Was it eating it or just resting? Sometimes they go off the host plant to rest and molt. Many predators are searching host plants for caterpillars. I saw a book listed for "Butterflies of Trinidad and Tobago", but there are only out of print copies and they are expensive. Do you have a library? Maybe you could check out some field guides there about butterflies. In North American, the caterpillars eat: "Habitat tropical to Transition Zone woodland and Citrus groves. Ranges south to Colombia, Bimini, and Cuba. Hostplants tree and herb Rutaceae: Citrus sinensis, limon, aurantium, grandis (?), Zanthoxylum fagara, americanum, clava-herculis, hirsutum, Ptelea trifoliata, Ruta graveolens, Casimiroa edulis, Dictamnus albus, Amyris elemifera, Choisya; Piperaceae: four spp. of Piper in Cuba. R. Sim observed eggs laid on (and larvae eating) Staphyleaceae: Staphylea trifolia. Records of Nyssa sylvatica (Nyssaceae), Persea borbonia (Lauraceae), and Populus (Salicaceae) seem erroneous. Both P. cresphontes and P. zelicaon(6) fed on Citrus in the San Joaquin Valley, Calif., though cresphontes was a stray there and is now eradicated." This is from "The Butterflies of North America, A Natural History and Field Guide" by Dr. James A. Scott. I also searched on the host plant database. You can put in the plant or the butterfly/moth, but you have to use the latin name. Here is a link that might be useful: Host Plant Database for Papilio cresphontes Giant Swallowtail

If you can take pictures, you can almost always have any caterpillar identified in a flash right here in the gardenweb. Learning the identity of what you find munching away on your plants is key to your decision making process of whether to let them be or not. Your description of the behavior of your pests is what led Jean to reasonably suggest that you have Fall webworms without seeing an image. Time of year, as well as the location of these webby nests were important clues. Next, you need to learn about caterpillar host plants. Many moths and butterfly adults will only deposit their eggs on certain plants. Thus, identifying the PLANT becomes important. There is nothing wrong whatsoever in sacrificing a host plant to support a caterpillar species. If you visit the Butterfly Forum, you will find a large community of people who plan and plant for that very reason. Many of the folks over there take it to the next step and collect caterpillars of favored butterflies and moths to rear in the safety of cages. If you visit that forum, and I suggest that you do, be SURE to read up in the FAQ section first, as many of your questions will be answered there. Butterfly caterpillars don't typically 'nest'. When they reach a predetermined size, they begin a occupation process which usually means the formation of a chrysalis. Moths make cocoons out of silk. The large 'nests' that you are witnessing are normal for a very few species...like tent caterpillars and webworms.

Oops. I see that you've already been there. http://forums.gardenweb.com/forums/load/butterfly/msg0916492919311.html?3 Here is a link that might be useful: the answer at the Butterfly Forum

Floral, you are right to warn gardeners about using any forms of purple loosestrife here (horticultural as well as wild) in gardens. Perennials.all, I do have a good idea of the distribution of wild purple loosestrife in Ontario and elsewhere in North America. I think you missed the point of my post. The 1992 Manitoba Study took plants of the horticultural cultivar 'Morden Pink' and planted them near a river and adjacent to wetlands, both of which had wild purple loosestrife present. The furthest any of the cultivar plants were away from wild purple loosestrife was 200 meters. 'Morden Pink' was thought to be sterile. It was presumed that the ability of plants of the cultivar to self-fertilize themselves, in the way the wild type does, had been bred out of them; namely that they were "self-incompatable". In fact, it was found that the cultivar plants lost their self-incompatability (hence sterility) in these locations in the presence of the wild type and, like the wild form, produced very large numbers of seeds. Some gardeners then wondered whether there was any evidence for sterile purple loosestrife cultivars losing their self-incompatatbility in garden situations, when the gardens were located nowhere near wetlands nor wild-growing purple loosestrife. Some simply dismissed this question out of hand. Either you believed in "the purple plague" or you didn't. Personally I'm more impressed by evidence. Are the researchers correct in extrapolating from purple loosestrife cultivars placed in research locations to those in average gardens? Over the years, I've seen almost no relevant evidence presented by gardeners. My post presents such evidence; that is seedlings which have been produced by one plant of a horticultural purple loosestrife cultivar that lost it's self-incompatability in a garden.