I would think it depends on what type of defense compounds you suspect to find in your host and non-host. There is a lot of great literature out there for cardenoloids that may be of help.

Hello Amar,

There is no simple answer to your question because Insect-Plant interactions are usually of highly multivariate determinism. Especially at the inter generic level that you are highlighting. In your example, are you speaking of one non-host plant or of many, and if one or few, why/how did you chose them ?

Your question is kind-of (almost literally) statistical issue. The biochemical difference between tea and "no-tea" plant are "immense", and only limited by your analytical procedure. So you need a filter to draw an hypothesis (usually from the literature: the chemistry of tea is probably know, the other plant might or might not be as well know as tea). From this literature, you might extract knowledge-based candidates (such as say the cardenolides quoted by Rhoda- which are by-the-way probably absent from your plants :-). Then you may draw hypothesis to test, and test it (ie a correlation between your classes and your "acceptance" index, whatever it may be). But a correlation is by no means a proof of effect, the n you have to test the results and bio-assay the compound families you have selected; usually in vitro...).

I am sure you are aware of this standard methodology, and my advice would be:

1: Without hypothesis (ie a thourough litterature survey of your plant panel), do not go !

2: If you have only too plant categories, I would also advice you not to go... This is just a statistical argument (which is: between two points, you have 50% chance to draw a line that goes the way you hypothesized it :-). This is far over any statistically testable hypothesis... So if you have more (biologically sound) plant taxa, then you might go somewhere (but also, the degrres of freedom of the chemistry will increase if you have very distant plant taxa...

I hope that I am not losing you ion my "statistical" answer (I am a biochemist actually), but it is really important to know of the "likelihood" to be able to decipher the biochemical signal from the noise...

Hope this helps

We can tell more about the chemical methodologies, but this is really not the limiting step at your question...

Sincerely

Yvan

You said you are working on a particular insect pest. If this particular pest uses plant volatiles to locate its host plant (tea plant), you might want to collect volatiles from the host plant and look at the biological activities of the most abundant volatile compounds on the insect using GC-EAD. That could be your starting point. If you can not access the GC- EAD, then you can use a simple olfactometer apparatus to determine the most active compunds.

I hope this helps

Nadir

It will depend on whether you are thinking of attraction or repulsion. Many pests are totally host-specific while close relatives are generalists. For instance the pea aphid vs the peach aphid. Then even within one species of plant there are wide differences in susceptibility as with wheat cultivars and Hessian fly (another aphid) while the Hessian fly may survive on miscanthus in wheat growing areas yet not be know as a pest.

You don't say what kind of insect you are looking at. Herbivores that eat leaves are very different from sucking insects or those that lay eggs which invade fruits or seeds as maggots.

Insects also habituate. Manduca sexta the tobacco hornworm develops taste preferences and generally if raised on tomato will not eat tobacco and vice versa. But not always. I got one from a garden pepper plant which adapted to tobacco.

Sometimes you can use an extract of leaf, if leaf is the target of the pest. Test fractions for their attractiveness in association with an artificial rearing diet. That is the class is way to sort out the biochemicals, either attractant or repusive factor.

If you talk about host and non-host plants, it means that you have already done a very important thing in this topc which is the observation in nature of the problem. This means that you have observed that a certain pest insect has chosen a plant species, or variety, but not other to develop some activity (feeding, oviposition, etc.) on it. Then, the point is how to distinguish what makes this particular species attractive to the insect.

In my opinion you have to follow two parallel strategies to gain insights in this problem: a) the evaluation of those characters of the plant that may "attract" the insect, and b) the confirmation of this atraction in artificial devices like olfactometers or multiple-choice systems.

Attraction may be based on visual or chemical clues, and at this point again the observation of the biological system is essential: Are adults or larvae the first stages in stablishing in the host plants? Whether adults fly or not may suggest what type of clues they are following. Also, many larvae are blind, so they should only follow chemical or mechanical clues. But we can not discard that in many cases, larvae are "eating" and "tasting" leaves until they find one they like and not always an "attractive" stimulus is clearly found.

So, observe your systems and as other colleagues suggested draw an hypothesis.

If your objective is the biochemical differentiation of host and non-host plants and in view of the analytical technologies that you refer you may use there is an alternative design that you could implement: a metabolomic-based survey of host and non-host plants. For instance a metabolite fingerprinting of the two types of plants using 1H-rmn and analysis by principal component analysis may reveal you differences among them without the need to identify each of metabolites present in the extracts. If effectively your PCA reveals differences then you may plann to go further with the identification of singular metabolites applying selective extraction process and either GC-MS or FT-IR analysis.

Thank you everyone for your kind suggestions. I'm grateful to all of you.

Dear Yvan Rahbe, the insect I've been working on feeds principally on tea (a sucking pest) besides feeding on a number of alternate hosts which neither belong to a particular taxonomic category (like family, tribe, etc.) nor they look alike. The most interesting thing about its behaviour is that the pest does not lay egg in all the alternate hosts where it feeds on. So I've been trying to know the criteria (either chemical or morphological) that should be there in a plant to be its host/alternate host or a non-host.

Dear Lawrence Davis, I've been trying to know the chemical composition that attracts the insect towards the plant; and as you said, the pest is exclusively a sucking pest and the target of the pest is the young leaves where it feeds on and the soft delicate stem for oviposition.

Dear Jose L. Casas, your suggestion is really appreciative.

Thank you.

Hello again Amar,

Then it is a sucking-pest (you may know that I am an "aphid man"...! :-)

Then also, like Jose stressed it, an important item is the thourough knowledge of the biology of your model insect, and of the parent family. From the features you are stressing (ie real polyphony, am I right ?), your insect should not be an aphid. And as you are pointing out to egg-laying, then an important feature to investigate is the determinism of egg-laying in your model (general fitness -ie say antibiosis on non-hosts ?-, or odor-based cues, or gustatory-based cues ?). You may build a bioassay on this, and then run for a bioassay-guided fractionation (which is a way I would prefer on pure correlation-based chemical analyses -unless you might have interesting a priori hypotheses to test ...). Of course chemistry is not all (you mentioned morphology)...

Hope this helps too.

Note: If it an aphid, it is really peculiar (and interesting !)...

To differeniate biochemically host and non-host plants is a herculean task. One should rember that non-host plants not necessarily contain the defensive chemicals. Even carbohydrates, protenins etc if not present in proper ratio as desired by the insect, will qualify for a non-host plant. In this regard one has to read the basic and fundamentals concepts on insect nutrition.

Good luck.