Biomass and Bioenergy 136 (2020) 105526 Available online 14 March 2020 0961-9534/© 2020 Elsevier Ltd. All rights reserved. Research paper Impact of nitrogen and phosphorous on biomass yield, nitrogen effciency, and nutrient removal of perennial grasses for bioenergy Guillermo Siri-Prieto a, * , Mauricio Bustamante a , Valentín Picasso b , Oswaldo Ernst a a Agronomy Faculty, Universidad de la República, Uruguay b Agronomy Department, University of Wisconsin, Madison, USA A R T I C L E INFO Keywords: Giant reed Elephantgrass Switchgrass Biomass yield Fertilization Nitrogen use effciency ABSTRACT Perennial grasses are the promising source of bioenergy in South America which could provide several envi- ronmental benefts such as reduction in the greenhouse gasses emissions and reduction of nutrients and soil losses. Our objective was to determine the impact of N and P fertilization on biomass yield, N use effciency (NUE), apparent N recovery (ANR), and nutrient removal (NR) on three perennial grasses: elephantgrass (Pen- nisetum purpureum Schum.), giant reed (Arundo donax L.), and switchgrass (Panicum virgatum L.). Four fertility treatments were evaluated in a 8-year feld study in northwestern Uruguay: 1. control (No fertilizer), 2.100 kg N ha 1 year 1 , 3.100 kg P 2 O 5 ha 1 year 1 , and 4.100 kg N þ 100 kg P 2 O 5 ha 1 year 1 . Across years, elephantgrass had the highest biomass yield followed by giant reed and switchgrass (18.9, 16.3, and 14.1 Mg ha 1 , respec- tively). Biomass yield increased 46% when N fertilizer was added, compared to the control. A low response was detected for P fertilization on all grasses, probably for initial P soil content (�9 g kg 1 ). Elephantgrass had the highest NUE (70 kg kg 1 N), however, it had the highest total NR on these eight years (899, 226, and 2800 kg ha 1 for N–P–K, respectively) among the grasses, indicating a potential for increased fertilization input over time. Switchgrass had the lowest average ANR (19%) and NR (334, 45, and 166 kg ha 1 , respectively). There- fore, even though switchgrass presented the lowest biomass yield, it is an excellent option as low-input bio- energies grass for temperate regions. 1. Introduction The increasing demand and prices for oil, and the competition be- tween fuel and food production in land suitable for crop production are encouraging the production of perennial grasses for bioenergy uses in marginal lands [1]. These energy grasses can be used for the production of electricity, biofuels, and biogas. Besides, these grasses are more ecologically suitable than annual crops for their higher total biomass per unit area, lower establishment costs, reduced soil erosion, and increased water quality [2]. Perennial grasses have been proposed as key bio- energy crops in Europe and the US, based on their low management costs in fertility and pest management and high productivity in a diverse set of edaphoclimatic conditions [1,3]. Elephantgrass (EG) is a perennial C 4 grass native to Africa, which can produce high biomass yield per unit area under different soil and weather conditions [4,5]. Moreover, it has a high energy potential, and it has been traditionally used for animal feed [6,7]. Giant reed (GR) is a perennial rhizomatous C 3 grass, native to East Asia, and considered one of the most promising bioenergy species for Southern Europe [8], for its adaptability to different environmental conditions, quick growth and high biomass productivity [1,9–11]. Switchgrass (SW), a C 4 grass native to North America, has shown medium-high biomass productivities for potential feedstock, either for producing ethanol or electricity via co-fring [2,12]. Additionally, recent long-term (i.e., >10 yr) multi-site trials conducted in Oklahoma, New York, South Dakota and Virginia in the US across different landscape positions, indicate that switchgrass production could maintain or improve overall soil quality, increase carbon sequestration and reduce the risk of N loss to the environment [13,14]. Fertilizer use effciency, specially that of N, is an important factor determining both the selection of suitable dedicated energy crops to be grown, and their potential biomass productivity [15,16]. For EG, most of the studies indicated a strong and positive biomass productivity response to N application [9,17,18], but a non signifcance response to P application [19]. Reported fndings on the GR biomass yield response to N and P application have been contrasting. Working in Central Italy, Angelini et al. [9] reported a GR biomass yield increase of 4 Mg ha 1 * Corresponding author. E-mail address: gsiri46@gmail.com (G. Siri-Prieto). Contents lists available at ScienceDirect Biomass and Bioenergy journal homepage: http://www.elsevier.com/locate/biombioe https://doi.org/10.1016/j.biombioe.2020.105526 Received 3 September 2019; Received in revised form 10 February 2020; Accepted 25 February 2020