. If the species has outbreaks, it is possible to collect those from nature. But then there is a probability to carry also parasitoids and/or diseases with the herbivorous insects.
Permits: I have always in my studies with birches in nature had the larvae inside mesh bags that covered one branch (photos below). The branch inside the bag was usually around 80 cm long. For this you almost always need a permission/licence from the land owner. No one wants that pest insects would spread to nature from your experiment. I always collected the insects away after the experiment. They were then usually in their last larval instar or already pupated.
Amount: How many herbivorous insects are enough to cause an effect in the host plant? That depends on the species (of both plant and insect), instar, how long the insects are defoliating and probably a dozen other things. In my experiments, I usually have put around 30 individuals of 2nd instar autumnal moth caterpillars inside one mesh bag and removed them when they are in their last (5th) instar. The time of defoliation is then usually two weeks and defoliation percentage of the leaves has been between 30 to 100 %. Getting total defoliation is not necessarily a good thing. A good percentage could be something like 50-70 %. For small trees two branches with mesh bags has been enough, and for bigger trees I’ve used 3-4 bagged branches. And remember that control trees (with no herbivores) will need the same amount of mesh bags but with no insects inside. If adult insects are released inside the mesh bags to deposit eggs or insect eggs are brought there, then it is more difficult to control that the amount of caterpillars/larvae will be similar in each mesh bag.
Systemic induction: In my experiments I’ve always been interested in the systemic induction of the plants, i.e. the reaction of intact branches of trees that have had herbivorous insects on other branches. The systemic induction is usually strongest in those branches that have a close connection (i.e. not via the trunk) to the herbivore-damaged branch. This is something that should be thought of already when deciding which branches will get the mesh bags.
This is a short list of what to consider when planning an experiment. Please, leave comments if you want to add some more information.
© Elina Mäntylä (firstname.lastname@example.org), 13 June 2016