CAMILA DA SILVA GRASSMANN MAIZE-BASED SYSTEMS AS AFFECTED BY FORAGE GRASS AND NITROGEN FERTILIZATION: ELUCIDATING 15N-FERTILIZER RECOVERY, GREENHOUSE GAS EMISSIONS, N-CYCLE FUNCTIONAL GENES IN SOIL AND CROP YIELDS Botucatu 2021 CAMILA DA SILVA GRASSMANN MAIZE-BASED SYSTEMS AS AFFECTED BY FORAGE GRASS AND NITROGEN FERTILIZATION: ELUCIDATING 15N-FERTILIZER RECOVERY, GREENHOUSE GAS EMISSIONS, N-CYCLE FUNCTIONAL GENES IN SOIL AND CROP YIELDS Doctoral thesis submitted to the College of Agricultural Sciences, UNESP - Botucatu Campus, to obtain the degree of Doctor in Agriculture. Advisor: Prof. Dr. Ciro Antonio Rosolem Co-advisor: Dr. Eduardo Mariano Botucatu 2021 G769m Grassmann, Camila da Silva Maize-based systems as affected by forage grass and nitrogen fertilization: Elucidating 15N recovery, greenhouse gas emissions, N-cycle functional genes in soil and crop yields / Camila da Silva Grassmann. -- Botucatu, 2021 159 p. : il., tabs. Tese (doutorado) - Universidade Estadual Paulista (Unesp), Faculdade de Ciências Agronômicas, Botucatu Orientador: Ciro Antonio Rosolem Coorientador: Eduardo Mariano 1. Soil fertility. 2. Nitrogen cycle. 3. Forage plants Soils. 4. Greenhouse gases. I. Título. Sistema de geração automática de fichas catalográficas da Unesp. Biblioteca da Faculdade de Ciências Agronômicas, Botucatu. Dados fornecidos pelo autor(a). Essa ficha não pode ser modificada. TÍTULO DA TESE: UNIVERSIDADE ESTADUAL PAULISTA Câmpus de Botucatu CERTIFICADO DE APROVAÇÃO MAIZE-BASED SYSTEMS AS AFFECTED BY FORAGE GRASS AND NITROGEN FERTILIZATION: ELUCIDATING 15N-FERTILIZER RECOVERY, GREENHOUSE GAS EMISSIONS, N-CYCLE FUNCTIONAL GENES IN SOIL AND CROP YIELDS AUTORA: CAMILA DA SILVA GRASSMANN ORIENTADOR: CIRO ANTONIO ROSOLEM COORIENTADOR: EDUARDO MARIANO Aprovada como parte das exigências para obtenção do Título de Doutora em AGRONOMIA (AGRICULTURA), pela Comissão Examinadora: Prof. Dr. CIRO ANTONIO ROSOLEM (participação Virtual) Produção Vegetal / Faculdade de Ciências Agronômicas de Botucatu - UNESP Pesquisador Dr. HEITOR CANTARELLA (participação Virtual) Centro de Pesquisa e Desenvolvimento de Solos e Recursos Ambientais / Instituto Agronômico de Campinas Prof. Dr. RAFAEL OTTO (participação Virtual) Ciência do Solo / Escola Superior de Agricultura Luiz de Queiroz Prof. Dr. PAULO SÉRGIO PAVINATO (participação Virtual) Departamento de Ciência do Solo / ESCOLA SUPERIOR DE AGRICULTURA PROF. DR. GUSTAVO CASTOLDI (Participação Virtual), POLO DE INOVAÇÃO / INSTITUTO FEDERAL GOIANO Botucatu, 29 de janeiro de 2021 Faculdade de Ciências Agronômicas - Câmpus de Botucatu - Avenida Universitária, , 3780, 18610034, Botucatu - São Paulo I dedicate my dissertation work to my family and friends. A special feeling of gratitude to my loving parents Carlos and Sandra Grassmann whose words of encouragement and unconditional love was essential. My special brother Guilherme, who never left my side. I also dedicate this dissertation to my boyfriend, for supporting me, understanding and loving me. ACKNOWLEDGEMENTS Firstly, and always to God, for health and strength to work, and enlightenment to discern between right and wrong. Prof. Dr. Ciro Antonio Rosolem, for his friendship, advice, teaching, trust, partnership and excellent support in these almost 8 years of orientation that started in the undergrad. I will be forever grateful for the opportunity granted. Prof. Dr. Karl Ritz, from the University of Nottingham, for the excellent guidance and teaching, as well as friendship. Dr. Eduardo Mariano, who I feel honored and grateful to have had as a co-supervisor. I will be forever grateful for so much learning and friendship. Prof. Dr. Paulo Roberto Miranda Meirelles and Dr. André Michel de Castilhos, for the support in analysis involving zootechnical matters. Bruno Rosolen Gilli, for all the help, encouragement and companionship in field sampling and laboratory determinations. São Paulo State University (UNESP), College of Agricultural Sciences, for the teaching, infrastructure and support offered. Graduate Program in Agriculture - UNESP/FCA, for the PhD opportunity. Professors and employees of the Graduate Program in Agronomy (Agriculture) of the São Paulo State University “Júlio de Mesquita Filho” (Botucatu Campus), for the teaching and collaboration. Employees of the Plant Production Department, especially Adelina, Casemiro, Dariele, Vinícius, Ciro, Dorival, Eliane, Iara, Júlia, Mateus, Valéria, and Vinícius for their friendship, collaboration and teaching. All members of NUCLEUS: A virtual joint centre to deliver enhanced N–use efficiency via an integrated soil–plant systems approach for the United Kingdom and Brazil. Funded in Brazil by FAPESP–São Paulo Research Foundation [grant number 2015/50305–8]. This work was carried out with the support of the São Paulo State Research Support Foundation (FAPESP), for granting a PhD scholarship (grant number 2016/25253-7) and BEPE scholarship (grant number 2018/09622-8) and Coordination for the Improvement of Higher Education Personnel-Brazil (CAPES) – Financing code 001. To the friends of the Graduate Program, especially Eduardo Mariano, Carlos Nascimento, Bruno Gazola, Sérgio Freitas, Gustavo Ferreira, Laudelino Motta, Barbara Silva, Priscila Bahia, Beatriz Borges, and Clóvis Borges, for the moments of partnership, friendship, collaboration and support. To the interns Bruno Rosolen Gilli, Rodrigo Martins, Lucas Brunini, Thiago Benetom, Fernando Thome, and Matheus Miziara for their friendship and help in carrying out this work. Employees and teachers of UNASP/SP, who contributed to my personal and school education. All those who, directly or indirectly, contributed to the development realization of this study, my sincere and eternal thanks. THANKS! ABSTRACT Due to the interest in N use efficiency (NUE) and sustainable agricultural systems, the adoption of integrated systems, such as the intercropping of maize with forage grasses can be of great relevance, allowing the use of the land throughout the year, besides avoiding losses of N through nitrate (NO3 -) leaching, nitrous oxide (N2O) emissions, ammonia volatilization (NH3), and immobilization. Tropical forage grasses of the genus Megathyrsus and Urochloa can suppress soil–nitrification by releasing inhibitory substances, reducing N losses and increasing fertilizer N recovery of the cash crop in rotation. In this way, understanding the N transformations in the soil by microorganisms and the fertilizer recovery in the system are very important. Firstly, the first two chapters are about a 3-year (2014-2017) field experiment conducted in southeastern Brazil, were forage grasses Guinea grass (Megathyrsus maximus cv. Tanzânia), palisade grass (Urochloa brizantha cv. Marandu), and ruzigrass (Urochloa ruziziensis cv. Comum) were cultivated in rotation with maize for grain in summer, to analyze the influence of forage grass and N fertilization in each study. In first chapter, maize was fertilized with 140 kg ha-1 N as (15NH4)2SO4 or not fertilized, and recovery of residual 15N was quantified in the second season. In second chapter, the change was that the N source used was ammonium sulphate not labeled, and were analyzed nitrous oxide (N2O), methane (CH4), and NH3 emissions from the system. In the third and fourth chapter, maize was intercropped with the same grasses previously mentioned. The N rates were 90, 180 and 270 kg ha-1 N and treatments without N fertilization. The objective was also to ascertain the effect of grasses and N fertilization from the analyzes carried out. The third chapter characterized the changes in N-cycle genes in the soil and measured the N2O emissions. The fourth chapter assessed maize grain yield and forage production, bromatological quality, and estimated meat production. In the first season after 15N application, 21%, 65%, and 33% of the N in maize grain, stover, and shoots, respectively, was derived from fertilizer. In the next season, of the total N found in maize grain, stover, and shoots, 2.2%, 1.9%, and 2.0%, respectively, was derived from the residual fertilizer applied in the previous year. There were no differences between forage grass species in the amount of 15N recovered by maize, soil, and total N. In the first season of maize in rotation with forage grasses, Guinea grass, palisade grass, ruzigrass did not affect N2O and NH3 emission due to their apparent inability to suppress soil nitrification. However, N fertilization slightly increases cumulative N2O emission. In maize intercropping with grasses, N fertilization increases the abundance of AOB (amoA of bacteria) more than AOA (amoA of archaea). N2O emission was influenced by AOB, water-filled pore space (WFPS) and N fertilization. Nitrogen fertilization positively affects forage growth and nutritional quality, resulting in a higher maize grain yield, higher forage production and quality, and eventually higher estimated meat production. Moreover, Guinea grass resulted in the highest estimated meat production when fertilized with 270 kg ha-1 N. However, no evidence of biological inhibition by the grasses were confirmed. Keywords: Zea mays L. Urochloa. Megathyrsus. N2O. AOB. AOA. Estimated meat production. RESUMO Devido ao interesse na eficiência no uso do N (NUE) e em sistemas agrícolas sustentáveis, a adoção de sistemas integrados, como o consórcio de milho com gramíneas forrageiras, pode ser de grande relevância, permitindo o uso da terra ao longo do ano, além de evitar perdas de N por lixiviação de nitrato (NO3 -), emissões de óxido nitroso (N2O), volatilização de amônia (NH3) e imobilização. Gramíneas forrageiras tropicais do gênero Megathyrsus e Urochloa podem suprimir a nitrificação do solo ao liberar substâncias inibidoras, reduzindo as perdas de N e aumentando a recuperação de N fertilizante da cultura comercial em sucessão. Desta forma, o entendimento das transformações do N no solo por microrganismos e a recuperação do fertilizante no sistema são muito importantes. Em primeiro lugar, os dois primeiros capítulos são a respeito de um experimento de campo de 3 anos (2014-2017) conduzido no sudeste do Brasil, onde gramíneas forrageiras capim colonião (Megathyrsus maximus cv. Tanzânia), capim braquiária (Urochloa brizantha cv. Marandu) e capim braquiária (Urochloa ruziziensis cv. Comum) foram cultivadas em rotação com milho para grão no verão, para analisar a influência da gramínea forrageira e da fertilização com N em cada estudo. No primeiro capítulo, o milho foi fertilizado com 140 kg ha-1 de N na forma de (15NH4)2SO4 ou não fertilizado, e a recuperação do 15N residual foi quantificada na segunda safra. No segundo capítulo, a mudança foi que a fonte de N utilizada foi o sulfato de amônio não rotulado, e foram analisadas as emissões de óxido nitroso (N2O), metano (CH4) e NH3 do sistema. No terceiro e quarto capítulos, o milho foi consorciado com as mesmas gramíneas mencionadas anteriormente. As doses de N foram 90, 180 e 270 kg ha-1 e os tratamentos sem adubação nitrogenada. O objetivo também foi verificar o efeito das gramíneas e da fertilização com N a partir das análises realizadas. O terceiro capítulo caracterizou as mudanças nos genes do ciclo N no solo e mediu as emissões de N2O. O quarto capítulo avaliou o rendimento de grãos de milho e a produção de forragem, a qualidade bromatológica e a estimativa da produção de carne. Na primeira safra após a aplicação de 15N, 21%, 65% e 33% do N no grão de milho, palha e brotos, respectivamente, foi derivado de fertilizante. Na safra seguinte, do total de N encontrado nos grãos, caules e ramos de milho, 2,2%, 1,9% e 2,0%, respectivamente, foram derivados do fertilizante residual aplicado no ano anterior. Não houve diferenças entre as espécies de gramíneas forrageiras na quantidade de 15N recuperado pelo milho, solo e N. total. Na primeira temporada de milho em rotação com gramíneas