Research paper Aboveground tree additive biomass models in Ecuadorian highland agroforestry systems Jos  e Riofrío a, * , Celia Herrero a, b , Jorge Grijalva c, d , Felipe Bravo a, e a Sustainable Forest Management Research Institute, Universidad de Valladolid & INIA, Av. Madrid 44, 34004 Palencia, Spain b ECM Ingeniería Ambiental S.L., c/Manuel Martinez de Azcoitia 35, 34003 Palencia, Spain c Facultad de Medicina Veterinaria y Zootecnia, Universidad Central del Ecuador, Ecuador d Programa Nacional de Forestería, Instituto Nacional Aut onomo de Investigaciones Agropecuarias, Panamericana Sur km1, Quito, Ecuador e Departamento de Producci on Vegetal y Recursos Forestales, E.T.S de Ingenierías Agrarias, Universidad de Valladolid, Campus, Palencia, Spain article info Article history: Received 30 May 2014 Received in revised form 13 May 2015 Accepted 27 May 2015 Available online xxx Keywords: Andean species Multi-stemmed species Allometric models Weighted regression SUR abstract Agroforestry land-use systems in the Andean region have great socioeconomical and biophysical rele- vance due to the abundance of products and services they provide. Biomass estimation in these systems constitutes a priority concern as it facilitates assessment of carbon sink potential and functionality for biomass production. In this paper, a set of equations were fitted to enable easy and reliable estimation of the total aboveground biomass of four frequently used species in Andean agroforestry systems: Acacia melanoxylon L., Alnus acuminata Kunth., Buddleja coriacea Remy. and Polylepis racemosa Ruiz&Pav. The best models for each biomass component (stem, thick branches, thin branches and leaves) per species were fitted simultaneously according to SUR methodology (seemingly unrelated regressions). All models showed high goodness of fit statistics and more than 70% of the observed variation in biomass com- ponents was explained by the independent variables. The inclusion of height as a predictive variable in the models improved their predictive reliability and expanded the application range. The models developed here are useful for assessing the sustainability of agroforestry systems and could support governmental or non-governmental forest conservation incentive programs and initiatives. © 2015 Elsevier Ltd. All rights reserved. 1. Introduction In the Andean region, agroforestry strategy integrates produc- tion and conservation in order to diversify products and services, optimize resources and reduce risks. It generates income, ensures food security and favors the preservation of biodiversity [1,2]. Numerous initiatives to promote soil conservation and agricultural sustainability on steeply sloping lands have been implemented by governmental and non-governmental institutions in the Andean region. They mainly focus on wood productivity, natural resource management (biodiversity, watershed management and soil con- servation) or multiple land use through agroforestry systems (AFS) [3]. However, the lack of basic research and an appropriate approach to the biophysical and socio-economic circumstances of farmers continues to undermine technical proposals [3,4]. There is a clear need for more complete knowledge regarding structural and functional interactions among AFS components. Sustainability assessment through indicators that quantify the magnitude of the positive or negative effects of AFS has been suc- cessfully implemented in the region [5]. A set of sustainability in- dicators related to the resource base, the function of the system itself and its impact on other systems has made possible to evaluate whether a given system is sustainable or not [6]. In this framework, priority is given to developing efficient methods for estimating biomass and carbon (C) indicators that are easily applicable across AFS, in order to evaluate carbon sink potential and improve func- tional analysis. Carbon is quantified in the tree biomass of forest ecosystems by estimating the total amount of living material. Allometric equations can relate biomass to easily-measured tree level variables such as diameter at breast height (dbh) and total height (ht). On a local or site scale, this is the most practical option for assessing above- ground biomass (AGB) from inventory data. Some data for AGB and components are available from allome- tric equations previously developed for Andean species [7e9]. * Corresponding author. E.T.S. de Ingenierıas Agrarias, Universidad de Valladolid, Campus de Palencia, Av. Madrid 44, C.P. 34004, Spain. E-mail address: joseguillermo.riofrio@alumnos.uva.es (J. Riofrío). Contents lists available at ScienceDirect Biomass and Bioenergy journal homepage: http://www.elsevier.com/locate/biombioe http://dx.doi.org/10.1016/j.biombioe.2015.05.026 0961-9534/© 2015 Elsevier Ltd. All rights reserved. Biomass and Bioenergy 80 (2015) 252e259