Forest Ecology and Management 494 (2021) 119341 0378-1127/© 2021 Elsevier B.V. All rights reserved. Biotic and abiotic drivers of carbon, nitrogen and phosphorus stocks in a temperate rainforest Jorge F. Perez-Quezada a, b, * , Cecilia A. P´ erez b , Carla E. Brito a , Juan P. Fuentes c , Aurora Gaxiola b, d , David Aguilera-Riquelme a , Javier Lopatin a, e, 1 a Department of Environmental Science and Renewable Natural Resources, University of Chile, Avenida Santa Rosa 11315, Santiago, Chile b Institute of Ecology and Biodiversity, Alameda 340, Santiago, Chile c Department of Silviculture and Nature Conservation, University of Chile, Avenida Santa Rosa 11315, Santiago, Chile d Department of Ecology, Pontifcia Universidad Cat´ olica de Chile, Alameda 340, Santiago, Chile e Center for Climate Resilience Research (CR) 2 , University of Chile, 8370449 Santiago, Chile A R T I C L E INFO Keywords: Biomass Deadwood Patagonia Evergreen broadleaf forest Nutrients Vines ABSTRACT Forest ecosystems are recognized for their large capacity to store carbon (C) in their aboveground and below- ground biomass and soil pools. While the distribution of C among ecosystem pools has been extensively studied, less is known about nitrogen (N) and phosphorus (P) pools and how these stocks relate to each other. There is also a need to understand how biotic and abiotic ecosystem properties drive the magnitude and distribution of C- N-P stocks. We studied a temperate rainforest in southern South America to answer the following questions: 1) how are C-N-P total stocks distributed among the different ecosystem pools?, 2) how do C:N, C:P and N:P ratios vary among ecosystem pools?, and 3) which are the main biotic and abiotic drivers of C-N-P stocks? We established 33 circular plots to estimate C, N, and P stocks in different pools (i.e. trees, epiphytes, understory, necromass, leaf litter, and soil) and a set of biotic (e.g., tree density and richness) and abiotic variables (e.g., air temperature, humidity and soil depth). We used structural equation modeling to identify the relative importance of environmental drivers on C-N-P stocks. We found that total ecosystem stocks (mean ± SE) were 1062 ± 58 Mg C ha 1 , 28.8 ± 1.5 Mg N ha 1 , and 347 ± 12.5 kg P ha 1 . The soil was the largest ecosystem pool, containing 68%, 92%, and 73% of the total C, N, and P stocks, respectively. Compared to representative temperate forests, the soil of this forest contains the largest concentrations and stocks of C and N. The low P stock and wide soil C:P and N:P ratios suggest that P may be limiting forest productivity. The ecosystem C-N-P stocks were mainly driven by abiotic properties measured in the study area, however for N stocks, variables such as plant diversity and canopy openness were also relevant. Our results provide evidence about the importance not only of under- standing the differences in C, N, and P stocks but also of the factors that drive such differences. This is key to inform conservation policies related to preserving old-growth forests in southern South America, which indeed are facing a rapid land-use change process. 1. Introduction The study of ecosystem pools of chemical elements is key for modelling climate change because these pools determine ecosystem productivity and the exchange of greenhouse gases (GHG) between terrestrial ecosystems and the atmosphere. Forests are major contribu- tors to several important biosphere processes (Weathers et al. 2013), including the carbon (C), nitrogen (N), and phosphorus (P) biogeo- chemical cycles. It is estimated that forests contain about 80% of global aboveground C (IPCC 2007), around 40% of belowground terrestrial C (IPCC 2007), and fx around 10% of atmospheric CO 2 each year (Waring and Running 2007). Temperate forests are known to have lower C allocated to vegetation but greater C allocated to soils (62%) than tropical forests (IPCC 2000). Carbon stocks in soils are constrained by a complex interaction of two alternative C fuxes. On the one hand, increasing CO 2 concentrations in the atmosphere increases C seques- tration in the soil, determined by an increase in primary productivity, although this process is very sensitive to the age of soil C (He et al. * Corresponding author. E-mail address: jorgepq@uchile.cl (J.F. Perez-Quezada). 1 Present address: Faculty of Engineering and Science, Universidad Adolfo Ib´ a˜ nez, Santiago, Chile. Contents lists available at ScienceDirect Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco https://doi.org/10.1016/j.foreco.2021.119341 Received 28 November 2020; Received in revised form 29 April 2021; Accepted 4 May 2021