2006 Philadelphia Annual Meeting (22–25 October 2006)

Paper No. 10
Presentation Time: 10:45 AM

THE FOUR PHASES OF PLANT-ARTHROPOD HERBIVORY IN DEEP TIME


LABANDEIRA, Conrad C., Dept. of Paleobiology, Smithsonian Institution, Washington, DC 20013, labandec@si.edu

Vascular-plant hosts, their arthropod herbivores, and associated functional feeding groups have been distributed globally into four major herbivore expansions on land during the past 420 m.y. They are: a Late Silurian to Late Devonian (Phase 1, 60 m.y. long) interval of myriapod and apterygote hexapod herbivores on several clades of primitive vascular-plant hosts; a Late Mississippian to end-Permian (Phase 2, 125 m.y. long) interval of mites and apterygote and basal pterygote herbivores on pteridophyte and basal gymnospermous plant hosts; a Middle Triassic to Recent (Phase 3, 245 m.y. long) interval of mites, orthopteroids and basal hemipteroid and holometabolan herbivores on pteridophyte and gymnospermous plant hosts; and a mid Early Cretaceous to Recent (Phase 4, 115 m.y. long) interval of orthopteroids and derived hemipteroid and holometabolous herbivores on angiospermous plant hosts. These host-plant and arthropod-herbivore associations are mediated by eight functional feeding groups: external foliage feeding, piercing-and-sucking, boring and palynivory (all early Phase 1 origins); galling, seed predation and nonfeeding oviposition (all early Phase 2 origins); and leaf mining (an early Phase 3 origin). Within about 20 m.y. of each herbivore expansion, a major biota is present for which evidence documents the full or almost complete spectrum of later plant-arthropod associations. These four associational phases are linked to paleoclimatologic variables of greenhouse/icehouse cycles and atmospheric O2 and CO2 levels by uncertain causes, albeit some relationship probably is present. The eight functional feeding groups persist through most of the 420 m.y. sampled interval, but contain host-plants and arthropod herbivores that are spatiotemporally ephemeral, indicating considerable iterative evolution of associations. Poor understanding of associations in Phases 1 to 3 is attributed to disproportionate focus on both extant and fossil angiosperm–holometabolan insect associations.