forrageiras, capim-Guiné, capim-paliçada, ruzigrass não afetou a emissão de N2O e NH3 devido à sua aparente incapacidade de suprimir a nitrificação do solo. No entanto, a fertilização com N aumenta ligeiramente a emissão cumulativa de N2O. No consórcio de milho com gramíneas, a fertilização com N aumenta a abundância de AOB (amoA de bactérias) mais do que AOA (amoA de arquéias). A emissão de N2O foi influenciada por AOB, espaço poroso cheio de água (WFPS) e fertilização de N. A fertilização nitrogenada afeta positivamente o crescimento das forragens e a qualidade nutricional, resultando em maior rendimento de grãos de milho, maior produção e qualidade das forragens, e eventualmente maior produção de carne estimada. Além disso, o capim colonião resultou na maior produção de carne estimada quando fertilizado com 270 kg ha-1 N. No entanto, não foram confirmadas provas de inibição biológica por parte das gramíneas. Palavras-chave: Zea mays L. Urochloa. Megathyrsus. N2O. AOB. AOA. Produção estimada de carne. FIGURE LIST Figure 1 - Monthly precipitation, and average minimum and maximum air temperatures in the first (2015–2016) and second (2016–2017) growing seasons, and long-term average (period 1955-2015)………………………………………32 Figure 2 -Schematic representation of the whole plot and its microplot (deployed exclusively in N–fertilized treatments), in addition to the sampling procedures of plant, litter, and soil…………………………………………….36 Figure 3 - Maize (partitioned into grains and stover) and litter N derived from fertilizer as affected by forage grass in the first (2015–2016) and second (2016-2017 growing season after application of 15N-labeled fertilizer………………….47 Figure 4 - Soil N derived from fertilizer at four depth layers as affected by forage grass in the first (2015–2016) and second (2016–2017) growing season after application of 15N-labeled fertilizer…………………………….……………..48 Figure 5 -The fate of 15N–labeled fertilizer in plant–litter–soil system as affected by forage grass in the first (2015–2016) and second (2016–2017) growing season after application of 15N-labeled fertilizer……………………………49 Figure 6 - Daily minimum and maximum air temperatures and rainfall recorded during the field experiment period (2014-2017)……………………………………..69 Figure 7 - Seasonal variation of water-filled pore space (a) and soil temperature (b) at depth of 5.5 cm in forage grass-maize rotations from 2015 to 2017 as affected by forage grass specie and N fertilization………………………….70 Figure 8 - Seasonal variation of NH4 +-N (a) and NO3 --N content in the soil (b) at depth of 0-10 cm in forage grass-maize rotations from 2014 to 2017 as affected by forage grass specie and N fertilization…………………………………….71 Figure 9 - Seasonal variation of N2O (a) and CH4 flux (b), in addition to NH3 loss rate (c) in forage grass-maize rotations from 2014 to 2017 as affected by forage grass specie and N fertilization………………………………………………..73 Supplementary Figure 1 - Cumulative emissions (or uptake) of N2O (a), CH4 (b), and NH3 (c) in forage grass-maize rotations as affected by forage grass specie and N fertilization over the experimental period…………………………….76 Supplementary Figure 2 - Cumulative crop biomass (forage plus maize) production (a) and yield-scaled emission (or uptake) of N2O (b), CH4 (c), and NH3 (d) in forage grass-maize rotations as affected by forage grass specie and N fertilization over the experimental period…………………………………….79 Figure 10 - Heatmap showing Spearman rank-order correlation coefficients (r) between greenhouse gas flux and soil properties (n = 237) throughout the experimental period (2015-2017)……………………………………………..80 Figure 11 - Crop succession occurred in the experimental site from 2014 to 2018….96 Figure 12- Daily minimum, maximum and average air temperatures and rainfall recorded during the growing season period (Dec.2017-May 2018)…….103 Figure 13 -Contents of NO3 --N (a), NH4 +-N (b), and total N (c) at the 0-20 cm soil layer as affected by forage grass specie, N fertilization, and time after planting……………………………………………………………………..….105 Figure 14 - Abundance of 16S rRNA of bacteria (a) and archaea (b and c), nifH (d), amoA of bacteria (e), amoA of archaea (f), nirS (g) and nosZ (h) at the 0-20 cm soil layer as affected by forage grass species, N fertilization and days after planting............................................................................................107 Figure 15 - Cumulative N2O emission in forage grass-maize intercropping as affected by forage specie and N fertilization………………………………………….109 Supplementary Figure 3 - Variation of water-filled pore space (a) and soil temperature at depth of 5.5 cm (b) as well as N2O flux (c) in forage grass-maize intercropping as affected by forage grass specie and N fertilization…….110 Figure 16 - Heatmap of the Spearman rank correlation order among the relative abundance of N-cycle functional genes, soil properties, and N2O flux. *, **, and *** significant at P≤0.05, P≤0.01, and P≤0.001, respectively……….111 Figure 17 - Principal component analysis of the soil microbial genes (a) and chemical properties (b) with N2O fluxes, water-filled pore space (WFPS) and temperature (TºC) from soil of 72 sampling points across three times………………………..………………………………………….....……113 Figure 18 - Monthly minimum, maximum and average air temperatures and rainfall recorded during the growing season period (Dec.2017-Nov.2018)……..135 Figure 19 - Forage dry matter yield (a) and maize grain yield (b) at maize harvest. The error bars represent the SEM (n = 8 and 4 for panels a and b, respectively)…………………………………………………………………..138 Figure 20 - Light interception in Guinea grass (a) and palisade grass (b) from 14 days after maize harvest until the forage grass reach 95% of LI; SPAD index in forage grass cut (c)…………………………………………………………...141 Figure 21 - Estimated meat production by the LRNS Cornell model…………….....…143 TABLE LIST Table 1 - Net nitrification rate from 10–d incubation of soil samples taken at the 0–20 cm depth layer before maize planting in the first (2015–2016) and second (2016–2017) growing season, in which the latter was also affected by fertilizer N rate………………………………………………………………………..…....39 Table 2 - Dry matter yield and N accumulation of forage grass species grown before maize in the first (2015–2016) and second (2016–2017) growing season, in wich the later was also affected by fertilizer N rate…………………………..40 Table 3 -Maize (partitioned into grains and stover) and litter dry matter yield and harvest index (HI) as affected by forage grass and fertilizer N rate in the first (2015-2016) and second (2016-2017) growing season……………………..42 Table 4 – Maize (partitioned into grains and stover) and litter N accumulation and N harvest index (NHI) as affected by forage grass and fertilizer N rate in the first (2015-2016) and second (2016-2017) growing season……………………..44 Table 5 - Seasonal N2O and CH4, and NH3 emissions (or uptake) in forage grass-maize rotations from 2014 to 2017 as affected by forage grass specie and N fertilization………………………………………………………………………..74 Table 6 - Seasonal crop biomass and yield-scaled (YS) emission (or uptake) of N2O, CH4, and NH3in forage grass-maize rotations from 2014 to 2017 as affected by forage grass specie and N fertilization ……………………………………77 Table 7 - Seasonal N2O emission factor in forage grass-maize rotations as influenced by forage grasses………………………………………………………………..80 Table 8 - Description of primers, standards DNA and amplification conditions used in qPCR analysis……………………………………………………………………98 Supplementary Table 1 - Soil properties as affected by N fertilization and days after planting (DAP)………………………………………………………………….104 Supplementary Table 2 - Probability (P) values for microbial genes in the soil as affected by forage grass species, N fertilization and days after planting (DAP)...………………………………………………………………………….108 Supplementary Table 3 - Scores of principal components analysis relative to temperature (TºC), water-filled pore space (WFPS), N2O flux, and abundance of microbial genes as presented in Fig. 17a…………………………………114 Supplementary Table 4 - Scores of principal components analysis of the soil chemical characteristics, water-filled pore space (WFPS), temperature (TºC), and N2O flux presented in Fig. 17b………………………………………………………115 Table 9 - Bromatological quality of forage grass at maize harvest. The error bars represent the SEM (n = 16, 8 and 4 for forage grasses, N rate and forage grass x N rate, respectively)……………………………………….………..…139 Table 10 - Dry matter yield and bromatological quality of forage grass at cutting in Nov. 2018………………………………………………….……………………….....142 SUMMARY GENERAL INTRODUCTION……………………………………………………...25 CHAPTER 1: FATE OF 15N FERTILIZER APPLIED TO MAIZE IN ROTATION WITH TROPICAL FORAGE GRASSES…….....………………………………...29 1.1 Introduction……………………………………………………………………….…30 1.2 Material and methods………….……………………………….………………….31 1.2.1 Study site……………………………………………..……………………….…….31 1.2.2 Study design………….……………..……………………………..……………..…33 1.2.3 Crop management………………………….……………………..………............33 1.2.4 Soil nitrification assessed by laboratory incubation………………………….....34 1.2.5 15N microplots…………………………………………..………………..………….34 1.2.6 Sampling procedure and 15N analyses …….……………….………………….…36 1.2.7 Calculations and statistical analysis………………….…………………….…….37 1.3 Results…………...………………………………………………………………….38 1.3.1 Net nitrification rates……………………………………………..…………………38 1.3.2 Forage yield and N accumulation………………………………..……………..…39 1.3.3 Maize yield and N accumulation………………………………..……………...….40 1.3.4 Distribution and fate of 15N–labeled fertilizer applied to maize…………………46 1.4 Discussion……..…………………………………………………………………....49 1.4.1 Dry matter yield and n accumulation…………………….………………..……..49 1.4.2 Distribution of 15N–Labeled fertilizer………….…………………………….…….51 1.4.3 Fate of 15N–labeled fertilizer…………………………….….…………….…..…..52 1.5 Conclusion…………………………………………..………...…...……….……….55 References……………………………………………………………………….....56 CHAPTER 2: EFFECT OF TROPICAL GRASS NITROGEN FERTILIZATION ON NITROUS OXIDE, METHANE, AND AMMONIA EMISSIONS OF MAIZE- BASED-ROTATION SYSTEMS…………………………………………….…….61 2.1 Introduction……………………………………….………………………....…...….62 2.2 Material and methods…………………………………………...…...……….…….64 2.2.1. Study site and experimental setup……………………….……………………...…64 2.2.2 Gas sampling procedure and auxiliary measurement……………………..….…65 2.2.3 Ammonia measurements………………………………………….………………..66 2.2.4. Calculations and statistical analysis………………………………….……………67 2.3 Results……………...……………………………………….……………….……….68 2.3.1 Environmental conditions and soil properties…………………………………….68 2.3.2 Nitrous oxide, methane, and ammonia emissions………………..……….……...71 2.3.3 Crop biomass, yield-scaled emissions, and N2O emission factor………………77 2.3.4 Relationships between greenhouse gas flux and soil properties……………….80 2.4 Discussion………………..…………………………………………………………..81 2.4.1 Greenhouse gas flux and rate of ammonia volatilization………………………..81 2.4.2 Cumulative emissions of greenhouse gases and ammonia…………………….82 2.4.3 N2O emission factor……………….……………………………………………......84 2.5 Conclusion…..………………………………………….………………….……….84 References…………………………………….………..…………..………...…….85 CHAPTER 3: FUNCTIONAL N-CYCLE GENES IN SOIL AND N2O EMISSIONS IN A MAIZE/TROPICAL FORAGE GRASSES INTERCROPPING SYSTEM……….……………………………..………………………………….….91 3.1 Introduction………………………...…………………………………..…….…..….92 3.2 Material and methods………….…………………………………...…………....…94 3.2.1 Site description and experimental design……………..….………………………94 3.2.2 Soil sampling…………………………………………………..…………………….95 3.2.3 Soil chemical analysis………………….…………………………………………...96 3.2.4 Isolation of dna and quantitative real-time PCR…………………………............96 3.2.5 Gas sampling, temperature and water-filled pore space procedure…………..101 3.2.6 Statistical analysis……………..……………………………………………..……102 3.3. Results……………………………………..……………………………..