The impact of soil disturbance on nitrogen partitioning and carbon secuestration in Mediterranean soils

THE IMPACT OF SOIL DISTURBANCE ON NITROGEN PARTITIONING AND CARBON SECUESTRATION IN MEDITERRANEAN SOILS Sevilla, 2017 Memoria que presenta: María López Martín para optar al título de Doctor por la Universidad de Sevilla en: Recursos Naturales y Medio Ambiente THE IMPACT OF SOIL DISTURBANCE ON NITROGEN PARTITIONING AND CARBON SECUESTRATION IN MEDITERRANEAN SOILS Visado en Sevilla, a 12 de mayo de 2017 LA DIRECTORA LA TUTORA Dra. Dña. Elena Fernández Boy Profesor Titular de la Universidad de Sevilla Memoria que presenta Dña. María López Martín para optar al grado de Doctor en Recursos Naturales y Medio Ambiente Dra Dña. Heike Knicker Profesor de Investigación del CSIC IRNAS-CSIC DOCTOR D. JOSÉ ENRIQUE FERNÁNDEZ LUQUE, DIRECTOR DEL INSTITUTO DE RECURSOS NATURALES Y AGROBIOLOGÍA DE SEVILLA DEL CONSEJO SUPERIOR DE INVESTIGACIONES CIENTÍFICAS. Certifica: Que la presente Memoria de Investigación titulada “The impact of soil disturbance on nitrogen partitioning and carbon sequestration potential in Mediterranean soils”, presentada por Dña. María López Martín para optar al grado de Doctor en Recursos Naturales y Medio Ambiente, ha sido realizada en el Departamento de Cristalografía, Mineralogía y Química Agrícola, bajo la dirección de la Dra. Dña. Heike Knicker, reuniendo todas las condiciones exigidas a los trabajos de Tesis Doctorales. En Sevilla, a 12 de mayo de 2017 El presente trabajo se ha realizado en el marco del proyecto CGL2009-10557 del Ministerio de Ciencia e Innovación, gracias a una beca predoctoral dentro del Programa Nacional de Formación de Personal Investigador (BES-2010-042581). “Hay que perseverar y sobre todo, tener confianza en uno mismo” Marie Curie - Índice - página Resumen _____________________________________________________________ i Abstract _____________________________________________________________ iv 1. Introducción _________________________________________________________ 1 2. Objetivos específicos del trabajo_________________________________________ 19 Publicaciones científicas 3. Variability of the quality and quantity of organic matter in soil affected by multiple wildfires _____________________________________________________ 27 3.1. Introduction 30 3.2. Material and methods 32 3.3. Results and discussion 38 3.4. Conclusions 46 3.5. References 47 4. Distribution and transformation of Black Carbon and Black Nitrogen in physical soil fractions from fire-affected soils after seven years of recuperation _________ 53 4.1. Introduction 56 4.2. Material and methods 58 4.3. Results and discussion 62 4.4. Conclusions 75 4.5. References 76 5. Can Black Nitrogen serve as a N-source for plant growth in fire-affected areas and how is its use affected by the presence of inorganic N? ___________________ 81 5.1. Introduction 84 5.2. Material and methods 85 5.3. Results and discussion 88 5.4. Conclusions 103 5.5. References 104 6. Incorporation of N from burnt and unburnt 15N-grass residues into the peptidic fraction of fire affected and unaffected soils ______________________ 107 6.1. Introduction 110 6.2. Material and methods 112 6.3. Statistical analysis 116 6.4. Results and discussion 117 6.5. Summary and conclusions 127 6.6. References 128 7. Resultados generales _________________________________________________ 133 8. Discusión general ____________________________________________________ 141 9. Conclusions ________________________________________________________ 151 Appendix __________________________________________________________ 157 i -ResumenLos incendios forestales pueden ser considerados como uno de los factores más influyentes en la degradación de los suelos en la cuenca Mediterránea. El fuego puede modificar las propiedades físicas y químicas del suelo, y, en particular, alterar la calidad y cantidad de la materia orgánica del suelo (MOS), a corto, medio y largo plazo. La quema de la vegetación produce residuos altamente alterados químicamente conocidos genéricamente como material orgánico pirogenético (PyOM, acrónimo del término inglés pirogenic organic matter), y el nitrógeno (N) orgánico que contiene es llamado Black Nitrogen (BN). Químicamente, el BN está compuesto por compuestos aromáticos