Roger Kitching’s bookcould be regarded asthe first field guide tothe fascinating worldof life in ephemeralbodies of water.Phytotelmata,aquatic organismsliving in containers ofplant origin, whichrange from flatworms to crustaceans toaquatic larvae of insects and frogs, occurnot only in tree holes, bamboo internodesand basins of buttressed trees, but also inbromeliad tanks, palm fronds and otherwatertight axils of leaves and floral bract,and even in the highly acidic fluid in thepitchers of pitcher plants. How unique arethe assemblages of such organisms inphytotelm and do they function as acommunity, showing the emergentcharacters, such as stable food webs, thatare seen in other communities?This book presents evidence andtantalizing facts about the speciesdiversity, trophic relations and evolution ofphytotelmata. Based on his own extensivework, Kitching constructs species-basedfood webs of the saprophagous and thepredatory animals found in tree holes andpitcher plants, and examines global,continental and regional patterns in theseand others reported in the literature.Disregarding microbes, rotifers andnematodes, he singles out competition andpredation as the key forces that structurethe phytotelm community and shows howbiogeographical and ecological factorsinfluence the pattern and process inthese communities.A drawback to this approach is that it isentirely binary in linking species (havingor not having a connection) and suffersfrom a lack of quantitative information onthese links. It is difficult to assess the rolesof individual species in these habitatsbecause nutrient input will differdepending on both the host plant and thenature of its container (pans or rot holes).All species of saprophytes need to beknown and their competitive andcomplementary functions in nutrient andenergy consumption worked out toproduce comprehensive food webs forthese systems. However, the top-downeffects of predators are easier to assessand here an experimental approach can beused to manipulate the food-webstructure. Not surprisingly, impacts ofpredators are more dramatic in smallerphytotelm units than in larger ones, butnot all predators, even those occurring inhigh density, have a keystone role incontrolling the overall food-web structure.Since 1960 (Ref. 1), ecologists havefocused on the analysis of food webs inmicrocosms, such as small springs,decaying logs and cowpats. This was anattempt to produce an accurate descriptionof the entire community in a semi-closedsystem so that all species present in thatsystem could be accounted for in terms ofenergy transfer and trophic relations. ThisLindeman-style approach to ecosystemanalysis