………..102 3.3.1 Environmental conditions and soil characteristics………………………………102 3.3.2 N-cycle functional genes………...……………………………………..………....106 3.3.3 Nitrous oxide emissions…………………………………………………………...108 3.3.4 Correlations between the abundance of N cycle genes, soil properties and N2O flux…….………………………………………………………………………….…111 3.3.5 Principal component analysis…………………………………………...……….112 3.4 Discussion………………………………………….………….……………………115 3.4.1 Linking N-cycle genes, forage grass and N fertilization………………………...115 3.4.2 N-cycle gene abundance and soil chemical characteristics……………………118 3.4.3 Relationships of N2O emissions with N-cycle genes, N fertilization and soil properties………….……………………..………………………………………...119 3.5….Conclusions…….....…………………………………………..…………………...120 References…………..……………..….……………….………………………....122 CHAPTER 4: BROMATOLOGICAL QUALITY AND ESTIMATED MEAT PRODUCTION IN MAIZE INTERCROPPING WITH TROPICAL FORAGE GRASSES WITH NITROGEN FERTILIZATION……………………………….131 4.1 Introduction…………………………………………….…………………………..132 4.2 Materials and methods……………………………….…………………………...133 4.2.1 Study site…………………………………………………….……………………..133 4.2.2 Study design………………………………………….……………………………134 4.2.3 Crop management…………………………………………..…………………… 134 4.2.4 Forage dry matter and maize grain yield……………………………………….135 4.2.5 Bromatological analysis of forage grass……………………………….……….135 4.2.6 Light interception, SPAD index and height of grasses………………………..136 4.2.7 Estimated meat production………………………………………….…………...136 4.2.8 Statistical analysis……………………………..………………………….………137 4.3 Results………...……………………………………...…………………….……...137 4.3.1 Forage dry matter and bromatological quality, and grain yield at maize harvest…..………………………………………………………………………….137 4.3.2 Forage light interception and SPAD index in grass cut…………………………140 4.3.3 Forage dry matter yield and bromatological quality at the final………………..140 4.3.4 Estimated meat production…………………………………………...……..…...143 4.4 Discussion…………………………………………………….……………………143 4.4.1 Forage yield and bromatological quality yield at maize harvest………………143 4.4.2 Light interception, height and SPAD index after maize-harvest……..………..145 4.4.3 Dry matter, bromatological quality and estimated meat production………….146 4.5 Conclusions.......................................................................................................147 References………..………………………...…………………………….……...….148 FINAL CONSIDERATIONS……………………………………………...………...155 REFERENCES……………………………..………………...……………….…….157 25 GENERAL INTRODUCTION Maize (Zea Mays L.) is among the most important food sources in the world, playing a key role (Ileri et al., 2018) in ensuring food security. It has been used for animal feed, human nutrition and, more recently, for bioethanol production (Ranum et al., 2014). As a C4 plant, maize is capable of achieving high dry matter yields accumulating large amounts of nutrients (Uzun et al., 2020). Usually, nitrogen (N) is the nutrient most required by maize (Teixeira et al., 2014; Wang et al., 2017). Nitrogen defficiency can limit crop yields, since this element is a constituent of important molecules such as amino acids, proteins, nucleic acids, nitrogenous bases, and chlorophyll (Moreira and Siqueira, 2002). One of the most studied topics regarding N in agricultural systems refers to strategies to improve N use efficiency (NUE) by crops, which can be achieved through conservation practices, like systems with different species intercropped (Adewopo et al., 2014; Rosolem et al., 2017). The adoption of integrated systems, such as the intercropping of maize with forage grasses has been of great relevance for tropical agriculture, allowing the use of land throughout the year (Crusciol et al., 2009; Kichel et al., 2009). However, increasing the NUE through conservation systems is paramount, as it is necessary to ensure adequate soil N availability for plants and to reduce losses in the agricultural systems (Rosolem et al., 2017). The volatilization of NH3 from fertilizer sources is an important issue when urea is applied on the soil surface in no-till systems (Mariano et al., 2012), due to the higher activity of the urease enzyme when compared to conventional tillage (Silva et al., 2017). However, in Brazil, with the N rates currently used and split applications, the risk of NO3 - reaching the groundwater is relatively low (Villalba et al., 2014). In addition, since N fertilizer is critical to sustain or increase the crop yield, the application of high N rates can lead to subsequent high N2O emissions in N-fertilized soils compared with those unfertilized (McSwiney and Robertson, 2005; Martins et al., 2015). No-till systems also largely affect the organic matter in the soil (Sá et al., 2015., Souza et al., 2016), and during the anaerobic decomposition of organic matter, CH4 production can occur (Dutaur and Verchot, 2007). Therefore, crop management strategies to decrease leaching losses and N2O emissions, as well as to increase NUE are fundamental for achieving adequate sustainability levels in agricultural systems (Rosolem, et al., 2017). 26 Despite the several benefits from the no-till systems in relation to conventional cropping systems, such as improvements in the chemical, physical and biological properties of the soil, and reduction of CO2 emissions (Lal et al., 2007), maize-grass intercropped systems can result in competition between species for N, which can compromise crop yields (Borghi et al. 2014). The rotation and intercropping of maize with tropical grasses of the genus Urochloa (syn. Brachiaria) and Megathyrsus (syn. Panicum) is very common for integrated crop-livestock systems in tropical Brazil (Salton et al., 2014). In addition, it has been suggested that forage species of the genus Urochloa and Megathyrsus can affect microbiological processes within the N cycle, as well as N availability and losses (Subbarao et al., 2012). Subbarao et al. (2012) reported that biological nitrification inhibition (BNI) capacity was much higher in U. humidicola than in U. decumbens, M. maximus, Lolium perenne, U. brizantha, cereal and vegetable crops studied in a sand-vermiculite culture for 60 days. Thus, using 15N to study fertilizer N recovery by cash crops intercropped with forage species is fundamental (Rocha et al., 2019) to assess whether the fate of fertilizer N in the soil- plant system is affected by BNI, when this process is active. The N transformations in the soil by microorganisms can occur in several ways (Zhang et al., 2006), and the understanding of these processes is essential in the search for efficient and sustainable agricultural systems. In the atmosphere, N is found as a diatomic molecule (N2), which can be fixed by a specialized microbiota, by a process known as biological N fixation (BNF; Cardoso, 1992). In the soil, this element can be found in organic and mineral forms (Cantarella, 2007). The soil N dynamics is complex, mostly driven by soil microorganisms as follows: i) BNF: the enzyme nitrogenase, which is encoded by the nifH gene, breaks the N2 triple bond to reduce N2 to ammonia (NH3; Zhang et al., 2006); ii) nitrification: NH4 + is converted to NO3 - via the action of ammonia monooxygenase (amoA). This process comprises two phases: oxidation of NH4 + to nitrite (NO2 -) and oxidation of NO2 - to NO3 -; iii) denitrification: copper nitrite reductase (nirK), iron nitrite reductase (nirS) and nitric oxide reductase (norB; Levy-Booth et al., 2014) oxidize NO3 - successively to NO2 -, NO and finally N2O; iv) the only route for converting N2O to N2 (Sun et al., 2019) is nitrous oxide reductase (nosZ; Henry et al., 2006). Considering all these aspects, the following hypotheses can be formulated: 27 1. Forage species with high BNI capacity would increase maize grain yield and 15N recovery in the soil-plant system by suppressing soil nitrification and therefore decreasing fertilizer N losses; 2. Tropical forage grasses would affect N cycle-associated genes and mitigate N2O emissions in N-fertilized maize; 3. Nitrogen fertilization would increase maize yield, as well as dry matter yield and bromatological quality of forage grasses. Based on these hypotheses, the objectives of this study were: i) to estimate N2O, CH4, and NH3 emissions in maize-based rotation systems as affected by tropical forage grasses; ii) assess whether rotation with tropical forage grasses influences maize dry matter yield, N accumulation, 15N recovery, and the fate of the N fertilizer in the plant-litter-soil system over two growing seasons; iii) to characterize changes in total bacterial and archaeal abundances and in microbial populations involved in N- fixation (nifH), ammonia oxidation (AOA and AOB), and denitrification (nirS and nosZ) and to measure N2O emissions in the maize intercropped with forage grasses; and iv) to assess maize grain yield and forage production, bromatological quality and estimated meat production in maize-forage grass intercropped systems. The first chapter of this thesis, entitled: “Fate of 15N fertilizer applied to maize in rotation with tropical forage grasses” was published in Field Crops Research. Chapter 2, entitled: “Effect of tropical grass and nitrogen fertilization on nitrous oxide, methane, and ammonia emissions of maize-based rotation systems” was published in Atmospheric Environment; and the third chapter, entitled: “Functional N-cycle genes in soil and N2O emissions in a maize/tropical forage grasses intercropping system” was recently submitted to Science of the Total Environment. Lastly, chapter 4, entitled “Bromatological quality and estimated meat production in maize intercropping with tropical forage grasses with N fertilization” will be submitted in due course to Grass and Forage Science. 28 29 CHAPTER 1 FATE OF 15N FERTILIZER APPLIED TO MAIZE IN ROTATION WITH TROPICAL FORAGE GRASSES Published in Field Crops Research (doi: 10.1016/j.fcr.2019.04.018) ABSTRACT Tropical forage grasses of the genus Megathyrsus and Urochloa can suppress soil– nitrification by releasing inhibitory substances, reducing N losses and increasing fertilizer N recovery of the cash crop in rotation. In contrast, ruzigrass (Urochloa ruziziensis) has been reported to decrease the yield and N accumulation of the subsequent crop and hence can affect N use efficiency and the fate of applied N. We investigated the effects of Guinea grass (M. maximum), palisade grass (U. brizantha), and ruzigrass on succeeding crop yield, N accumulation, and the fate of 15N–labeled fertilizer applied to maize (Zea mays L.) in a 2–year field experiment in Brazil. Maize was fertilized with 140 kg ha–1 N as (15NH4)2SO4 or not fertilized, and recovery of residual 15N was quantified in the second season. Net nitrification rates through an incubation study had no differences among grasses. Nitrogen application increased maize yield and N accumulation in both seasons, whereas maize yield decreased by 9.5% following ruzigrass compared with the other forages. The grasses had no effect on 15N recovery by maize or in the system. On average, the recovery of 15N in maize and soil was 34% and 46% in the first growing season and 2.9% and 20% in the second season, respectively. Our results indicated that tropical perennial grasses had no differential effects on nitrification rates and the fate of 15N–labeled fertilizer in the plant– litter–soil system in the season of application nor in the subsequent crop (residual effect). Keywords: Zea mays L.; Brachiaria; 15N; nitrogen uptake efficiency; soil N loss. 30 1.1 Introduction The benefits of no–till over conventional tillage systems include improvements in chemical, biological, and physical soil properties, such as increased C sequestration, microbial activity, and water and nutrient availability and reduced CO2 emissions, soil erosion, and weed incidence (Lal et al., 2007). These benefits are attained by growing crops without disturbing the soil and by maintaining plant residues