Population responses to productivity, food chain length, and complexity in aquatic microcosm

The role of productivity in limiting food chain length, and the subsequent influence of food chain length on trophic cascades, are central controversies in ecology. Classic theory predicts that food chain length will increase with productivity, and that the resultant number of trophic levels will determine population responses to further productivity increases. Recent theory questions productivity's involvement in food chain length and whether complexity within trophic levels prevents trophic cascades. Experimental tests to assess productivity and complexity's role in determining food chain length and population abundances are rare.;I assembled one-, two-, and three-level food chains with one type of species per trophic level along a productivity gradient in laboratory microcosms to examine the relationship between productivity and food chain length. Bacteria occupied the first trophic level, the bacterivorous ciliate Colpidium the second trophic level, and the predatory ciliate Didinium the third trophic level. Increased productivity increased food chain length from two to three trophic levels. In food chains with two-trophic levels Colpidium increased as productivity increased, limiting bacterial increase. In three-level food chains, Didinium increased, preventing Colpidium from increasing, and allowing bacteria to increase. These results support predictions of classic prey-dependent food chain models.;I assembled microbial food webs with different trophic architectures and measured population responses to altered productivity to explore the importance of trophic level complexity. A linear food chain consisting of bacteria, Colpidium, and Didinium served as a reference system for comparison with more complex systems. Three additional food web configurations varied both the number of species per trophic level and the pattern of feeding linkages among trophic levels (additional species included the bacterivorous ciliate Loxocephalus and the algae Chlorella or Scenedesmus). Complex food webs responded differently than the linear food chain to increased productivity. In two of three complex webs, top predators did not respond to increased productivity. Two webs, one linear and one complex, performed as predicted by theory, while the remaining two complex webs departed from the responses predicted by Abrams (1993). Trophic complexity clearly affected population responses to altered productivity, sometimes producing responses not predicted by simple models.