heterocíclicos, y su incorporación a la MOS va a afectar presumiblemente el ciclo del N en el suelo. En una primera fase de este trabajo fueron muestreados Cambisoles localizados en la Sierra de Aznalcóllar (Sevilla) bajo un bosque típico Mediterráneo propenso a los incendios, que habían sido afectados por distinto número de incendios. Los suelos elegidos fueron afectados por el fuego por primera vez en 1996, y posteriormente una segunda vez en 2004. Con objeto de evaluar la reproducibilidad de nuestros datos sobre el impacto del fuego en la calidad y cantidad de la MOS en las áreas estudiada, así como la influencia de la variabilidad espacial, principalmente en cuanto a la distribución del combustible antes del fuego se compararon dos métodos de muestreo diferentes En el primero de ellos (método 1), las esquinas y el centro de un cuadrado elegido al azar con una longitud de lado de 15 m fueron muestreados y analizados por separado. En el segundo, se utilizaron muestras compuestas (método 2) obtenidas de tres suelos. Los análisis efectuados permitieron describir las propiedades físicas y químicas de los suelos, así como sus características espectroscópicas mediante resonancia magnética nuclear (RMN), que fueron comparadas estadísticamente usando ANOVA. Ambos métodos de muestreo arrojaron resultados comparables con similares desviaciones estándar. No se observaron grandes diferencias con respecto a las propiedades físicas y químicas, contenido de carbono (C) y N entre los suelos no afectados y afectados por el fuego. Sin embargo los resultados de la RMN en estado sólido de 13C mostraron que los suelos quemados presentaban un incremento en aromaticidad. En base a estos resultados se concluyó que el muestreo con número reducido de réplicas (método 1) puede ser adecuado para aportar datos representativos del comportamiento por RMN de los suelos. Los resultados indicaron que en el área de estudio ii afectada por el fuego recuperó rápidamente las propiedades típicas del suelo y de su MOS, y que los cambios provocados por el fuego no se mantuvieron a largo del tiempo. Para elucidar la estabilidad bioquímica del Black Carbon (BC) y del BN de los suelos estudiados, muestras seleccionadas de la campaña del 2011 fueron estudiadas por RMN en estado sólido de 13C y 15N, y sus espectros fueron comparados con aquellos que fueron adquiridos de las muestras muestreadas en la misma zona, 4 semanas después del fuego en 2004. El análisis confirmó una incorporación rápida del material quemado en la MOS, justo después del incendio. Como consecuencia, el N tipo pirrol dominó la fracción de N orgánico del suelo. Después de siete años de recuperación del suelo se produjo una considerable disminución del C aromático, aunque su contribución al C total fue mayor en las zonas quemadas que en el suelo control. El hecho de la pérdida de PyOM durante la recuperación del suelo cuestionó la longevidad de los residuos quemados en el suelo. El BN mostró incluso menor recalcitancia bioquímica. El fraccionamiento por densidad y tamaño de partícula de la materia orgánica (MO) de los suelos que se están recuperando del incendio indicaron que el material quemado remanente se acumuló principalmente en la fracción orgánica particulada. Sin embargo, la baja recalcitrancia bioquímica debido a la falta de interacción entre la fracción mineral, el BN y el BC requiere de futuras investigaciones. Con el fin de aportar más luz sobre el impacto del BN en el ciclo del N en el suelo, se llevó a cabo un experimento en macetas durante 0.5, 1, 5, 8, 12 y 16 meses bajo condiciones controladas en invernadero. Para ello, Lolium perenne fue cultivado en suelos quemado y no quemado después de ser cubiertos con MO fresca enriquecida con 15N (15N-MO) o su 15N-PyOM. Paralelamente, los contenedores fueron fertilizados con KNO3 no enriquecido (Ni), mientras que otros fueron cubiertos con PyOM y MO junto con K15NO3 (15Ni) y otros con sólo 15Ni. La determinación de la recuperación del 15N añadido (15Nadd) junto con un análisis estadístico de los resultados mostró que el impacto del fuego no produjo cambios en la distribución del 15Nadd entre plantas (hojas y raíces) y suelo. En las hojas de las plantas, las mayores recuperaciones de 15Nadd se observaron en las macetas tratadas con 15Ni y 15N-MO en comparación