on the soil surface. In addition, eliminating bare fallow periods in favor of growing leguminous or non– leguminous cover crops is a widely recognized method for increasing soil C stocks and improving nutrient cycling (Tonitto et al, 2006). Forage grasses grown as cover crops in the off–season have been also successfully used in integrated crop–livestock systems (Moraes et al., 2014). Leguminous cover crops can increase soil N supply through biological N fixation (Baligar and Fageria, 2007), but the rapid decomposition of their residues in tropical regions (Thomas and Asakawa, 1993) is a drawback in terms of soil protection. Therefore, non–leguminous species such as tropical perennial grasses have been introduced to increase the amount and persistence of the litter layer over the soil surface. In tropical Brazil, Guinea grass (Megathyrsus maximum cv. Tanzânia; syn. Panicum maximum cv. Tanzânia), palisade grass (Urochloa brizantha cv. Marandu; syn. Brachiaria brizantha cv. Marandu), andruzigrass (U. ruziziensis cv. Comum; syn. B. ruziziensis cv. Comum) are the main forage grass species used as cover crops in the off–season (April–September), while soybean [Glycine max (L.) Merrill] and maize (Zea mays L.) are the typical cash crops grown in summer (October–March). These tropical grasses have good drought tolerance during fall and winter due to their deep root systems (Fisher et al., 1994). Several earlier studies reported that forage grasses from the genus Urochloa, especially U. humidicola (syn. B. humidicola), can suppress soil nitrification through the exudation of inhibitory substances (Subbarao et al., 2012; 2015). Inhibition of soil nitrification has been proposed as a practical way to decrease environmental pollution caused by N fertilization (e.g., denitrification and NO3 – leaching) and to improve N uptake and fertilizer N recovery, primarily for plants preferring NH4 + over NO3 – (Subbarao et al., 2012). Despite the potential benefits of U. humidicola for plant N acquisition and reduced N loss to the environment, its cultivation in Brazil is essentially 31 limited to seasonally flooded soils. However, Guinea grass roots also release moderate amounts of nitrification inhibitors, whereas the suppressive effects of palisade grass seem to be lower (Subbarao et al., 2012). Although the inhibitory effects of ruzigrass on biological nitrification are unknown, adverse effects of its residues on the succeeding crop have been observed, such as lower yields and lower N accumulation (Echer et al., 2012; Souza et al., 2014; Marques et al., 2019). Microbial immobilization of mineral N during decomposition of forage residues and release of allelopathic substances have been suggested to be responsible for this effect (Echer et al., 2012; Souza et al. 2014). Various field studies have used the 15N method to assess fertilizer N recovery (here termed 15N recovery) by maize in monoculture or intercropped with grass species, and no interference of these forage crops has been confirmed (Coser et al., 2016; Almeida et al., 2017). However, the extent to which previously grown tropical forage grasses influence fertilizer N acquisition by maize and its fate in systems with crop rotation remain unclear. Furthermore, the ability of forage species to alter the recovery of residual fertilizer–derived N in the subsequent maize crop is also poorly understood. We hypothesized that forage species with high biological nitrification inhibition capacity could increase maize grain yield and15N recovery in the soil–plant system, thus decreasing N fertilizer losses. We aimed to (i) test the nitrification inhibition from Guinea grass, palisade grass, and ruzigrass; and (ii) assess whether rotation with the tropical forage grasses influences maize dry matter yield, N accumulation, 15N recovery, and the fate of 15N–labeled fertilizer in the plant–litter–soil system over two growing seasons. 1.2 Material and Methods 1.2.1 Study site A rainfed field experiment was conducted in Botucatu, SP, Brazil (22°49’ S, 48°26’ W; 700 m a.s.l.) for two consecutive cropping seasons (October 2015–May 2017) of maize affected by previously grown of tropical grasses. The local soil is a clay Rhodic Hapludox (Soil Survey Staff, 2014), with 190, 196, and 614 g kg–1 of sand, silt, and clay, respectively, at a depth of 0–20 cm. The clay fraction has ~70% kaolinite, ~15% gibbsite, and small amounts of vermiculite and illite. The study region typically 32 experiences dry winters and hot summers, with historical annual average minimum and maximum temperatures of 15.3 and 26.1°C, respectively. The average annual precipitation is 1360 mm. During the first (2015–2016) and second growing seasons (2016–2017), the average annual minimum temperatures were 17.0 and 16.4 °C, and the average annual maximum temperatures were 26.1 and 26.1°C, respectively (Fig. 1). The annual precipitation was 1859 mm and 1683 mm in the first and second seasons, 37% and 24% higher, respectively, than the long–term average (Fig. 1). Maize accumulated 1820 growing–degree days (GDD) and received 74% of the annual precipitation in the first growing season and 1919 GDD and 54% of the annual precipitation in the second season. The weather station used to measure the climate parameters was located 2.6 km from the study site. Prior to the experiment, the basic soil properties at the top 20 cm were: pH 5.9, total C 19 g kg–1, total N 1.3 g kg–1, NH4 +– N 5.4 mg kg–1, NO3 ––N 6.4 mg kg–1, P 15 mg dm–3, K 1.3 mmolc dm–3, Ca 35 mmolc dm–3, and Mg 24 mmolc dm–3, H+Al 37 mmolc dm–3, cation exchange capacity 97 mmolc dm–3, and base saturation 61%. Fig. 1. Monthly precipitation, and average minimum and maximum air temperatures in the first (2015–2016) and second (2016–2017) growing seasons, and long–term average (period 1955–2015). 33 1.2.2 Study design The experiment was conducted in split plots arranged in completely randomized blocks, with four replicates. Forage grass species were grown for eleven months (2014–2015) as cover crops, followed by planting of maize over the residues. In 2016, grasses were grown only in the maize off–season, followed by this grain crop. The forage species Guinea grass, palisade grass, and ruzigrass were grown in the main plots, while the subplots were assigned to N fertilization (140 kg ha–1 N) or the unfertilized control in maize. The subplots measured 4.5 m ×10 m (Fig. 2). Within each subplot, a microplot was set up to follow the fate of the 15N–labeled fertilizer applied to maize. A static plot design was deployed, with forage species and N fertilization treatments assigned repeatedly to the same plots. 1.2.3 Crop management Forage grasses were planted in November 2014 using a no–till drill at 7 kg of live seeds ha–1 with a row spacing of 0.17 m and no application of fertilizer. The forage grasses were cut twice, in April and June of the following year, at a height of 30 cm. Of the total dry matter yield of Guinea grass, palisade grass, and ruzigrass, 32%, 34%, and 50% was removed through cuts, while N removal was 43%, 45%, and 63% of the total accumulated N, respectively. In September 2015, the forage grasses were terminated using glyphosate (2.9 kg ha–1a.i.) and a mixture of paraquat and diuron (0.6 and 0.3 kg ha–1 a.i., respectively). The crop residues were left on the soil surface. Maize (hybrid 2B810PW, Dow AgroSciences, São Paulo, Brazil) was planted in October using the above-mentioned drill at a row spacing of 0.75 m to achieve a final stand of 65,000 plants ha–1. The hybrid used is glyphosate–resistant and insect–tolerant. Each main plot received 53 kg ha–1 P as triple superphosphate and 100 kg ha–1 K as potassium chloride at planting. The N fertilizer (granular ammonium sulfate) application was split twice, 30 kg ha–1 N at planting and 110 kg ha–1 N topdressed at growth stage V5 (five leaves with visible leaf collars). The topdressed N fertilizer was hand–applied to the soil surface in single–side banding (3 cm wide), ~5 cm from the crop row. The crop was hand–harvested in April 2016, and the maize stover (leaves, stems, and 34 cobs) was left in the field. Due to unfavorable climatic conditions after the first maize harvest, forage grasses were planted in October of the second season and desiccated 60 d after plant emergence. The forage grasses were not cut in the second season because growth was much less than in the first season. Maize (cv. hybrid 2B587PW, Dow AgroSciences, São Paulo, Brazil) was planted in December 2016. Apart from the maize cultivar, all agricultural practices (row spacing, plant density, and rate and timing of NPK fertilizers) were the same as in the first growing season. The maize was harvested in May of the following year. 1.2.4 Soil nitrification assessed by laboratory incubation To assess the influence of the forage grasses in the microbial oxidation of NH4 + over NO3 –, a soil incubation study was performed. Soil samples at the 0–20 cm were taken before maize planting in 2015 and 2016, oven–dried at 40°C to constant weight, and ground (< 2 mm mesh sieve). Two subsamples of 7 g of dry soil were transferred to 50–mL polypropylene centrifuge tubes and rewetted to 65% of water–holding capacity (Mariano et al., 2017). Soil samples were pre–incubated at 25°C for 10 d to decrease the mineral N flush. One subsample received 500 µL of 71 mM (NH4)2SO4 (140 µg N g−1) and was reincubated, while the other (untreated soil) was shaken with 2 M KCl (at a soil:solution ratio of 1:5; w/v) on an orbital shaker (200 rev min–1, 1 h). The supernatant was filtered using No. 42 filter paper, and NO3 ––N content at zero– time was determined by colorimetry (Miranda et al., 2001). Ten days following N addition, soil samples were extracted and analyzed for NO3 – as above. Net nitrification rate was calculated by subtraction of NO3 ––N of treated from untreated samples, divided by the incubation period. 1.2.5 15N microplots Unconfined microplots measuring 2.25 m × 1.50 m were set up in each N– fertilized (140 kg ha–1 N) subplot in the first growing season (Fig. 2). All agricultural factors in the microplots matched those of the subplot. Each microplot contained three rows of maize, with seven plants in each row. 15N–labeled ammonium sulfate [(15NH4)2SO4] with an abundance of 4.5 atom % 15N excess was obtained from Sigma– 35 Aldrich Inc. (St. Louis, MO, USA). Powder 15N–labeled fertilizer was applied at planting (30 kg ha–1 N) and as topdressing (110 kg ha–1 N) at the V5 stage of the maize. In the second growing season, unlabeled ammonium sulfate was applied on the microplots to assess the recovery of residual 15N–labeled fertilizer from the first season. All other agricultural practices remained the same as in the first season. 