Field paleontologists are often at their best (and happiest) when free to roam over promising exposures, letting experience and instinct lead them to the places where fossils occur. Will Downs (Figure 1) was a master of this approach, and the field of vertebrate paleontology benefited greatly from his talents at locating and collecting highly productive fossil sites. However, areas where fossils are relatively abundant and dispersed over the outcrops also lend themselves to a different collecting strategy, one in which controlled surface surveys enhance the traditional focus on the discovery of richly productive sites or unusual fossils. This methodology, which will be described below, was developed by the Harvard – Geological Survey of Pakistan – Smithsonian research team in the 1970s (Barry et al. 1980) and was used initially to document biostratigraphic changes through time in the Miocene Siwaliks. It is a tribute to Will that in spite of his aversion to “controls” of any kind, he participated in these surveys and contributed to the development of the methodology. Fundamentally, he was interested in the science and what the team could learn with new approaches, even when these might put a damper on his preferred way of doing things.

The basic goal of controlled surface sampling of the vertebrate record is to document, in a manner as free of collecting biases as possible, the fossil assemblage that occurs on the ground surface at a particular stratigraphic level. Documentation targets any remains that have been naturally exposed and are visible on outcrop surfaces – from whole bones and teeth to scraps of bone, as well as coprolites or other trace fossils. Identification at any level of taxonomic resolution is potentially useful, and the resulting data can address questions such as the proportion of fish versus reptile versus mammal, the frequency of tooth versus limb and vertebra fragments, or the relative abundance of equid versus bovid teeth. The controlled surface surveys provide information that usually cannot be recovered from museum catalogues or traditional, taxon- or body-part specific collecting strategies, and the two approaches are complimentary when applied to the same strata. The surface surveys have their own sets of biases, which will be discussed below, but these biases differ from those in collections oriented toward the recovery of the more complete and identifiable specimens. Perhaps the most important benefit of controlled surface surveys is that they can be repeated at different stratigraphic intervals, as was done in the Siwaliks, thereby providing information on changes in the taxonomic composition of the fossil assemblages through time, i.e., biostratigraphic trends. The same methods can also be used for comparisons of contemporaneous faunas and skeletal part assemblages in different areas or lithofacies. The key is to eliminate noise from inconsistent collecting strategies so that such comparisons result in reliable information about the bone assemblages themselves.

In this paper, we describe the methods that were developed by the Siwalik research team to investigate biostratigraphic change through time in the vertebrate paleocommunity. We also provide examples of some of the results of these sampling methods and guidelines for using them in other regions and time intervals.