19 Chapter 19 Long-Term Geological Evolution and Mass-Flow Balance of the South-Central Andes Johannes Glodny · Helmut Echtler · Oscar Figueroa · Gerhard Franz · Kirsten Gräfe · Helga Kemnitz · Wolfgang Kramer Charlotte Krawczyk · Jo Lohrmann · Friedrich Lucassen · Daniel Melnick · Matthias Rosenau · Wolfgang Seifert Whereas oceanic crust recycling into the mantle is in- herent in the concept of plate tectonics, and has geochemi- cally been proven (Hofmann and White 1980; Sobolev et al. 2005), destruction of the continental crust and net mass loss has been demonstrated at convergent margins using mass balance studies (von Huene and Scholl 1991). A very influential hypothesis for continental growth is the ‘andesite model’, introduced by Taylor (1967), which sug- gests significant horizontal and volumetric growth by the input of andesitic arc material at active continental mar- gins. This concept has later been modified (Taylor and McLennan 1985), largely dismissed (Reymer and Schubert 1984), or restricted to ridge-subduction settings (Defant and Drummond 1990; Drummond and Defant 1990). However, the general idea of regarding active margins as sites of continental growth appears to be persistent (cf. Clift and Vannucchi 2004). Recent studies have shown that active margins can episodically, within less than a few million years, shift between modes of accretion and tectonic erosion (Bangs and Cande 1997). However, the short-term net effects of convergence are difficult to assess as continent-derived sediment is lost into the mantle even during accretion. Geological indications for dominance of either continen- tal growth or tectonic erosion are sparse. They comprise mainly of relative shifts in the position of the magmatic arc or of fixed points of the fore-arc, compared to the trench (e.g., Mpodozis and Ramos 1989; Meschede et al. 1999; Clift and Vannucchi 2004), or of basin formation by fore-arc subsidence caused by tectonic erosion (von Huene and Lallemand 1990; Laursen et al. 2002; Wells et al. 2003). Other convergence-related tectonic effects at active mar- gins are difficult to interpret in terms of crustal growth. In particular, the emergence of morphological fore-arc highs and their extensional deformation may occur at both accretive and erosive margins (von Huene and Scholl 1991; Hartley et al. 2000). Long-term, average rates of mass flux (over the 100 Myr timescale) are more relevant for estimating the overall behaviour of a margin segment than short-term events and fluctuations. Therefore, when assessing the relevance of active, continental margin systems for crustal growth, Abstract. In south-central Chile (36–42° S), the western edge of South America has evolved as an active margin since the Pennsylvanian (~305 Ma). Active margins are considered as sites of both potential continental growth and continental destruction. Continental growth in a margin setting can proceed by accretionary offscraping of juvenile material from the oceanic plate and by magmatic addi- tions, whereas net mass loss can be achieved by subducting conti- nental material, delamination, and chemical weathering. In south- central Chile, margin evolution was never interrupted by island-arc accretion or continental collision. Thus, the area provides an excel- lent field laboratory for studying mass flux through a long-term, persistent, convergent-margin system. Using new isotopic age data, we summarize the current knowl- edge of the geological evolution of the south-central Chilean mar- gin, from subduction initiation to the ongoing Andean morphotec- tonic processes, with emphasis on mechanisms of mass transfer. It is inferred that net crustal growth and mass losses alternated in time and space, and that dominance of one of the other process might have even occurred contemporaneously within a short distance along the same margin, controlled by factors such as sediment avail- ability in the trench, lower-plate morphology, upper-plate tecton- ics, and climate. In south-central Chile, the margin north of 38° S is characterized by a landward trench migration of ~100 km that occured mainly in the early Permian, whereas farther south, the modern and the late Palaeozoic magmatic arcs are superimposed. For most of its life- time, the margin evolved in a delicate balance between construc- tive and destructive processes. Over the long term, the south-cen- tral Chilean continental margin has not been a site of net growth, but, rather, a site of continental mass wasting, crustal recycling and crustal rejuvenation. 19.1 Introduction How, and how fast, continental crust grows is of ongoing interest in the world of geodynamics. Usually, concepts are developed by identifying particular growth mecha- nisms, such as tectonic accretion of igneous oceanic crust, various mechanisms for adding basaltic plume material, and continent versus juvenile arc collision, and then com- bining them with geochemical, isotope-geochemical and geochronological inferences (cf. Albarède 1998). Most growth concepts imply an overall increase in the total volume of continental crust with time (Hurley and Rand 1969; De Paolo and Wasserburg 1979; Reymer and Schu- bert 1984; see compilation in Tarney and Jones 1994).