36 Fig. 2. Schematic representation of the whole plot and its microplot (deployed exclusively in N–fertilized treatments), in addition to the sampling procedures of plant, litter, and soil. Fertilizer N was applied over the litter layer in single–side bandings, ~3 cm from the maize row. 1.2.6 Sampling procedure and 15N analyses The plant, litter, and soil sampling procedures were the same in both growing seasons. At physiological maturity (R6 growth stage), three maize plants in the middle of each microplot were clipped at the stem base. The plants were partitioned into grains, cob cores, stem, and leaves (including sheaths). Fresh samples were oven– dried at 65°C to constant weight to assess the dry weight. The dry biomass was ground in a Wiley mill and passed through a 0.50–mm sieve for total N concentration and 15N measurements. The cob cores were added to the stem and leaves fraction to form the stover sample. Three plants from each control plot were randomly harvested and subjected to the protocol described above to assess the natural 15N abundance. The litter on the soil surface of each microplot was also sampled. All litter biomass (forage residues from the first growing season and forage plus maize residues from the second season) found in a central area of 0.75 m2 (0.75 m × 1.00 m; Fig. 2) of each microplot was collected and weighted. A similar protocol was used for the control plots. A subsample of the litter biomass was oven–dried at 65°C for dry weight, ground in a 37 Wiley mill, and passed through a 0.50–mm sieve for 15N analysis. The remaining fresh litter was returned to the field. Soil samples were taken using a core sampler at depths of 0–10, 10–20, and 20– 40 cm from four points in each microplot: (i) two samples from the central maize row, which received 15N–labeled fertilizer and (ii) two samples from the middle of the two outer maize rows (Fig. 2). Samples from the same depth and sampling position were combined (n = 2), oven–dried at 40°C, ground in a ball mill, and passed through a 0.0059mm sieve (equivalent to 100 mesh) for total N concentration and 15N analyses. To estimate the soil N accumulation, the soil bulk density at each soil depth and position was assessed using the volumetric ring method (Blake and Hartge, 1986) after the maize harvest. The natural 15N abundance in the soil was also measured. The grain, stover, litter, and soil samples were analyzed for total N concentration and 15N abundance using an automatic N analyzer (PDZ Europa ANCA–GSL, Sercon Ltd., Crewe, UK) interfaced with an isotope ratio mass spectrometer (PDZ Europa 20– 20, Sercon Ltd., Crewe, UK). 1.2.7 Calculations and statistical analysis The grain harvest index (HI) and N harvest index (NHI) of the maize were calculated as follows:  G SHI DM DM 100  (1)  G SNHI NC NC 100  (2) in which HI is the grain harvest index; DMG and DMS are the grain and shoot dry matter (kg ha–1), respectively; NHI is the N harvest index; and NCG and NCS are the N accumulation (kg ha–1) in the grains and shoots, respectively. The amount of N derived from fertilizer (Ndff), 15N recovery in maize, litter, and soil, and unrecovered N were calculated using the following equations:  1Ndff (kg ha ) a b NC   (3)  15N recovery (%) Ndff FNR 100  (4) 15 FS TFSUnrecovered N (%) 100 N recovery  (5)   SS TFS GFS TSSUnrecovered N (%) (Ndff Ndff ) Ndff FNR 100    (6) 38 where Ndff is the N derived from fertilizer; a and b are the 15N enrichment (atom % 15N excess) in the product (plant, litter, or soil) and substrate (fertilizer), respectively, both obtained by deducting the natural abundance (~0.368 atom % 15N); NC is the N accumulation (kg ha–1) in the product; 15N recovery is the percentage of fertilizer N recovery; FNR is the fertilizer N rate applied (kg ha–1); Unrecovered NFS and Unrecovered NSS are the percentage of fertilizer N unaccounted for (i.e., potential losses) in the first and second growing seasons after application of 15N–labeled fertilizer, respectively; 15N recoveryTFS is the total N recovery (%; sum of plant, litter and soil) in the first growing season; NdffTFS and NdffGFS are the N derived from fertilizer (kg ha–1) in the plant–litter–soil system and grains, respectively, in the first growing season; and NdffTSS is the N derived from fertilizer (kg ha–1) in the plant–litter–soil system in the second growing season. Generalized linear models were performed using the GLM procedure of SAS (version 9.3, SAS Institute, Inc., Cary, NC, USA). Block, forage grass, and N fertilization were considered fixed effects. Net nitrification rate, dry matter yield, and N accumulation of forage in the first and second growing seasons was subjected to one– way (effect of forage grass) and split plot (effect of forage grass and N fertilization) ANOVA, respectively. Splitplot ANOVA was conducted for dry matter yield and N accumulation of maize and litter, in addition to maize harvest indices (HI and NHI). The Ndff, 15N recovery, and unrecovered N results were subjected to one–way ANOVA. Fisher's least significant difference (LSD) test was used to compare least square means through the LS MEANS statement. Statistical significance is reported at the 5% level of significance. 1.3 Results 1.3.1 Net nitrification rates Net nitrification rate of soils sampled before maize planting did not differ among forage grasses in the first growing season, while nitrification increased by 34% following N fertilizer application relative to the control in the subsequent season (Table 1). 39 Table 1. Net nitrification rate from 10–d incubation of soil samples taken at the 0–20 cm depth layer before maize planting in the first (2015–2016) and second (2016–2017) growing season, in which the latter was also affected by fertilizer N rate. Values represent means ± SEM (n = 8 for the main effect of forage grass; n = 12 for the main effect of N rate; and n = 4 for the interaction between the forage grass and N rate). Growing season Forage grass N rate Net nitrification rate kg ha–1 mg NO3 ––N kg–1 d–1 First Guinea grass Control 3.2 ± 0.2 Palisade grass Control 3.2 ± 0.3 Ruzigrass Control 3.5 ± 0.5 P = 0.763 Second Guinea grass – 2.7 ± 0.4 Palisade grass – 2.2 ± 0.3 Ruzigrass – 2.6 ± 0.5 P = 0.788 – Control 2.2 ± 0.2b – 140 2.9 ± 0.4a P = 0.050 Guinea grass Control 2.5 ± 0.6 Palisade grass Control 1.8 ± 0.1 Ruzigrass Control 2.2 ± 0.5 Guinea grass 140 2.9 ± 0.7 Palisade grass 140 2.7 ± 0.6 Ruzigrass 140 3.0 ± 1.0 P = 0.800 Means followed by a common letter within a column are not significantly different by the LSD–test at the 5% level of significance. 1.3.2 Forage yield and N accumulation In the first growing season, the dry matter yield of palisade grass was 17% higher than that of the other forage grass species on average (Table 2). Nitrogen accumulation followed the results observed for dry matter yield and was highest for Guinea grass and palisade grass. In the second growing season, the dry matter yield and N accumulation of ruzigrass were 42% and 47% higher than those of palisade grass, respectively (Table 2). 40 Table 2. Dry matter yield and N accumulation of forage grass species grown before maize in the first (2015–2016) and second (2016–2017) growing season, in which the latter was also affected by fertilizer N rate. Growing season Forage grass N rate Dry matter yield N accumulation kg ha–1 Mg ha–1 kg ha–1 First Guinea grass Control 11.7 ± 0.7b 118 ± 9a Palisade grass Control 13.1 ± 0.5a 111 ± 12a Ruzigrass Control 10.6 ± 0.2b 80 ± 5b P= 0.008 P = 0.006 Second Guinea grass – 4.7 ± 0.6ab 48 ± 6ab Palisade grass – 4.2 ± 0.6b 38 ± 6b Ruzigrass – 6.3 ± 0.7a 63 ± 8a P = 0.041 P = 0.028 – Control 5.1 ± 0.5 51 ± 5 – 140 5.0 ± 0.7 48 ± 7 P = 0.892 P = 0.757 Guinea grass Control 4.6 ± 0.9 51 ± 9 Palisade grass Control 4.6 ± 0.5 42 ± 1 Ruzigrass Control 6.1 ± 1.1 61 ± 11 Guinea grass 140 4.8 ± 0.9 46 ± 8 Palisade grass 140 3.7 ± 1.2 34 ± 12 Ruzigrass 140 6.5 ± 1.1 65 ± 14 P = 0.861 P = 0.898 Means followed by a common letter within a column are not significantly different by the LSD–test at the 5% level of significance. Values represent means ± SEM (n = 8 for the main effect of forage grass; n = 12 for the main effect of N rate; and n = 4 for the interaction between the forage grass and N rate). 1.3.3 Maize yield and N accumulation The maize grain yield was 11% higher in succession to Guinea grass and palisade grass compared with ruzigrass in the first season, on average (Table 3). Grain and stover yield increased by 206% and 84%, respectively, following fertilizer N addition (140 kg ha–1 N)compared with the unfertilized control. An interaction of grass species with the N rate was observed for maize shoot biomass, which was greatest following palisade grass with N application and lowest following ruzigrass without N addition. Litter biomass was not affected by forage or N fertilization. However, HI 41 increased by 32% following N fertilization over the unfertilized control. In the second season, maize (grain, stover, and shoots) and litter yields and HI increased in response to N fertilization, and there was no effect of grass species (Table 3). Nitrogen accumulation in maize grain, stover, shoots, and NHI increased by 239%, 95%, 179%, and 22%, respectively, following fertilizer N compared with the unfertilized control in the first growing season (Table 4). In the second season, similar to the effects on dry matter yield, N accumulation in maize and litter increased following fertilizer N addition but had no effect on N accumulation in forage grasses (Table 4). 42 Table 3. Maize (partitioned into grains and stover) and litter dry matter yield and harvest index (HI) as affected by forage grass and fertilizer N rate in the first (2015–2016) and second (2016–2017) growing season. Values represent means ± SEM (n = 8 for the main effect of forage grass; n = 12 for the main effect of N rate; and n = 4 for the interaction between the forage grass and N rate). Growing season Maize (partitioned into grains and stover) and litter dry matter yield, N accumulation, harvest index (HI), and N harvest index (NHI) as affected by forage grass (cover crop) and fertilizer N rate in the first (2015– 2016) and second (2016–2017) growing season. Growingseason Forage grass N rate Dry matter yield (Mg ha–1) HI kg ha–1 Grains Stover Shoots Litter First Guinea grass – 7.1 ± 1.2a 8.0 ± 0.9 15.1 ± 2.0 3.8 ± 0.3 0.46 ± 0.03 Palisade grass – 7.1 ± 1.4a 8.4 ± 1.1 15.5 ± 2.5 5.1 ± 0.3 0.43 ± 0.03 Ruzigrass – 6.4 ± 1.3b 8.0 ± 1.0 14.4 ± 2.3 4.7 ± 0.4 0.42 ± 0.03 P = 0.016 P = 0.737 P = 0.291 P = 0.078 P = 0.226 – Control 3.4 ± 0.2b 5.7 ± 0.4b 9.1 ± 0.4b 4.5 ± 0.3 0.37 ± 0.01b – 140 10.4 ± 0.2a 10.5 ± 0.3a 20.9 ± 0.4a 4.6 ± 0.3 0.50 ± 0.01a P< 0.001 P< 0.001 P< 0.001 P = 0.710 P< 0.001 Guinea grass Control 4.0 ± 0.2 6.1 ± 1.0 10.0 ± 1.0c 4.1 ± 0.5 0.41 ± 0.03 Palisade grass Control 3.3 ± 0.1 5.6 ± 0.6 9.0 ± 0.6cd 5.1 ± 0.6 0.37 ± 0.02 Ruzigrass Control 2.9 ± 0.1 5.5 ± 0.3 8.3 ± 0.3d 4.3 ± 0.6 0.35 ± 0.01 Guinea grass 140 10.3 ± 0.3 9.9 ± 0.4 20.1 ± 0.3b 3.5 ± 0.3 0.51 ± 0.02 Palisade grass 140 10.9 ± 0.3 11.2 ± 0.6 22.1 ± 0.5a 5.2 ± 0.2 0.49 ± 0.02 Ruzigrass 140 10.0 ± 0.2 10.5 ± 0.4 20.5 ± 0.6ab 5.2 ± 0.4 0.49 ± 0.01 P = 0.053 P = 0.116 P = 0.046 P = 0.315 P = 0.411 Second Guinea grass – 6.2 ± 0.8 8.5 ± 1.0 14.8 ± 1.9 6.7 ± 0.8 0.43 ± 0.02 Palisade grass – 6.6 ± 1.0 8.5 ± 1.1 15.1 ± 2.0 6.6 ± 0.8 0.43 ± 0.01 Ruzigrass – 6.2 ± 1.1 8.6 ± 1.3 14.7 ± 2.4 5.8 ± 0.7 0.41 ± 0.02 43 P = 0.593 P = 0.996 P = 0.963 P = 0.420 P = 0.996 – Control 4.0 ± 0.3b 6.2 ± 0.5b 10.2 ± 0.8b 5.0 ± 0.4b 0.40 ± 0.01b – 140 8.6 ± 0.4a 10.9 ± 0.6a 19.6 ± 1.0a 7.7 ± 0.6a 0.45 ± 0.01a P< 0.001 P< 0.001 P< 0.001 P = 0.005 P = 0.003 Guinea grass Control 4.2 ± 0.6 6.3 ± 1.0 10.4 ± 1.6 5.0 ± 0.5 0.40 ± 0.01 Palisade grass Control 4.4 ± 0.7 6.2 ± 0.8 10.6 ± 1.5 5.5 ± 1.1 0.41 ± 0.01 Ruzigrass Control 3.5 ± 0.3 6.0 ± 1.2 9.6 ± 1.4 4.5 ± 0.4 0.41 ± 0.01 Guinea grass 140 8.3 ± 0.3 10.7 ± 0.8 19.1 ± 1.0 8.4 ± 1.1 0.46 ± 0.02 Palisade grass 140 8.8 ± 0.8 10.8 ± 1.1 19.6 ± 1.8 7.7 ± 1.0 0.45 ± 0.02 Ruzigrass 140 8.8 ± 1.0 11.1 ± 1.5 19.9 ± 2.4 7.0 ± 1.0 0.45 ± 0.01 P = 0.696 P = 0.944 P = 0.871 P = 0.775 P = 0.662 Means followed by a common letter within a column are not significantly different by the LSD–test at the 5% level of significance. 