LAUSR

We are grateful for the insightful comments provided by Dr. Bellezza on our paper. He raises vital points about the nature of climate and continentality. We agree with Bellezza that the low density of available weather station data from the central and western Tibetan Plateau would result in a lower degree of accuracy in any modeled reconstruction. Our analysis focused, for these reasons, on the better-understood eastern Tibetan Plateau, where higher numbers of available weather stations improve the accuracy of reconstructions. We hope that in the future more weather data from the central and western Tibetan Plateau, both areas crucial to modern Tibetan agriculture, will be made available to foreign researchers so that researchers can create a more accurate picture of changing cropland across this area. In his comment Bellezza scrutinized our definition of the altitudinal boundary of the Tibetan Plateau and how this affects interpretations of percentage of cropland allotted to agriculture. In our paper we followed the methods used by Chen et al. (2015b), who defined the northeastern Tibetan Plateau as having a lower altitudinal limit of 1,500 m asl. We agree that this lower altitudinal limit is problematic. As Bellezza points out, setting the altitudinal limit this low includes large tracts of agricultural land such as the QaidamBasin and Xining on the northeastern Tibetan Plateau. We have recreated the analysis in our supplemental information by setting the altitudinal limit at 2,700 m asl to eliminate these areas of lower altitude. According to GLC-SHARE, only ∼770 km of cropland-dominant land on the Tibetan Plateau is above 4,000 m asl, or ∼4.7% of cropland-dominant land is above 2,700 m asl (∼16,500 km). Figure 1 presents a revised empirical cumulative probability density distribution of the elevations of cropland-dominant land on the Tibetan Plateau above 2,700m asl; 95% of crop-dominant land above 2,700 m asl is below 3,980 m asl. If we constrained the area in our analysis to east or west-central Tibet as suggested by Bellezza, this might change these figures further. Regardless of how one defines the Tibetan Plateau, Bellezza’s call to ground truth contemporary and ancient agricultural cropland data is well warranted. On the archaeological front, we have argued that it is necessary for archaeologists to move beyond the assumption that seeds found in sites represent in situ cultivation, as this neglects the fact that early Tibetans likely engaged in extensive networks of trade much as they do