查看文章

To resume collection after the drought in January 1997, a secondary course was built to bypass the site (Figure S1) in July 2009. In February and March 2011 plastic bags were filled with the dirt removed and then piled up to work as dams on both ends of the site. The remaining water, along with that continuously supplied by the natural aquifer, was pumped out all throughout the fieldwork, which took place in late March 2011 and more recently in late January-early February 2012. In 2011, the site was uncovered and photographed and bones exposed in an area of about 30 m 2 in the streambed, at an average depth of 2 metres. Those bones regarded as too vulnerable to be left without risk of loss were collected. A metre grid system was established to map the specimens (Figure S2). In addition, samples were taken for ongoing analyses, including grain size