44 Table 4. Maize (partitioned into grains and stover) and litter N accumulation and N harvest index (NHI) as affected by forage grass and fertilizer N rate in the first (2015–2016) and second (2016–2017) growing season. Values represent means ± SEM (n = 8 for the main effect of forage grass; n = 12 for the main effect of N rate; and n = 4 for the interaction between the forage grass and N rate). Growing season Forage grass N rate N accumulation (kg ha–1) NHI kg ha–1 Grains Stover Shoots Litter First Guinea grass – 70 ± 13 31 ± 4 101 ± 16 46 ± 7 0.67 ± 0.02 Palisade grass – 68 ± 15 32 ± 4 100 ± 19 67 ± 9 0.64 ± 0.03 Ruzigrass – 63 ± 14 32 ± 5 96 ± 19 48 ± 6 0.63 ± 0.03 P = 0.118 P = 0.881 P = 0.524 P = 0.239 P = 0.319 – Control 30 ± 2b 22 ± 1b 52 ± 2b 52 ± 8 0.59 ± 0.02b – 140 103 ± 2a 42 ± 2a 146 ± 3a 56 ± 5 0.71 ± 0.01a P< 0.001 P< 0.001 P< 0.001 P = 0.581 P< 0.001 Guinea grass Control 37 ± 2 23 ± 4 59 ± 5 48 ± 14 0.62 ± 0.03 Palisade grass Control 29 ± 1 22 ± 2 51 ± 2 66 ± 18 0.56 ± 0.03 Ruzigrass Control 26 ± 1 20 ± 1 46 ± 1 41 ± 7 0.57 ± 0.03 Guinea grass 140 103 ± 4 40 ± 2 142 ± 5 45 ± 7 0.72 ± 0.01 Palisade grass 140 106 ± 5 42 ± 4 148 ± 8 68 ± 9 0.72 ± 0.01 Ruzigrass 140 101 ± 2 45 ± 3 146 ± 4 55 ± 8 0.69 ± 0.01 P = 0.251 P = 0.302 P = 0.279 P = 0.607 P = 0.418 Second Guinea grass – 96 ± 18 53 ± 11 149 ± 29 43 ± 6 0.66 ± 0.01 Palisade grass – 95 ± 19 39 ± 6 133 ± 25 42 ± 7 0.69 ± 0.02 Ruzigrass – 81 ± 16 41 ± 6 123 ± 21 37 ± 5 0.65 ± 0.03 P = 0.052 P = 0.136 P = 0.104 P = 0.758 P = 0.196 45 – Control 47 ± 4b 27 ± 3b 74 ± 6b 31 ± 3b 0.65 ± 0.02b – 140 135 ± 6a 62 ± 5a 196 ± 10a 51 ± 4a 0.70 ± 0.01a P< 0.001 P< 0.001 P< 0.001 P = 0.001 P = 0.009 Guinea grass Control 49 ± 8 28 ± 6 78 ± 14 32 ± 5 0.64 ± 0.01 Palisade grass Control 50 ± 8 26 ± 4 76 ± 12 33 ± 8 0.66 ± 0.01 Ruzigrass Control 41± 4 27 ± 6 68 ± 7 28 ± 3 0.66 ± 0.01 Guinea grass 140 143 ± 6 78 ± 9 221 ± 15 54 ± 8 0.67 ± 0.02 Palisade grass 140 139 ± 16 52 ± 8 191 ± 22 52 ± 9 0.73 ± 0.02 Ruzigrass 140 122 ± 8 56 ± 2 177 ± 10 47 ± 7 0.68 ± 0.01 P = 0.825 P = 0.232 P = 0.576 P = 0.901 P = 0.351 Means followed by a common letter within a column are not significantly different by the LSD–test at the 5% level of significance. 46 1.3.4 Distribution and fate of 15N–labeled fertilizer applied to maize In the first season after application of 15N–labeled fertilizer, Ndff in maize and litter did not differ in response to the preceding grass species (Fig. 3). Overall, 21%, 65%, and 33% of the N in maize grain, stover, and shoots, respectively, was derived from fertilizer. Of the total 15N detected in maize shoots (48 kg ha–1, on average), 43% was in grain, and the remainder was in stover. In the litter, only 10% of the N was derived from 15N–labeled fertilizer. Comparable results were observed in the second season, with no effect of forage species on maize and litter Ndff (Fig. 3). Overall, of the total N found in maize grain, stover, and shoots, 2.2%, 1.9%, and 2.0%, respectively, was derived from the residual fertilizer applied in the previous year. In contrast to the first season, most of the 15N accumulated in maize shoots (73%, on average) was found in grains rather than in stover. In the litter, 4.1% of the N was derived from the fertilizer. There was no difference in soil Ndff among the forage grasses (Fig. 3). Of the 140 kg ha–1 N applied in the first season, on average 64 kg ha–1 was found in the soil profile of 0 to 40 cm, of which 69% was in the topsoil (0–10 cm), 17% was in the 10–20 cm layer, and 14% was in the 20–40 cm layer (Fig. 4). Similarly, most of the residual 15N found in the 0–40 cm soil profile (29 kg ha–1, on average) at harvest in the second season was recovered from the upper layer (0–10 cm; Fig. 4). There were no differences among the forage grass species in the amount of 15N recovered in maize, soil, and total N in the first season (Fig. 5). However, 15N recovery from litter was 33% higher for palisade grass than for the other forage grasses. The average 15N recovery in maize, litter, and soil was 35%, 4%, and 46%, respectively, whereas 15% (21 kg ha–1) was unaccounted for (unrecovered N). Similarly, in the second season, forage grasses did not affect 15N recovery in maize (grain, stover, and shoots), litter, soil, and total N (Fig. 5). Overall, 2.9%, 1.5%, and 20% of the 15N–labeled fertilizer applied in the first growing season was recovered from maize shoots, litter, and soil, respectively, whereas 43% (60 kg ha–1) was unaccounted for in the plant– litter–soil system. 47 Fig. 3. Maize (partitioned into grains and stover) and litter N derived from fertilizer as affected by forage grass in the first (2015–2016) and second (2016–2017) growing season after application of 15N–labeled fertilizer. The error bars indicate the SEM (n = 4). NS: no significantly differences between forage grasses by the LSD–test at the 5% level of significance. 48 Fig. 4. Soil N derived from fertilizer at four depth layers as affected by forage grass in the first (2015–2016) and second (2016–2017) growing season after application of 15N–labeled fertilizer. The error bars indicate the SEM (n = 4). NS: no significantly differences between forage grasses by the LSD–test at the 5% level of significance. 49 Fig. 5. The fate of 15N–labeled fertilizer in plant–litter–soil system as affected by forage grass in the first (2015–2016) and second (2016–2017) growing season after application of 15N–labeled fertilizer. The error bars indicate the SEM (n = 4). Means followed by a common letter are not significantly different, while NS indicates no significantly differences between forage grasses, both by the LSD–test at the 5% level of significance. 1.4 Discussion 1.4.1 Dry matter yield and N accumulation The dry matter yield of the three grass species in the first and second growing seasons was within the range of 5.3 to 13.1 Mg ha–1 reported in other studies (Borghi et al., 2013; Pacheco et al., 2017; Marques et al., 2019), whereas N accumulation was lower than that reported by Marques et al. (2019). The wide variability of forage yields among these studies is directly related to the planting time (in–season and off– 50 season), duration of the growth cycle, and climatic conditions. Hence, the much lower forage dry matter yield recorded in the second season compared to the previous year (57% lower, on average) is explained by the short period of forage growth and the significant drought period during fall and winter. The lower maize grain yield following ruzigrass compared with Guinea grass and palisade grass in the first growing season is congruent with the findings of Marques et al. (2019). Two alternative hypotheses have been suggested to explain this observation: (i) a decrease in N availability to maize arising from high microbial N immobilization in the soil due to crop residues of ruzigrass (Echer et al., 2012); and (ii) allelopathic suppression due to secondary metabolites of forage grasses that inhibit maize growth (Weston and Duke, 2003; Souza et al., 2014). However, it is virtually impossible to separate allelopathic interferences from interferences by competition (i.e., utilization or competition for space, light, nutrients, and moisture) under field conditions (Weston and Duke, 2003). In addition to the lower grain yield, the lowest maize shoot biomass following ruzigrass in the unfertilized control is further evidence of the adverse effect of this grass species. The lack of an effect of forage grass species on maize yield and shoot biomass in the second growing season probably reflects the low dry matter yield of the forage crops. Increased dry matter yield and N accumulation by maize is a well–known and documented effect of N fertilizer application (Setiyono et al., 2010; Ciampitti and Vyn, 2012). In an extensive review, Ciampitti and Vyn (2012) reported that the average (n = 2074) shoot biomass production and N accumulation of modern maize hybrids at physiological maturity were 18 Mg ha–1 and 170 kg ha–1, respectively. The lack of an effect of forage species and N fertilization on litter dry matter yield and N accumulation in the first growing season may be attributed to the following mechanisms: (i) the narrow dry matter yield range (10.6–13.1 Mg ha–1) after desiccation; and (ii) the low amount of fertilizer–N retained in the litter layer following maize fertilization, which could lead to faster litter decomposition (Kuzyakov et al., 2000) in the case of significant retention. However, the higher dry matter yield and N accumulation in the litter with N fertilizer application in the subsequent season were due to the inclusion of maize stover from the previous season as a major component of the total litter biomass. In the first season, the dry matter yield and N accumulation of stover at harvest were, on average, 54% and 51% higher in the N–fertilized treatments than in the control, respectively, supporting this hypothesis. The average values of 0.43 for HI and 0.60 51 for NHI obtained in this study are within the ranges reported for modern maize hybrids (Setiyono et al., 2010; Ciampitti and Vyn, 2012). The lower HI and NHI values in the control treatments compared with N fertilization are associated with suboptimal N rates, which result in reduced grain filling and lower N remobilization from the stover to ears during the critical phases bracketing the silking period (Setiyono et al., 2010). 1.4.2 Distribution of 15N–labeled fertilizer In the first growing season, the Ndff in maize shoots (48 kg ha–1, on average) was within the range of 34 to 64 kg ha–1 found in previous studies (Coelho et al., 1991; Schindler and Knighton, 1999; Gava et al., 2006; Liu et al., 2015). The lower percentage of Ndff in grains (43%, on average) compared with NHI (0.60, considering the N–fertilized treatments) was probably due to partial 15N remobilization from other sources and/or plant N uptake at later stages (e.g., post–silking), which is typical for modern maize hybrids (Ciampitti and Vyn, 2012). In line with previous studies (Gava et al., 2006; Dourado–Neto et al., 2010; Wang et al., 2016), most of the plant N in the present study was derived from soil, primarily through mineralization processes. However, N inputs from litter decomposition, biological N fixation, and wet and dry deposition cannot be neglected as additional N sources for maize (Dentener et al., 2006; Silva et al., 2008; Montañez et al., 2009). The comparable and low values of Ndff in litter (5.6 kg ha–1, on average) among the forage grasses indicate that fertilizer N was weakly retained in this crop residue, even though microbial N immobilization and/or adsorption of NH4 + from ammonium sulfate in organic residues may be possible (Mariano et al., 2016). Therefore, the frequent rainfalls after fertilization most likely leached fertilizer N from plant residues into the soil. Accordingly, the substantially higher amount of 15N found in the 0–10 cm depth relative to the subjacent soil layers, irrespective of treatment, suggests considerable immobilization and some adsorption of mineral N forms (presumably NH4 +) in exchangeable sites of the topsoil, in line with the conclusions of other studies (Coelho et al., 1991; Liu et al., 2015; Karwat et al., 2017). However, the potential for NH4 + fixation in this soil (i.e., a highly weathered tropical soil) is low due to the high proportion of kaolinite, a 1:1 clay mineral (Nieder et al., 2011). While sampling deeper soil layers could increase residual N by around 22% in the first season (Reddy and Reddy, 1993; Liu et al., 2015; Wang et al., 2016), deep N leaching below 1.0 m was shown to be low in most of Brazilian regions, usually less 52 then 5.0% of the N applied (Villalba et al., 2014). The amount of NO3 ––N leached could not be related to rainfall (Rosolem et al., 2017). Furthermore, when deep rooted grasses are introduced in the system, N leaching is strongly decreased, and it is cycled to the topsoil (Rosolem et al., 2017). Although a large proportion of the fertilizer N (97 kg ha–1, on average) remained in the system after the first season, only a small amount of residual 15N (4.1 kg ha–1, on average) was taken up by maize. Considering the low percentage of Ndff in the shoots (2.1% of total N, on average), it can be inferred that 97.9% of the total N taken up by maize was derived from other sources (e.g., soil, unlabeled fertilizer, atmospheric deposition, etc) than the residual labeled N. In contrast to the first growing season, the values of the Ndff percentage in grains (73%, on average) and NHI (0.69, considering N–fertilized treatments) were similar, indicating that the residual 15N was taken up by maize mainly at later stages (e.g., during grain filling). Despite the retention of maize stover on the soil surface after the previous harvest, which resulted in Ndff of 27.3 kg ha–1, only 7.7% (considering all treatments) was recovered in the litter biomass in the following year as stover became a component of this organic layer. Although the isotope distribution in the soil in the second season was similar to that in the previous season, with most of the residual 15N found in the upper layer (0–10 cm), a substantial decrease in Ndff (55%, on average) was observed in the 0–40 cm soil layer. Different loss pathways most likely explain these observations, as will be detailed below. 