The late Eocene to early Miocene Renova Formation records initial post-Laramide sediment accumulation in the intermontane basin province of southwest Montana. Recent studies that postulate deposition of the Renova Formation were restricted to a broad, low-relief, tectonically quiescent basin on the eastern shoulder of an active rift zone vastly differ from traditional models in which the Renova Formation was deposited in individual intermontane basins separated by basin-bounding uplands. This study utilizes detrital zircon geochronology to resolve the paleogeography of the Renova Formation. Detrital zircon was selected as a detrital tracer that can be used to differentiate between multiple potential sources of similar mineralogy but with distinctly different U-Pb ages. Laser ablation-multicollector-inductively coupled plasma mass spectrometry (LA-MC-ICPMS) U-Pb detrital zircon ages were determined for 11 sandstones from the Eocene-Oligocene Renova Formation exposed in the Sage Creek, Beaverhead, Frying Pan, Upper Jefferson, Melrose, and Divide basins. Detrital zircon ages, lithofacies, paleoflow, and petrography indicate that provenance of the Renova Formation includes Paleogene volcanics (Dillon volcanics and Lowland Creek volcanics), Late Cretaceous igneous intrusions (Boulder batholith, Pioneer batholith, McCartney Mountain pluton), Mesozoic strata (Blackleaf Formation, Beaverhead Group), Belt Supergroup strata, and Archean basement. The oldest deposits of the Renova are assigned Bridgerian to Uintan North American Land Mammal (NALM) ages and contain detrital zircons derived from volcanic, sedimentary, and metamorphic rocks constituting the “cover strata” to uplift-cored Late Cretaceous plutonic bodies. Regional unroofing trends are manifested by a decreased percentage of cover strata–sourced zircon and an increased percentage of pluton-sourced zircon as Renova deposits became younger. Zircon derived from Late Cretaceous plutonic bodies indicate that initial unroofing of the McCartney Mountain pluton, Pioneer batholith, and Boulder batholith occurred during Duchesnean time. Facies assemblages, including alluvial fan, trunk fluvial, and paludal-lacustrine lithofacies, are integrated with detrital zircon populations to reveal a complex Paleogene paleotopography in the study area. The “Renova basin” was dissected by paleo-uplands that shed detritus into individual intervening basins. Areas of paleo-relief include ancestral expressions of the Pioneer Range, McCartney Mountain, Boulder batholith–Highland Range, and Tobacco Root Range. First-order alluvial distributary systems fed sediment to two noncontiguous regional-trunk fluvial systems during the Chadronian. A “Western fluvial system” drained the area west of the Boulder batholith, and an “Eastern fluvial system” drained the area east of the Boulder batholith. Chadronian paleodrainages parallel the regional Sevier-Laramide structural grain and may exhibit possible inheritance from Late Cretaceous fluvial systems. Detrital zircons of the Renova Formation can be confidently attributed to local sources exposed in highlands that bound the Divide, Melrose, Beaverhead, Frying Pan, Upper Jefferson, and Sage Creek basins. The data presented in this study do not require an Idaho batholith provenance for the Renova Formation.

Standard petrographic (e.g., QFL) and chemical (e.g., bulk chemistry) methods for provenance determination successfully classify approximately 55 to 85% of the empirical data on which the methods are based. New approaches in provenance analysis include: (1) age dating of single zircon grains or rock fragments; (2) calculating the P-T-t paths of metamorphic clasts; (3) interpreting rare earth element distribution and Sm-Nd isotope systematics to infer mantle precursors; and (4) analyzing single groups of minerals; Fe-Ti oxide minerals appear to be an especially promising group of minerals because, despite extensive chemical alteration, intergrowth patterns assist in identifying ilmenite and magnetite precursors. The goals of quantitative provenance analysis are to estimate both the proportions of source rocks represented in a body of detrital sediments and the rates at which the detritus was derived. We have quantitatively compared the distribution and the chemistry of selected heavy mineral species in Holocene sands and the sandstones of the Oligocene Renova Formation in southwestern Montana; field evidence suggests that the source rocks of these sands and sandstones were the same. Our previous statistical analysis of the chemical and textural properties of detrital opaque oxide minerals in the sands and sandstones suggested that the Renova sandstones consist of 77% igneous (mostly granodiorites and dacites) and 23% metamorphic (mostly amphibolite-facies schists and gneisses) detritus. We have developed a first approximation of an algebraic mass balance model, using characteristic index minerals for specific parent rocks, which would calculate the proportions of parent rocks that were eroded to produce a body of detrital sediments. The largest uncertainty in the model is in the estimation of the proportions of parent rocks that must be eroded to release equal amounts of the respective index minerals, and in the relative preservation potentials of the index minerals during weathering and diagenesis. We have applied this model using garnets and zircons as two index minerals to compare the proportions of parent rocks of Holocene sands and Renova sandstones in southwestern Montana. Microprobe analyses of 574 single grains of garnet and zircon show that the compositions of detrital garnets and zircons in the Holocene sands and Renova sandstones are similar, suggesting that diagenetic alteration of these two minerals in the Renova has not been extensive. Our model calculations, based on modal analyses for zircons and garnets in 143,607 heavy mineral grains, suggest that ratios of igneous and metamorphic source rocks that were eroded to produce the Holocene sands and Renova sandstones in the study area are 81:19 and 82:18, respectively. The close correspondence of these two ratios possibly suggests that relative contributions of igneous and metamorphic sources in this locality have not changed much since the Oligocene time.