Contents lists available at ScienceDirect Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco Forest diversity plays a key role in determining the stand carbon stocks of Mexican forests Ricard Arasa-Gisbert a,1, ⁎ , Jordi Vayreda a , Rosa María Román-Cuesta a,2 , Sergio Armando Villela b , Rafael Mayorga b , Javier Retana a,c a CREAF, Campus de Bellaterra (UAB) Edifici C, Cerdanyola del Valles 08193, Spain b Comision Nacional Forestal, Periférico Poniente No. 5360, Zapopan 45019, Mexico c Universitat Autonoma Barcelona, Cerdanyola del Valles 08193, Spain ARTICLE INFO Keywords: Carbon stock Climate drivers Disturbance factors Mexico Species richness Structural richness ABSTRACT Defining the most important factors related with forest carbon (C) stocks in different forest types is still a controversial topic. In this study we used data from 10,500 plots from The National Forest Inventory of Mexico encompassing the six main forest types in Mexico (conifer, broadleaf, mixed, evergreen and semi-deciduous, dry and semiarid forests) in order to identify the main factors related to the spatial pattern of C stocks, including climate (temperature and rainfall), forest diversity (structural and species richness), topographic and soil characteristics (soil depth, slope and land tenure) and disturbance factors (fires, pests and tree felling). We built two different types of models, one taking all plots into account (global model, R 2 = 0.54, P < 0.001) and others for each forest type separately. Overall, structural richness was the most important variable related to C stock both in the global model and in each forest type model. Tree richness had a strong relationship in tropical forests (both dry and evergreen) but not in temperate forests (conifer, broadleaf and mixed forests), where slope and climate variables had greater effects on C stocks. C stock was strongly and positively correlated with pre- cipitation in almost all forest types, while it was strongly and negatively correlated with temperature in broadleaf and mixed forests. Surprisingly, slope was the second most important factor positively correlated with C stock in broadleaf and mixed forests. Surprisingly, soil depth, land tenure and disturbance variables had a negligible effect in almost all models, partially due to the poor quality of disturbance and soil depth data available from INFyS. The results suggest that, in order to enhance C stock in Mexican forests, management techniques should encourage increases of the number of tree species and, especially, tree size inequality, since both these factors were shown to have a key role in C stock. 1. Introduction Forests are one of the largest reservoirs of organic carbon (C) in the medium and long term (Houghton, 2005; Peichl and Arain, 2007; Canadell and Raupach, 2008). Therefore, a better knowledge of the distribution patterns and variability of both aboveground and below- ground C stocks is of crucial importance for understanding how C stock changes through time (Houghton, 2005) and for increasing the effi- ciency and C stock of forest ecosystems (Zhao and Zhou, 2006). C sinks in forests are determined by processes that either allow C sequestration (photosynthesis, C accumulation in soils, tree growth) or induce C losses (disturbances, respiration, tree mortality, microbial decomposi- tion of litter; Malhi et al., 1999). These processes are influenced by many climatic, biological, topographic, and site factors, and even small alterations in these processes can alter the whole C cycle (Malhi et al., 1999). When considering large spatial scales, environmental factors are the main variables influencing C stocks and C dynamics (Keeling and Phillips, 2007; Vayreda et al., 2012). Specifically, temperature, pre- cipitation and light availability are the most determinant variables for forest growth (Malhi et al., 1999). Temperature has been described as the most determinant factor for C stocks (Raich et al., 2006; Zhao and https://doi.org/10.1016/j.foreco.2018.02.023 Received 9 November 2017; Received in revised form 12 February 2018; Accepted 13 February 2018 ⁎ Corresponding author at: Instituto de Investigaciones en Ecosistemas y Sustentabilidad, Universidad Nacional Autonoma de Mexico (UNAM), Antigua Carretera a Patzcuaro No. 8701, Ex-Hacienda de San José de la Huerta, 58190 Morelia, Michoacan, Mexico. 1 Present address: Instituto de Investigaciones en Ecosistemas y Sustentabilidad, Universidad Nacional Autonoma de Mexico (UNAM), Antigua Carretera a Patzcuaro No. 8701, Ex- Hacienda de San José de la Huerta, 58190 Morelia, Michoacan, Mexico. 2 Present address: Jalan CIFOR, Situ Gede, 16115 Bogor, Barat, Indonesia. E-mail addresses: rarasa@cieco.unam.mx (R. Arasa-Gisbert), j.vayreda@creaf.uab.cat (J. Vayreda), R.Roman-Cuesta@cgiar.org (R.M. Román-Cuesta), svillela@conafor.gob.mx (S.A. Villela), Rafaelmayorga.ute@conafor.gob.mx (R. Mayorga), Javier.Retana@uab.cat (J. Retana). Forest Ecology and Management 415–416 (2018) 160–171 0378-1127/ © 2018 Elsevier B.V. All rights reserved. T