1.4.3 Fate of 15N–labeled fertilizer To our knowledge, this is the first report on N recovery (using a 15N tracer) by maize grown in rotation with perennial tropical forage grasses. Previous studies have focused on land–use change, such as the shift from cultivation of U. humidicola or native vegetation to maize, in addition to maize monoculture as a control (Moreta et al., 2014; Karwat et al., 2017). The 15N recovery in maize shoots in the first growing season was within the previously reported range of 12%–57% (Coelho et al., 1991; Gava et al., 2006; Almeida et al., 2017; Karwat et al., 2017) and close to the global estimate of 33% (Raun and Johnson, 1999). Based on the lack of difference in Ndff and 15N recovery in shoots among forage grass species and the lower grain yield of maize succeeding ruzigrass, it may be inferred that ruzigrass residues did not affect the N uptake efficiency (i.e., 15N recovery) of maize but impaired crop yield, as 53 discussed above. The underlying mechanisms remain unclear but may involve allelopathy (Souza et al., 2014). Furthermore, the higher amount of palisade grass residue before maize planting likely explains the increased 15N recovery from litter. The similarities in 15N distribution and 15N recovery in the soil profile among the treatments show that either the decomposition of forage grass roots during maize growth had no effect on these two factors or that the potential effects were the same for all grasses. The lack of difference in the net nitrification rates between forage grasses clearly supports the15N recovery results. Conversely, the higher soil nitrification following N fertilization in the second season suggests an increased turnover rate of soil organic matter caused by the mineral fertilizer N applied to maize in the first season, increasing NH4 +oxidation to NO3 – (Kuzyakov et al., 2000).Thus, the similar soil nitrification rates of the forage grasses, despite the reported higher nitrification suppression capacity of Guinea grass over palisade grass (Subbarao et al., 2012) is possibly explained by two factors. First, the majority of studies evaluating biological nitrification inhibition were performed with U. Humidicola pastures established for years (more than 10–years– old), where the nitrification suppression is assumed to be high due to the cumulative release of inhibitory substances essentially from root exudation and root turnover(Subbarao et al., 2007, 2008; Moreta et al., 2014; Subbarao et al., 2015). Secondly, Karwat et al. (2017) postulated that the residual effect of biological nitrification inhibition is short and limited to the subsequent crop, as inhibitory substances can be leached or mineralized by microorganisms. Based on these factors, the residual nitrification suppression effect of the tropical forage grasses in rotation with maize (a plant with very low nitrification inhibition capacity) is questionable and may not reach the critical threshold levels to decrease soil nitrification rates and promote benefits to the agriculture and environment. The low recovery (<3.4%) of residual 15N–labeled fertilizer in maize shoots after harvest in the second season indicates that the contribution of previous N fertilization to a succeeding crop is limited or even negligible, since most of the residual 15N remains in the soil as organic N due to immobilization and will be released slowly by remineralization (Reddy and Reddy, 1993; Liu et al., 2015; Smith and Chalk, 2018). In a recent meta–analysis performed by Smith and Chalk (2018) with more than 100 studies on the residual value of 15N–labeled fertilizers, the authors reported that a consistent value of 5.4% of the initial applied N was recovered in subsequent crops. This result demonstrates the need for fertilizer application to each crop in order to 54 maintain or achieve high yields. The observed lack of influence of the forage grasses on 15N recovery in maize, litter, and soil in the second growth season confirms the low potential of these plants to alter fertilizer N dynamics in this rotation system, at least in the short term. The amount of unrecovered N (16%, or 22 kg ha–1, on average) in the first growing season is consistent with the results of many other studies (Gava et al., 2006; Wang et al., 2016; Almeida et al., 2017) but is lower than the value of 43 kg ha–1 reported in a meta–analysis by Gardner and Drinkwater (2009). The following pathways are associated with 15N deficits in the plant–litter–soil system: (i) ammonia volatilization from senescing leaves (Farquhar et al, 1980); (ii) ammonia volatilization following fertilizer addition (Sommer et al., 2004); (iii) leaching of NO3 – below the crop–rooting zone (Di and Cameron, 2002); (iii) nitrous oxide emission from plants during reduction of NO3 – (Smart and Bloom, 2001); and (iv) nitrous oxide emission from soil (Bouwman, 1996). However, during the first growing season, NO3 – leaching and nitrous oxide emission from the soil were negligible, whereas volatilization losses of ammonia from the canopy and soil accounted for ~3.0 kg ha–1 (Rocha, 2018). The substantial increase in unrecovered N in the second season compared with the first growing season (60 versus 22 kg ha–1, on average) may have been caused by the leaching of residual 15N below the sampling depth (40 cm.; Liu et al., 2015). We therefore suggest that the downward motion of hydrophilic organic N (Kalbitz et al., 2000) derived from fertilizer can be an important pathway of N loss from upper to lower soil layers in subsequent crops. 1.5 Conclusions The results of this study provide important information on the effects of tropical forage grasses grown in rotation with maize on the fate of 15N–labeled fertilizer in a no–till system. Soil nitrification assessed by laboratory incubation does not differ among forage species in the first and second growing seasons. While maize yield and N accumulation increase substantially following N fertilization than the unfertilized control, these crop parameters are generally not affected by the forage grass grown in rotation, except for ruzigrass, where maize grain yield in the first season is lower. The mechanism underlying this effect of ruzigrass is not completely understood. There is 55 no difference among the forage grass species in the distribution and fate of 15N in the plant–litter–soil system and, consequently, in unrecovered–N (i.e., potential losses). The amount of residual labeled N taken up by maize in the second growing season is very low. Guinea grass and palisade grass can be used interchangeably in rotation with summer maize; however, the effect of ruzigrass should be further investigated to prevent potential yield losses. 56 References Almeida, R.E.M., Oliveira, S.M., Lago, B.C., Pierozan Junior, C., Trivelin, P.C.O., Favarin, J.L., 2017. Palisadegrass effects on N fertilizer dynamic in intercropping systems with corn. An. Acad. Bras. Cienc. 89, 1917–1923. http://dx.doi.org/10.1590/0001–3765201720160811 Baligar, V.C., Fageria, N.K., 2007. Agronomy and physiology of tropical cover crops. J. Plant Nutr. 30, 1287–1339. http://dx.doi.org/10.1080/01904160701554997 Blake, G.R., Hartge, K.H., 1986. Bulk density. In: Klute, A. (Ed.), Methods of soil analysis: part 1. Physical and mineralogical methods, second ed. SSSA, ASA, Madison, pp 363–376. Borghi, E., Crusciol, C.A.C., Mateus, G.P., Nascente, A.S., Martins, P.O., 2013. Intercropping time of corn and palisade grass and Guineagrass affecting grain yield and forage production. Crop Sci. 53, 629–636. http://dx.doi.org/10.2135/cropsci2012.08.0469 Bouwman, A.F., 1996. Direct emission of nitrous oxide from agricultural soils. Nutr. Cycl. Agroecosyst. 46, 53–70. http://dx.doi.org/10.1007/BF00210224 Ciampitti, I.A., Vyn, T.J., 2012. Physiological perspectives of changes over time in maize yield dependency on nitrogen uptake and associated nitrogen efficiencies: A review. Field Crops Res. 133, 48–67. http://dx.doi.org/10.1016/j.fcr.2012.03.008 Coelho, A.M., França, G.E., Bahia, A.F.C., Guedes, G.A.A., 1991. Balanço de nitrogênio (15N) em um latossolo vermelho–escuro, sob vegetação de cerrado, cultivado com milho. R. Bras. Cienc. Solo 15, 187–193. Coser, T.R., Ramos, M.L.G., Figueiredo, C.C., Urquiaga, S., Carvalho, A.M., Barros, F.V., Mendonça, M.T., 2016. Nitrogen uptake efficiency of maize in monoculture and intercropped with Brachiaria humidicola and Panicum maximum in a dystrophic Red– Yellow Latosol of the Brazilian Cerrado. Crop Pasture Sci. 67, 47–54. http://dx.doi.org/10.1071/CP15077 Dentener, F., Drevet, J., Lamarque, J.F., Bey, I., Eickhout, B., Fiore, A.M., Hauglustaine, D., Horowitz, L.W., Krol, M., Kulshrestha, U.C., Lawrence, M., Galy– Lacaux, C., Rast, S., Shindell, D., Stevenson, D., Van Noije, T., Atherton, C., Bell, N., Bergman, D., Butler, T., Cofala, J., Collins, B., Doherty, R., Ellingsen, K., Galloway, J., Gauss, M., Montanaro, V., Müller, J.F., Pitari, G., Rodriguez, J., Sanderson, M., Solmon, F., Strahan, S., Schultz, M., Sudo, K., Szopa, S., Wild, O., 2006. Nitrogen and sulfur deposition on regional and global scales: A multimodel evaluation. Global Biogeochem. Cy. 20, GB4003. http://dx.doi.org/10.1029/2005GB002672 Di, H.J., Cameron, K.C., 2002. Nitrate leaching in temperate agroecosystems: sources, factors and mitigating strategies. Nutr. Cycl.Agroecosyst. 46, 237–256. http://dx.doi.org/10.1023/A:1021471531188 http://dx.doi.org/10.1590/0001–3765201720160811 http://dx.doi.org/10.1080/01904160701554997 http://dx.doi.org/10.2135/cropsci2012.08.0469 http://dx.doi.org/10.1007/BF00210224 http://dx.doi.org/10.1016/j.fcr.2012.03.008 http://dx.doi.org/10.1071/CP15077 http://dx.doi.org/10.1029/2005GB002672 http://dx.doi.org/10.1023/A:1021471531188 57 Dourado–Neto, D., Powlson, D., Abu Bakar, R., Bacchi, O.O.S., Basanta, M.V., thi Cong, P., Keerthisinghe, G., Ismaili, M., Rahman, S.M., Reichardt, K., Safwat, M.S.A., Sangakkara, R., Timm, L.C., Wang, J.Y., Zagal, E., van Kessel, C., 2010. Multiseason recoveries of organic and inorganic nitrogen–15 in tropical cropping systems. Soil Sci. Soc. Am. J. 74, 139–152. http://dx.doi.org/10.2136/sssaj2009.0192 Echer, F.R., Castro, G.S.A., Bogiani, J.C., Rosolem, C.A., 2012. Crescimento inicial e absorção de nutrientes pelo algodoeiro cultivado sobre a palhada de Brachiaria ruziziensis. Planta Daninha 30, 783–790. http://dx.doi.org/10.1590/S0100– 83582012000400012 Farquhar, G.D., Firth, P.M., Wetselaar, R., Weir, B., 1980. On the gaseous exchange of ammonia between leaves and the environment: determination of the ammonia compensation point. Plant Physiol. 