2001 06 15: Application of trophic transfer ef ciency and age structure in the trophic analysis of fossil assemblages. Lethaia, Vol. 34, pp. 97-118. Oslo. ISSN 0024-1164. We evaluate onshore-offshore trends in age-frequency distributions and trophic transfer ef ciencies using 11 modern death assemblages off the Texas coast. Trophic transfer ef ciencies within trophic levels offer little insight over that achieved by a size-frequency distribution. Production/biomass ratios will always be 1 in the fossil record. Within trophic-level estimates of paleogrowth ef ciency, the ratio of paleoproduction to paleoingestion (P i g lt /I i lt where i indicates the i th trophic level and lt indicates the timeaveraged value) follow the expected ecological trend precisely in that paleogrowth efciency is consistently higher in primary consumers than in predators in all 11 death assemblages. Paleoutilization ef ciency, the ratio of predator paleoingestion to prey paleoproduction, I 2°l t /P 1°g lt , may provide information on the degree of bias in the preservation of primary (1°) and secondary (2°) consumer trophic groups. I 2°l t /P 1°g lt fell below 0.1 in most cold-seep and bay assemblages, indicating a large surplus of primary consumers. In sharp contrast, I 2°l t /P 1°g lt was above 10.0 in most heterotrophic continental slope and inner continental shelf assemblages, indicating an overrepresentation of predators. In the classic food web, predator life span should exceed prey life span (¢L lt :. Predators should be more insulated from uctuations in food supply than their prey and the degree of this insulation should control ¢L lt . Of the 11 death assemblages covering the bay-to-continental slope transect, predators' life spans were longer than their prey only on the continental shelf and slope, indicating a requirement for a stable enough environmental regime to permit biological processes rather than physical factors to control life spans. Paleoproduction transfer ef ciency, the ratio of paleoproduction between two consecutive trophic levels, P 2ḡ lt =P 1ḡ lt , is the most accurately estimated trophic transfer ef ciency in paleoecology because it does not require an estimate of age-at-death. The relationships of paleoutilization ef ciency and paleoproduction transfer ef ciency are nearly identical among the 11 death assemblages. Although theoretically this need not be the case, these assemblages suggest that predator overrepresentation might be identi ed from a ratio dependent solely upon size-atdeath. &Age frequency, death assemblage, paleoecology, paleoenergetics, trophic transfer ef ciency, taphonomy.Eric N. Powell [eric@hsrl.rutgers.edu], Haskin Shell sh Research Laboratory, Rutgers University, 6959 Miller Ave.,