Before Farming 2008/2 article 2 1 A behavioural ecological approach to a proposed middle Holocene occupational gap Raven Garvey Department of Anthropology, University of California, Davis, One Shields Avenue, Davis, CA 95616, USA rgcarper@ucdavis.edu Keywords Argentina, behavioural ecology, climate change, middle Holocene, occupational hiatus Abstract Relative to other periods, there are very few recorded sites of middle Holocene age in Mendoza Province, Argen- tina. This analysis weighs the preliminary results of a regional study in Mendoza against three possible explana- tions for the small number of middle Holocene sites: poor visibility, reduced population density and behavioural adaptation. While changes in visibility and population density cannot be dismissed, evidence suggests that the scarcity of mid-Holocene sites in Mendoza reflects changed patterns of land and resource use triggered by climate change. Archaeological data from the region fit the predictions of Charnov’s marginal value theorem and a cost-benefit model of lithic procurement, and suggest that hunter-gatherer groups curtailed their mobility during the middle Holocene, producing a more cryptic archaeological record. 1 Introduction Archaeological and palaeoenvironmental data indi- cate that the middle Holocene (between 8000 and 4000 BP) was a time of warmer, more arid conditions in many parts of the world (Albanese & Frison 1995; Anderson et al 2007; Antevs 1948; Butzer 1957; Nuñez & Grosjean 1994; Sandweiss et al 1999). Indigenous populations living in areas affected by this trend ap- pear to have been impacted by resultant changes in resource availability (Antevs 1948; Deacon 1974; Grayson 1993; Humphreys & Thackeray 1983; Meltzer 1999; Sheehan 1994). Researchers in southern Mendoza Province, Argentina (figure 1) report a local expression of this larger phenomenon, most notably a conspicuous scarcity of archaeological sites dating to the middle Holocene (Gil et al 2005; Markgraf 1989; Nuñez & Grosjean 1994). (NB: Mendoza is the name of both the province and its capital city. Hereafter, Mendoza will be used to refer to the province. Thus, ‘southern Mendoza’ is the southern part of Mendoza Province, between approximately 34–37° south lati- tude.) The perceived gap in the Mendoza record resem- bles gaps reported for other parts of the western Hemi- sphere during the middle Holocene (Antevs 1948; Gil et al 2005; Grayson 1993; Humphreys & Thackeray 1983; Markgraf 1989; Meltzer 1999; Nuñez & Grosjean 1994; Sheehan 1994). The apparent similarity of re- sponses across so vast and varied an expanse in- spired a session devoted to middle Holocene adap- tations at the 2008 Annual Meeting of the Society for American Archaeology in Vancouver, British Colum- bia (Garvey et al, this issue). The project described here is still in its earliest stages and the data are consequently few, but the Mendoza study serves as a vehicle for a broader discussion of theoretical and methodological approaches to the middle Holocene. Two concepts guide the present research. First, occupational hiatus at scales detectable in the ar- chaeological record have profound implications for our understanding of human adaptive capabilities and it is correspondingly critical that we rule out non-be- havioural factors whose expression can mimic aban- donment or population diminution before interpreting the significance of perceived gaps. Second, behav- ioural ecological models offer a powerful means of exploring the relationships between people and their environments and of predicting responses to chang- ing resource availability during the middle Holocene. Testing archaeological data against these models’