66, 710–714. http//dx.doi.org/10.1104/pp.66.4.710 Fisher, M.J., Rao, M., Ayarza, M.A., Lascano, C.E., Sanz, J.I., Thomas, R.J., Vera, R.R., 1994. Carbon storage by introduced deep–rooted grasses in the South American savannas. Nature 371, 236–238. http://dx.doi.org/10.1038/371236a0 Gardner, J.B., Drinkwater, L.E., 2009. The fate of nitrogen in grain cropping systems: a meta–analysis of 15N field experiments. Ecol. Appl. 19, 2167–2184. http://dx.doi.org/10.1890/08–1122.1 Gava, G.J.C., Trivelin, P.C.O., Oliveira, M.W., Heinrichs, R., Silva, M.A., 2006. Balanço do nitrogênio da ureia (15N) no sistema solo–planta na implantação da semeadura direta na cultura do milho. Bragantia 65, 477–486. http://dx.doi.org/10.1590/S0006–87052006000300014 Kalbitz, K., Solinger, S., Park, J.H., Michalzik, B., Matzner, E., 2000. Controls on the dynamics of dissolved organic matter in soils: a review. Soil Sci. 165, 277–304. Karwat, H., Moreta, D., Arango, J., Núñez, J., Rao, I., Rincón, A., Rasche, F., Cadish, G., 2017. Residual effect of BNI by Brachiaria humidicola pasture on nitrogen recovery and grain yield of subsequent maize. Plant Soil 420, 389–406. http://dx.doi.org/10.1007/s11104–017–3381–z Kuzyakov, Y., Friedel, J.K., Stahr, K., 2000. Review of mechanisms and quantification of priming effects. Soil Biol. Biochem. 32, 1485–1498. http://dx.doi.org/10.1016/S0038–0717(00)00084–5 Lal, R., Reicosky, D.C., Hanson, J.D., 2007. Evolution of the plow power over 10,000 years and the rationale for no–till farming. Soil Till. Res. 93, 1–12. http://dx.doi.org/10.1016/j.still.2006.11.004 Liu, X., Li, X.G., Guo, R., Kuzyakov, Y., Li, F., 2015. The effect of plastic mulch on the fate of urea–N in rain–fed maize production in a semiarid environment as assessed by 15N–labeling. Eur. J. Agron. 70, 71–77. http://dx.doi.org/10.1016/j.eja.2015.07.006 http://dx.doi.org/10.2136/sssaj2009.0192 http://dx.doi.org/10.1590/S0100–83582012000400012 http://dx.doi.org/10.1590/S0100–83582012000400012 http://dx.doi.org/10.1038/371236a0 http://dx.doi.org/10.1890/08–1122.1 http://dx.doi.org/10.1590/S0006–87052006000300014 http://dx.doi.org/10.1007/s11104–017–3381–z http://dx.doi.org/10.1016/S0038–0717(00)00084–5 http://dx.doi.org/10.1016/j.still.2006.11.004 http://dx.doi.org/10.1016/j.eja.2015.07.006 58 Mariano, E., Jones, D.L., Hill, P.W., Trivelin, P.C.O., 2016. Mineral nitrogen forms alter 14C–glucose mineralisation and nitrogen transformations in litter and soil from two sugarcane fields. Appl. Soil Ecol. 107, 154–161. http://dx.doi.org/10.1016/j.apsoil.2016.05.019 Mariano, E., Otto, R., Montezano, Z.F., Cantarella, H., Trivelin, P.C.O., 2017. Soil nitrogen availability indices as predictors of sugarcane nitrogen requirements. Eur. J. Agron. 89, 25–37. http://dx.doi.org/10.1016/j.eja.2017.06.004 Marques, L.M., Crusciol, C.A.C., Soratto, R.P., Vyn T., Tanaka, K.S., Costa, C.H.M., Ferrari Neto, J., Cantarella, H., 2019. Impacts of nitrogen management on no–till maize production following forage cover crops. Agron. J. 111, 1–11. http://dx.doi.org/10.2134/agronj2018.03.0201 Montañez, A., Abreu, C., Gill, P.R., Hardarson, G., Sicardi, M., 2009. Biological nitrogen fixation in maize (Zea mays L.) by 15N isotope–dilution and identification of associated culturable diazotrophs. Biol. Fertil. Soils 45, 253–263. http://dx.doi.org/10.1007/s00374–008–0322–2 Moraes, A., Carvalho, P.C.F., Anghinoni, I., Lustosa, S.B.C., Costa, S.E.V.G.A., Kunrath, T.R., 2014. Integrated crop–livestock systems in the Brazilian subtropics. Eur. J. Agron. 57, 4–9. http://dx.doi.org/10.1016/j.eja.2013.10.004 Moreta, D.E., Arango, J., Sotelo, M., Vergara, D., Rincón, A., Ishitani, M., Castro, A., Miles, J., Peters, M., Tohme, J., Subbarao, G.V., Rao, I.M., 2014. Biological nitrification inhibition (BNI) in Brachiaria pastures: A novel strategy to improve eco– efficiency of crop–livestock systems and to mitigate climate change. Trop. Grasslands 2, 88–91. http://dx.doi.org/10.17138/tgft(2)88–91 Nieder, R., Benbi, D.K., Scherer, H.W., 2011. Fixation and defixation of ammonium in soils: a review. Biol. Fertil. Soils 47, 1–14. http://dx.doi.org/10.1007/s00374–010– 0506–4 Pacheco, L.P., São Miguel A.S.D.C., Silva, R.G., Souza, E.D., Petter, F.A., Kappes, C., 2017. Biomass yield in production systems of soybean sown in succession to annual crops and cover crops. Pesq. agropec. bras. 52, 47, 582–591. http://dx.doi.org/10.1590/S0100–204X2017000800003 Raun, W., Johnson, G.V., 1999. Improving nitrogen use efficiency for cereal production. Agron. J. 91, 357–363. http://dx.doi.org/10.2134/agronj1999.00021962009100030001x Reddy, G.B., Reddy, K.R., 1993. Fate of nitrogen–15 enriched ammonium nitrate applied to corn. Soil Sci. Soc. Am. J. 57, 111–115. http://dx.doi.org/10.2136/sssaj1993.03615995005700010021x Rocha, K.F., 2018. Dinâmica do nitrogênio em rotações de forrageiras com milho. Thesis (Doctorate in Agronomy–Agriculture). College of Agricultural Sciences. São Paulo State University, Botucatu. https://repositorio.unesp.br/handle/11449/152127 http://dx.doi.org/10.1016/j.apsoil.2016.05.019 http://dx.doi.org/10.1016/j.eja.2017.06.004 http://dx.doi.org/10.2134/agronj2018.03.0201 http://dx.doi.org/10.1007/s00374–008–0322–2 http://dx.doi.org/10.1016/j.eja.2013.10.004 http://dx.doi.org/10.17138/tgft(2)88–91 http://dx.doi.org/10.1007/s00374–010–0506–4 http://dx.doi.org/10.1007/s00374–010–0506–4 http://dx.doi.org/10.1590/S0100–204X2017000800003 http://dx.doi.org/10.2134/agronj1999.00021962009100030001x http://dx.doi.org/10.2136/sssaj1993.03615995005700010021x https://repositorio.unesp.br/handle/11449/152127 59 Rosolem, C.A., Ritz, K., Cantarella, H., Galdos, M.V., Hawkesford, M.J., Whalley, W.R., Mooney, S.J., 2017. Enhanced plant rooting and crop system management for improved N use efficiency. Adv. Agron. 146, 205–239. http://dx.doi.org/10.1016/bs.agron.2017.07.002 Schindler, F.V., Knighton, R.E., 1999. Fate of fertilizer nitrogen applied to corn as estimated by the isotopic and difference methods. Soil Sci. Soc. Am. J. 63, 1734– 1740. http://dx.doi.org/10.2136/sssaj1999.6361734x Setiyono, T.D., Walters, D.T., Cassman, K.G., Witt, C., Dobermann, A., 2010. Estimating maize nutrient uptake requirements. Field Crops Res. 118, 158–168. http://dx.doi.org/10.1016/j.fcr.2010.05.006 Silva, E.C., Muraoka, T., Buzetti, S., Espinal, F.S.C., Trivelin, P.C.O., 2008. Utilização do nitrogênio da palha de milho e de adubos verdes pela cultura do milho. R. Bras. Cienc. Solo 32, 2853–2861. http://dx.doi.org/10.1590/S0100– 06832008000700032 Smart, D.R., Bloom, A.J., 2001. Wheat leaves emit nitrous oxide during nitrate assimilation. Proc. Natl. Acad. Sci. USA 98, 7875–7878. http://dx.doi.org/10.1073/pnas.131572798 Smith, C.J., Chalk, P.M., 2018. The residual value of fertilizer N in crop sequences: an appraisal of 60 years of research using 15N tracer. Field Crops Res. 217, 66–74. http://doi.org/10.1016/j.fcr.2017.12.006 Soil Survey Staff, 2014. Keys to Soil Taxonomy, twelfth ed. USDA, Natural Resources Conservation Service, Washington. Sommer, S.G., Schjoerring, J.K., Denmead, O.T., 2004. Ammonia emission from mineral fertilizers and fertilized crops. Adv. Agron. 82, 557–622. http://dx.doi.org/10.1016/S0065–2113(03)82008–4 Souza, E.F.C., Fernandes, A.M., Souza–Schlick, G.D., Rosolem, C.A., 2014. Early growth of common bean cropped over ruzigrass residues. Planta Daninha 32, 775– 781. http://dx.doi.org/10.1590/S0100–83582014000400012 Subbarao, G.V., Rondon, M., Ito, O., Ishikawa, T., Rao, I.M., Nakahara, K., Lascano, C., Berry, W.L., 2007. Biological nitrification inhibition (BNI)–is it a widespread phenomenon? Plant Soil 294, 5–18. http://dx.doi.org/10.1007/s11104–006–9159–3 Subbarao, G.V., Nakahara, K., Ishikawa, T., Yoshihashi, T., Ito, O., Ono, H., Ohnishi–Kameyama, M., Yoshida, M., Kawano, N., Berry, W.L., 2008. Free fatty acids from the pasture grass Brachiaria humidicola and one of their methyl esters as inhibitors of nitrification. Plant Soil 313, 89–99. http://dx.doi.org/10.1007/s11104– 008–9682–5 Subbarao, G.V., Sahrawat, K.L., Nakahara, K., Ishikawa, T., Kishii, M., Rao, I.M., Hash, C.T., George, T.S., Srinivasa Rao, P., Nardi, P., Bonnett, D., Berry, W., Suenaga, K., Lata, J.C., 2012. Biological nitrification inhibition–a novel strategy to http://dx.doi.org/10.2136/sssaj1999.6361734x http://dx.doi.org/10.1016/j.fcr.2010.05.006 http://dx.doi.org/10.1590/S0100–06832008000700032 http://dx.doi.org/10.1590/S0100–06832008000700032 http://dx.doi.org/10.1073/pnas.131572798 http://doi.org/10.1016/j.fcr.2017.12.006 http://dx.doi.org/10.1016/S0065–2113(03)82008–4 http://dx.doi.org/10.1590/S0100–83582014000400012 http://dx.doi.org/10.1007/s11104–006–9159–3 http://dx.doi.org/10.1007/s11104–008–9682–5 http://dx.doi.org/10.1007/s11104–008–9682–5 60 regulate nitrification in agricultural systems. Adv. Agron. 114, 249–302. http://dx.doi.org/10.1016/B978–0–12–394275–3.00001–8 Subbarao, G.V., Yoshihashi, T., Wothington, M., Nakahara, K., Ando, Y., Sahrawat, K.L., Rao, I.M., Lata, J., Kishii, M., Braun, H., 2015. Suppression of soil nitrification by plants. Plant Sci. 233, 155–164. http://dx.doi.org/10.1016/j.plantsci.2015.01.012 Thomas, R.J., Asakawa, N.M., 1993. Decomposition of leaf litter from tropical forage grasses and legumes. Soil Biol. Biochem. 23, 1351–1361. http://dx.doi.org/10.1016/0038–0717(93)90050–L Tonitto, C., David, M.B., Drinkwater, L.E., 2006. Replacing bare fallows with cover crops in fertilizer–intensive cropping systems: A meta–analysis of crop yield and N dynamics. Agr. Ecosyst. Environ. 112, 58–72. http://dx.doi.org/10.1016/j.agee.2005.07.003 Villalba, H.A.G., Leite, J.M., Otto, R., Trivelin, P.C.O., 2014. Fertilizantes nitrogenados: novas tecnologias. Informações Agronômicas. 148, 12–18. http://www.ipni.net/publication/ia–brasil.nsf/issue/IA–BRASIL–2014–148. Wang, S., Luo, S., Yue, S., Shen, Y., Li, S., 2016. Fate of 15N fertilizer under different nitrogen split applications to plastic mulched maize in semiarid farmland. Nutr. Cycl. Agroecosyst. 105, 129–140. http://dx.doi.org/10.1007/s10705–016–9780–3 Weston, L.A., Duke, S.O., 2003. Weed and crop allelopathy. Crit. Rev. Plant Sci. 22, 367–389. http://dx.doi.org/10.1080/713610861 http://dx.doi.org/10.1016/B978–0–12–394275–3.00001–8 http://dx.doi.org/10.1016/0038–0717(93)90050–L http://dx.doi.org/10.1016/j.agee.2005.07.003 http://dx.doi.org/10.1007/s10705–016–9780–3 61 CHAPTER 2 EFFECT OF TROPICAL GRASS NITROGEN FERTILIZATION ON NITROUS OXIDE, METHANE, AND AMMONIA EMISSIONS OF MAIZE-BASED ROTATION SYSTEMS Published in Atmospheric Environment (doi: 10.1016/j.atmosenv.2020.117571) Abstract While tropical grasses were shown to inhibit the activity of soil nitrifiers, their role in greenhouse gas (GHG) and ammonia (NH3) emissions in N fertilized maize-based rotations are poorly understood. A 3-year (2014-2017) field experiment was conducted in southeastern Brazil to assess the influence of forage grass and N fertilization on nitrous oxide (N2O), methane (CH4), and NH3 emissions from maize (Zea mays L.)- grass rotations. Guinea grass (Megathyrsus maximus cv. Tanzânia), palisade grass (Urochloa brizantha cv. Marandu), and ruzigrass (Urochloa ruziziensis cv. Comum) were grown in the main plots, while an unfertilized control and 140 kg N ha-1 were applied annually to maize in sub-plots. No apparent nitrification suppression by the grasses was detected.N2O fluxes increased following N fertilizer addition in maize, particularly in the second season, where slightly higher cumulative N2O emission was observed with N fertilization in comparison with the control. CH4 fluxes showed high variation in the first forage and maize growing seasons. Residual N fertilizer decreased soil CH4 uptake of palisade grass and ruzigrass compared with unfertilized palisade grass in the second forage season. Cumulative NH3 emissions were unaffected by forage species and N fertilization. However, in both maize seasons, yield-scaled NH3 emission was the lowest following N addition. Throughout the seasons, the differences between the three grasses inN2O, CH4, and NH3 emissions were minimal. We conclude that the tropical perennial grasses rotated with maize were similar regarding GHG and NH3 emissions, while N fertilization slightly increased N2O emission and decreased soil CH4 uptake. Keywords: Zea mays L.; Brachiaria; Panicum; Nitrogen fertilizer; Nitrogen losses. 62 2.1 Introduction From the “Green Revolution” (starting in the 1950s) to 2012, the emission of