Agricultural potential as an indicator for 21st century land use change across South America?s ecoregions

LUCÍA ZARBÁ; HECTOR RICARDO GRAU; NESTOR IGNACIO GASPARRI; JORDAN GRAESSER; AIDE, T. MITCHELL

Posadas

Congreso; IUFRO CONFERENCE - Adaptive Management for Forested Landscapes in Transformation; 2018

Cropland and pastureland are expanding throughout Latin America, often replacing forestecosystems. Commodity production for export is proposed as one of the main drivers ofthis expansion, and indirectly these changes induce other land use changes (e.g. low profitagriculture abandonment, rural-urban migration). Previous studies found a segregatedpattern of expansion and reorganization of the different land uses across the continent,suggesting that geography plays an important role in these dynamics. Foresee whichecoregions move in each direction is important for contextualizing policy and planningefforts to maximize land use efficiency. In this study, we created a map that tried tocapture the agribusiness investor viewpoint as a major agent of change. Specifically, wewere interested in which ecoregions are more likely to undergo similar transformationdynamics and what type of agricultural activities may occur in the different ecosystems?To address this questions we proposed a typification of agricultural potential as anindicator of the overall land use change processes expected in each ecoregion based onthree attributes: (i) aptitude for mechanized agriculture, (ii) aptitude for rain-fed agriculture,and (iii) distance to consumption/distribution centers. In addition, alternative thresholdswith increasing level of tolerance were explored. We grouped the ecoregions throughcluster analysis, repeating the analysis for all combination of thresholds (n=48). Finally weanalyzed the sensitivity to each attribute and the correlation between the clusters andcropland cover area and its change between 2000-2014. Results as of now showed thatmaps with more flexible thresholds correlated better with patterns of cropland area.Correlation with cropland area showed big sensitivity to accessibility. In general patternscluster analyses found 4 big groups: mountain, well connected intermediate humidity andrelief, flat dry isolated, and flat wet well connected ecoregions. Ecosystems in the latter arethe most vulnerable to future agriculture expansion.