Abstract The principal purpose of this study is to determine from geological and biological evidence how a concentration of fossil bones originated. The Verdigre quarry of earliest Pli— ocene age in the Valentine Formation, northeastern Nebraska forms the basis for the investi— gation. Some 40 genera of mammals, 7 of reptiles, 2 of amphibians, and 6 of fishes are rep— resented in a collection of over 20,000 identifiable specimens recovered from a small exca— vation in a 4-5 foot-thick lens of extremely poorly sorted, gravelly, silty, clayey sand. The open framework and textural inversion of the fossiliferous stratum indicate very rapid deposition, probably in a single flood. Preferred orientation of long bones shows that the depositional current flowed from the NNW. The site of deposition was an off-channel de— pression on a broad floodplain which was subaerially exposed (mudcracked)prior to the flood. The great majority of the fossil remains do not represent animals killed by the flood as they are completely disarticulated and often somewhat abraded, except for 7 articulated skele— tons. Stream table experiments suggest that current sorting is probably responsible for the scarcity of elements such as ribs, vertebrae, sacra, and phalanges compared with rami, meta— podia, tibiae, etc. Discrete age groups, distinguished on the basis of tooth eruption and wear.(Kurten, 1953), are present in Verdigre mammals represented by large samples. Four hundred seventy-five Mery— aodus (antilocaprid)individuals fall into 7 year-classes: one juvenile (.8 year)and 6 adult (1.8- 6.8 years). Burrs on the horns of Meryaodus males are annual but cannot be used in population dynamics studies of collections of transported remains because they are easily re— moved by abrasion. Fifty-seven Protohippus (3-toed horse)individuals fall into 3 juvenile (-.2, .8, and 1.8 years)and at least 7 adult age classes.The lack of individuals of inter— mediate ages and the large proportion of animals of reproductive age in the samples indicate that non-selective catastrophic death accounts for most, and possibly nearly all, of the dis— articulated mammal remains in the quarry. Merycodus and Protohippus at least, died in the winter –– about 2-3 months before the season of giving birth. Growth ring study of catfish vertebrae and gar scales indicates that these aquatic forms were also winterkilled. Shotwell's (1955)method of determining the proximity of various habitats to vertebrate quarry sites is not applicable to the Verdigre assemblage, and probably not to other collections, because it fails to consider e.g.: sedimentary evidence, such as indications of size-sorting the possibility different effects of catastrophic and attritional mortality on the composition of quarry faunas the great mobility of grazing mammals. There is evidence that size sorting is responsible for differences in ‘completeness’ of various mammals in Shotwell's published samples as well as in the Verdigre quarry. Sites of deposition of late Tertiary quarry faunas were in stream valleys; the interstream grassland habitat was probably rarely if ever directly sampled by depositional processes. Mass occurrences of grazing animals such as in the Verdigre bonebed are interpreted as resulting from mortalities which occurred down on floodplains at times of environmental stress. Early Pliocene vegetation in the Verdigre area resembled that of parts of the Gulf Coast of southern Texas and northern Mexico; forests occupied the broad floodplains and savannas the interfluves (MacGinitie, 1962). Crocodiles and giant land tortoises in the fauna indicate a normally frost-free climate. The presence of sharks and batoids suggests marine influence; stream gradients may have been significantly lower and the shoreline of the Gulf may have been closer to Nebraska in the early Pliocene than at present.

Abstract The principal purpose of this study is to determine from geological and biological evidence how a concentration of fossil bones originated. The Verdigre quarry of earliest Pliocene age in the Valentine Formation, northeastern Nebraska forms the basis for the investigation. Some 40 genera of mammals, 7 of reptiles, 2 of amphibians, and 6 of fishes are represented in a collection of over 20,000 identifiable specimens recovered from a small excavation in a 4-5 foot-thick lens of extremely poorly sorted, gravelly, silty, clayey sand. The open framework and textural inversion of the fossiliferous stratum indicate very rapid deposition, probably in a single flood. Preferred orientation of long bones shows that the depositional current flowed from the NNW. The site of deposition was an off-channel depression on a broad floodplain which was subaerially exposed (mudcracked) prior to the flood. The great majority of the fossil remains do not represent animals killed by the flood as they are completely disarticulated and often somewhat abraded, except for 7 articulated skeletons. Stream table experiments suggest that current sorting is probably responsible for the scarcity of elements such as ribs, vertebrae, sacra, and phalanges compared with rami, meta-podia, tibiae, etc. Discrete age groups, distinguished on the basis of tooth eruption and wear (Kurten, 1953), are present in Verdigre mammals represented by large samples. Four hundred seventy-five Merycodus (antilocaprid) individuals fall into 7 year-classes: one juvenile (.8 year) and 6 adult (1.8 - 6.8 years). Burrs on the horns of Merycodus males are annual but cannot be used in population dynamics studies of collections of transported remains because they are easily removed by abrasion. Fifty-seven Protohippus (3-toed horse) individuals fall into 3 juvenile (-.2, .8, and 1.8 years) and at least 7 adult age classes. The lack of individuals of intermediate ages and the large proportion of animals of reproductive age in the samples indicate that non-selective catastrophic death accounts for most, and possibly nearly all, of the disarticulated mammal remains in the quarry. Merycodus and Protohippus, at least, died in the winter — about 2-3 months before the season of giving birth. Growth ring study of catfish vertebrae and gar scales indicates that these aquatic forms were also winterkilled. Shotwell’s (1955) method of determining the proximity of various habitats to vertebrate quarry sites is not applicable to the Verdigre assemblage, and probably not to other collections, because it fails to consider e.g.,: sedimentary evidence, such as indications of size-sorting the possibility different effects of catastrophic and attritional mortality on the composition of quarry faunas the great mobility of grazing mammals. There is evidence that size sorting is responsible for differences in ‘completeness’ of various mammals in Shotwell’s published samples as well as in the Verdigre quarry. Sites of deposition of late Tertiary quarry faunas were in stream valleys; the interstream grassland habitat was probably rarely if ever directly sampled by depositional processes. Mass occurrences of grazing animals such as in the Verdigre bonebed are interpreted as resulting from mortalities which occurred down on floodplains at times of environmental stress. Early Pliocene vegetation in the Verdigre area resembled that of parts of the Gulf Coast of southern Texas and northern Mexico; forests occupied the broad floodplains and savannas the interfluves (MacGinitie, 1962). Crocodiles and giant land tortoises in the fauna indicate a normally frost-free climate. The presence of sharks and batoids suggests marine influence; stream gradients may have been significantly lower and the shoreline of the Gulf may have been closer to Nebraska in the early Pliocene than at present.