INTRODUCTION

The Verde Formation of central Arizona, USA, is a Miocene-Pliocene sequence of predominantly lacustrine limestones and mudstones, with occasional interbedded volcanics, evaporites, fluvial sediments, and spring deposits (Twenter and Metzger 1963; Nations et al. 1981; Figure 1). Ancient Lake Verde is interpreted by these authors as having been a spring- and river-fed, shallow, algae-filled lake with marshy edges surrounded by terrestrial habitats including open oak-juniper savanna and grasslands with broadleaf deciduous gallery forest along the lake edge and inflowing streams. In general, vertebrate fossils are rare and known mostly from a few localities in the Pliocene portion of the formation, reflecting an early Blancan land mammal age (Czaplewski 1987a, 1987b, 1990). However, some localities in the Verde Formation might bear a late Hemphillian fauna (Lindsay and Czaplewski, this volume).

Twenter (1962) mentioned some of the first vertebrate fossils known from the Verde Formation. At one locality east of Cottonwood on the south flank of House Mountain, Yavapai County (Figure 2, Figure 3), he recovered fossils of small mammals, which were sent to C.W. Hibbard, who identified teeth of the rodents Dipodomys (= Prodipodomys), Bensonomys, and Sigmodon (Twenter and Metzger 1963). In 1979 I relocated this site and removed sediments from it. The sediments amounted to less than 10 kg of rock, but when screened underwater by standard techniques, produced a dozen isolated rodent molars from at least four different taxa. Further screening produced abundant rodent teeth and small vertebrate bone fragments. When small discrete samples of sediment were screenwashed, complete or nearly complete "sets" of maxillary or mandibular teeth showing equivalent degrees of wear and similar preservation were recovered. These results suggested that complete skulls and lower jaws of individual rats and mice were present but were destroyed during screenwashing. Part of the deposit was quarried in 1979-1980 in order to recover unique taphonomic data. The site is Museum of Northern Arizona (MNA) locality 318 in Coconino National Forest; all specimens are cataloged with MNA numbers.

Fluvial transport and accumulation by predators are often invoked as mechanisms involved in the accumulation of vertebrate bones (e.g., Dodson 1973; Wolff 1973; Mellett 1974; Mayhew 1977; Hunt 1978; Korth 1979; Hanson 1980; Andrews and Nesbit Evans 1983; Hoffman 1988). Owl pellets are commonly cited as a source for accumulations of small vertebrate fossils (Kusmer 1990; Denys et al. 1996; Terry 2004). However, effects such as weathering, disintegration, disturbance by insects or other invertebrates, transport, and burial of pellets after they have been regurgitated can bias analyses of owl pellet remains and might influence paleoecological interpretations (Terry 2004).

Cursory examination of the House Mountain fossils suggested that the assemblage represented an accumulation of bones from owl pellets. Although accumulation by owls is sometimes invoked as the cause when a concentration of microvertebrate fossils is discovered, few authors have actually attempted to support such claims with empirical data in open-air localities, probably because of the time and difficulties involved in quarrying microvertebrates localities. (However, some authors have addressed microvertebrate taphonomy in cave localities, e.g., Andrews 1990; Hockett 1991; Hoffman 1988; Jass 2002.) Walton (1990) gave a few examples of microvertebrate sites where owl pellet accumulation was implied, including mention of the House Mountain assemblage in the Verde Formation before the present analysis was completed.

Different taphonomic agents, including vertebrate predators, impose distinctive effects on vertebrate skeletons such that it may be possible for the paleontologist to identify the agent of accumulation of a given suite of bones (Dodson and Wexlar 1979). The purpose of the present paper is to document the concentration of early Pliocene small vertebrate remains at MNA locality 318 and to evaluate the taphonomic character of the assemblage. Production of microvertebrate quarry maps provided the opportunity to assess taphonomic data often unavailable from microvertebrate localities (e.g., bone orientation patterns). Few previous researchers have quarried sites with microvertebrate concentrations for the purpose of mapping the bones in situ (Irwin et al. 1997; Wilson 2006), rather than screenwashing quarried rock and thereby losing taphonomic information. No such studies, and no microvertebrate quarry maps, appear to have been published previously.