Edaphic Factors are a More Important Control on Surface Fine Roots than Stand Age in Secondary Tropical Dry Forests Jennifer S. Powers 1,2,3,4 and Daniel Pere ´ z-Aviles 1 1 Department of Ecology, Evolution and Behavior, University of Minnesota, 100 Ecology Building, 1987 Upper Buford Circle, St. Paul, MN, 55108, U.S.A. 2 Department of Plant Biology, University of Minnesota, St. Paul, MN, 55108, U.S.A. 3 Smithsonian Tropical Research Institute, Apartado, Balboa, 2072, Panama ABSTRACT Although there are generalized conceptual models that predict how above and belowground biomass increase during secondary succes- sion after abandonment from agriculture, there are few data to test these models for fine roots (defined as  2 mm diameter) in tropi- cal forests. We measured live and dead fine roots (0–10 cm depth) in 18 plots of regenerating tropical dry forest in Costa Rica that varied in age from 5 to 60 yrs, as well as in soil properties. We predicted that both stand age and soil fertility would affect fine roots, with greater values in older forests on low fertility soils. Across two sampling dates and locations, live fine roots varied from 0.35 to 3.53 Mg/ha and dead roots varied from 0.15 to 0.93 Mg/ha. Surprisingly, there was little evidence that surface fine roots varied between sampling dates or in relation to stand age. By contrast, total, live, and dead fine roots averaged across sampling dates within plots were negatively correlated with a multivariate index of soil fertility (Pearson correlations coefficients were 0.64, 0.58, and 0.68, respectively; P < 0.01) and other individual edaphic variables including pH, silt, calcium, magnesium, nitrogen, and phosphorus. These results suggest that soil fertility is a more important determinant of fine roots than forest age in tropical dry forests in Costa Rica, and that one-way these plant communities respond to low soil fertility is by increasing fine roots. Thus, simple conceptual models of forest responses to abandonment from agriculture may not be appropriate for surface fine roots. Abstract in Spanish is available in the online version of this article. Key words : belowground biomass; Costa Rica; secondary succession; soil nutrients. ONE OF THE MOST PERVASIVE CHANGES THAT TROPICAL LANDSCAPES ARE CURRENTLY UNDERGOING is forest regeneration following abandonment from agriculture (Brown & Lugo 1990). Under- standing the extent to which species composition, structure, and function of secondary forests resemble those of mature forests is a major challenge for tropical ecology (Chazdon 2003, Wright & Muller-Landau 2006). It is commonly assumed that both above and belowground biomass decrease rapidly following deforesta- tion or forest clearing and then increase when tree-based ecosys- tems reestablish (Greenland & Nye 1959, Melillo et al. 1996). This conceptual model may work well for aboveground stocks, such as stem biomass (Brown & Lugo 1990, Melillo et al. 1996, Silver et al. 2000), but it is less clear whether this applies to belowground components of ecosystems including soil organic carbon or fine roots. Such discrepancies in successional patterns between above and belowground biomass pools may result in part because residual soil organic matter (SOM) and fine roots may persist throughout transitions between dominant vegetation types (Castellanos et al. 2001). If so, total belowground stocks during old-field succession will consist of a mixture of pasture- and forest-derived sources. Alternatively, belowground stocks may be more affected by drivers other than land use, such as soil mineralogy (Lopez-Ulloa et al. 2005, Powers et al. 2011) or soil structure. Fine roots, commonly defined as roots  2 mm diameter, are important for water and nutrient acquisition (Hendricks et al. 1993), provide a major source of carbon to the soil (Gale & Cambardella 2000), and may comprise a variable, but large per- centage of net primary production (Vogt et al. 1996) (throughout the text we refer to quantities or stocks of fine root mass simply as fine roots). Measurements of fine roots at any one point in time reflect the sum of a number of dynamic processes including root production, biotic interactions (positive or negative interac- tions with herbivores, pathogens, and/or symbionts), senescence, and decomposition. Even though static measurements of root stocks do not account for these dynamic processes, much pro- gress has been made during the last few decades in identifying the coarse-scale controls on fine roots in unmanaged ecosystems, and consistent patterns are emerging. Across ecosystems fine roots are correlated with biomass (Cairns et al. 1997), basal area (Finér et al. 2011), climate (Yuan & Chen 2010), and soil fertility (Vito- usek & Sanford 1986). In general, among ecosystems soil fertility appears to be well correlated with fine roots, which tend to increase with decreasing nutrient availability (Powers et al. 2005, Jimenez et al. 2009), either because of increased allocation belowground on less fertile soils or constant allocation to fine roots across soil fertility gradients, coupled with faster root Received 16 September 2011; revision accepted 20 March 2012. 4 Corresponding author; e-mail: powers@umn.edu ª 2012 The Author(s) 1 Journal compilation ª 2012 by The Association for Tropical Biology and Conservation BIOTROPICA 0(0): 1–9 2012 10.1111/j.1744-7429.2012.00881.x