news and update ISSN 1948-6596 commentary Differences in climate and soil partly predict land use, popula- tion density, and conflict In The Wealth of Nations (1776), Adam Smith ar- gued that countries guided by the invisible hand of unencumbered economic markets would pros- per, but he acknowledged that climate, isolation, and lack of access to rivers and oceans would hin- der a nation’s development. Jared Diamond’s Guns, Germs, and Steel (1997) famously reinvigo- rated the consideration of how biogeographic variables have affected the fortunes of different world regions. Diamond argued that the size and orientation of Eurasia, together with its fortuitous endowment of climate, soil, and domesticable plants and animals, favored the earlier develop- ment there of agriculture, cities, and empires and that this “head start” affects economic fortunes to this day. Some studies support Diamond’s hy- pothesis that biogeographic variables strongly af- fect economic development (e.g. Hibbs and Olson 2004; Presbitero 2006) and that the effects of the Neolithic revolution, the first agriculture revolu- tion, still influence world incomes (Putterman 2008). Others argue that the “right” mix of institu- tions, meaning societal rules like property rights and competitive markets that influence behavior and economic incentives, internally promoted economic and technological development and were more important than biogeographic condi- tions (e.g. Acemoglu et al. 2002a; Easterly and Levine 2003). Beck and Sieber (2010, PLoS One) extend Diamond’s framework one step back, assessing whether climate and soil alone can predict the distribution of four basic land uses (agriculture, sedentary animal husbandry, nomadic pastoral- ism, and hunting-and-gathering), which reflect basic aspects of human cultural and economic practice. They then assess whether climate and soil affect population density, presumably through land use differences, and whether modeled re- gions of “overlap” in land use suitability predict locations of violent conflict. They focus on all but the Americas and Antarctica and divide this ex- panse into a 5km x 5km grid in a GIS framework. The authors creatively adapt ecological niche modeling (Peterson 2001) to predict the geo- graphic distribution of land use types based on the ecological distribution of climatic and soil condi- tions of a sample of sites of known land use, each site being a grid cell. In other words, knowing the ecological conditions of some places a land use is practiced, they predict where else it is likely to be practiced based on the presence of those condi- tions. The sample of land use types are records of locations, converted to grid cells, where that land use is currently practiced by a human society. They attempted to avoid spatial autocorrelation by using only a few records from each society. The authors chose a maximum entropy (Maxent) variety of ecological niche model, often generally termed a species distribution model (SDM). Maxent estimates the most uniform, or entropic, geographic distribution of land uses sub- ject to the constraint that the distribution of cli- matic and soil variable values among them must equal its empirical average from the sample sites (see Phillips et al. 2006). To test their model’s accuracy, they then a) compared its predictions of different human land use “niches” with actual presence records drawn from a variety of published sources, b) compared their model predictions to U.N. Food & Agricul- tural Organization records of actual agricultural practice, c) calculated an “index of shared (land use) suitability,” and compared predictions with a database of post-1945 conflict data, and d) calcu- lated regressions of the suitability for agriculture vs. log population density and suitability for agri- culture vs. GDP-PPP, an index of economic strength. Their findings support the hypothesis that the geographic template affects basic economic and political patterns of modern humans. The model was most accurate for predicting the pres- ence of sedentary animal husbandry and least ac- curate for nomadic pastoralists, but it was fairly strong in general. It modestly predicted popula- frontiers of biogeography 2.3, 2010 — © 2010 the authors; journal compilation © 2010 The International Biogeography Society