Contents lists available at ScienceDirect Urban Forestry & Urban Greening journal homepage: www.elsevier.com/locate/ufug Decoupling of the urban vegetation productivity from climate Leonardo Paolini a,b, ⁎ , Luitgard Schwendenmann a , Ezequiel Aráoz b,c , Priscila Ana Powell c a School of Environment, The University of Auckland, Auckland, New Zealand b Instituto de Ecología Regional (IER), Universidad Nacional de Tucumán (UNT)- Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Tucumán, Argentina c Facultad de Ciencias Naturales e Instituto Miguel Lillo, Universidad Nacional de Tucumán (UNT), Tucumán, Argentina ARTICLE INFO Handling Editor: Wendy Chen Keywords: Urban ecosystems Modis Primary productivity Time series analysis Timesat ABSTRACT Urbanization is causing profound alterations of ecosystem functions at local and regional scales. The need to maximize ecosystem services to improve the livability of cities is resulting in intensive management of urban vegetation, which is likely generating new conditions for the ecosystem functioning. In this paper, we address the association between vegetation functioning (i.e., primary productivity) and climate in urban and adjacent non-urban areas in Auckland, New Zealand. We used time series analysis of integrated vegetation indices derived from satellite images (MODIS) to estimate a proxy of primary productivity of urban and non-urban vegetation. We analyzed the interannual variability of vegetation productivity in relation to climate fluctuations. In Auckland’s urban area the variability of primary productivity was not associated with any of the climatic variables considered, while in the non-urban area (i.e. reference area) vegetation productivity was strongly associated with cumulative rainfall during the growing season. Our results suggest that the productivity of urban vegetation may be undergoing a decoupling from the regional climate. If a decoupling of ecosystem functions from climate becomes a general pattern in urban areas it could have significant effects on urban vegetation planning and management. In a context of increased variability, urban ecosystems could constitute stable ha- bitats and they will probably contribute to the viability of vulnerable populations. 1. Introduction Urbanization is one of the most extreme and persistent forms of alteration of the earth surface (McKinney, 2006; Palomino and Carrascal, 2006; Shochat et al., 2006, Hasse et al. 2014, Alberti et al., 2017; Eigenbrod et al., 2011). From a global perspective, urbanization may present a turning point for life on the planet, with yet unknown consequences on ecosystems functions (Alberti, 2015). With 75% of the human population predicted to live in cities by 2050 (United Nations Report, 2014), the urban land cover is forecast to increase by 1.2 mil- lion km 2 , almost tripling the global urban land area compared to 2000 (Seto et al., 2012). Although urban areas cover less than 5% of Earth’s land surface, they consume over 60% of the world's energy and account for more than 70% of CO 2 emissions (Churkina 2016), which alters climate (Gioia et al., 2014; Kotthaus and Grimmond, 2014), biodi- versity (Haedo et al., 2017), and ecosystem processes (Pickett et al., 2001; Alberti et al., 2008; Matthews et al., 2011; Paolini, 2012; Alberti, 2015; Paolini et al., 2016). Urban ecosystems are characterized by new, unique and complex interactions between built-up areas, social processes, natural elements and climate. These interactions define the structure and species com- position, as well as phenology and productivity of urban vegetation. The particular characteristics that distinguish urban ecosystems, con- trol photosynthetic activity and hence affect primary productivity of the vegetation (Imhoff et al., 2004) and could shape the responses and resilience of urban ecosystems to changes (Alberti et al., 2017). Gross Primary Productivity (GPP), defined as the total amount of CO 2 fixed by plants through photosynthesis (Gitelson et al., 2006, Running et al., 2004) mediates the relationship between land cover change and eco- system function (e.g. richness of both fauna and flora species) and is a good proxy for an important set of ecosystem services (e.g., climate regulation, biodiversity, etc.) (Alberti, 2015). Although this relation- ship varies with taxa and scale, and across biomes (Mittelbach et al., 2001; Haedo et al., 2017), GPP analyses are essential to understand ecosystem dynamics especially in urban areas where human interven- tion plays a central role. Phenology is one of the most important vari- ables to explain interannual variability of GPP (Fu et al., 2013, Zhang et al., 2014). Therefore, modelling phenological and GPP variations will https://doi.org/10.1016/j.ufug.2019.126428 Received 11 December 2018; Received in revised form 12 June 2019; Accepted 7 August 2019 ⁎ Corresponding author at: Science Centre, 302 - Bldg 302. 23 Symonds St, Auckland 1010. New Zealand. E-mail addresses: l.paolini@auckland.ac.nz (L. Paolini), l.schwendenmann@auckland.ac.nz (L. Schwendenmann), ezequielaraoz@gmail.com (E. Aráoz), priscilaapowell@gmail.com (P.A. Powell). Urban Forestry & Urban Greening 44 (2019) 126428 Available online 10 August 2019 1618-8667/ © 2019 Elsevier GmbH. All rights reserved. T