con 15N-PyOM. La adición del Ni no alteró el porcentaje de 15Nadd recuperado derivado de 15N-PyOM y 15N-MO. Comparativamente, la adición de PyOM and OM no alteró la eficiencia de 15Ni incorporado en los residuos vegetales. Después de 16 meses de incubación, los suelos cubiertos con 15N-PyOM mostraron el mayor secuestro de 15Nadd mientras que la RMN de estado sólido de 15N sugirió que parte del 15Nadd iii secuestrado en los suelos cubiertos con 15N-PyOM se encontraba en forma de N amida, probablemente perteneciente a residuos de biomasa microbiana. En resumen, nuestros resultados permiten concluir que la adición de PyOM rico en N y las cenizas incorporadas al suelo después de un incendio tienen un fuerte impacto a corto plazo en el ciclo del N. Por consiguiente, justo después de un incendio, el Ni de las cenizas reemplaza el N, que había sido suministrado por la degradación de la hojarasca. Esto prueba una rápida fertilización con N, necesaria para una rápida recuperación de la cobertura vegetal. Dado que el Ni no es usado directamente para la producción de biomasa, puede perderse rápidamente por volatilización o lixiviado. Por tanto el PyOM sirve como sumidero eficiente de N, que es liberado lentamente y así aporta una fertilización adicional a medio o largo plazo. A largo plazo, casi todo el N del PyOM será transformado en N orgánico del suelo (NOS), con propiedades y funciones comparables al NOS derivado de la enmienda no quemada. Finalmente, investigamos la incorporación of 15N derivado del 15N-MO o de 15N-PyOM en los aminoácidos extraíbles (AAs) de la MOS, derivados de suelos quemado y no quemado enmendados con dichos residuos con el fin de revelar el papel del BN en el crecimiento de la biomasa del suelo. Los AAs fueron extraídos con ácido clorhídrico de los suelos utilizados en el experimento con macetas. La cantidad de 15N añadido recuperado en la fracción de los AAs fue determinada mediante cromatografía de gases - espectrometría de masas. Este experimento confirmó que el N de PyOM es incorporado en la fracción peptídica de la MOS y que la eficiencia de su incorporación no se altera debido a la presencia de Ni y viceversa, pero si es afectada cuando el suelo ha sido quemado. INTRODUCCIÓN 1.1. Objetivo de la tesis El fuego es responsable de una de las mayores perturbaciones de los ecosistemas. En particular, produce cambios en la cantidad y calidad de la materia orgánica del suelo (MOS), afectada notablemente por la incorporación de la materia orgánica (MO) pirogénica (PyOM). El nitrógeno orgánico (Norg) como parte de la MOS va a sufrir cambios estructurales por el impacto del fuego dando lugar a la formación de compuestos heterocíclicos aromáticos de nitrógeno (N) o “Black Nitrogen” (BN). La formación del BN afecta a la biodisponibilidad del N y establece competición entre plantas, microorganismos y MOS. El objetivo principal de la presente tesis es dilucidar el impacto del fuego en la inmovilización del N, la cual va a afectar su tasa de degradación y por consiguiente al tamaño y composición de los distintos compartimentos de N del suelo. 1.2. El sistema Mediterráneo 1.2.1. Características y distribución El bosque mediterráneo se desarrolla en regiones con clima mediterráneo, que son: California, el centro de Chile, los países del mar Mediterráneo (Fig. 1.1), la ciudad de El Cabo y el sur de 4 The impact of soil disturbance on N and C in Mediterranean soils Australia, ocupando un 1.5% de la superficie forestal de la tierra (M'Hirit 1999). El clima mediterráneo se caracteriza por veranos secos y prolongados con temperaturas diurnas superiores a los 30 ºC, que reducen la humedad de la materia vegetal a menos de un 5%.(Vélez R 1990), lo que favorece el desarrollo de incendios forestales. El viento también contribuye al traslado de pavesas a otros puntos cercanos del incendio, lo que ayuda a propagar los fuegos. Los montes mediterráneos se caracterizan por su elevada heterogeneidad, lo que se explica por la riqueza florística y la gran variedad climática, geológica, geomorfológica y edáfica de las zonas en las que se desarrollan. Son sistemas con baja resiliencia, es decir tienen una capacidad limitada de respuesta ante las perturbaciones. 1.2.2. Vegetación Parte de la vegetación del bosque mediterráneo ha desarrollado adaptaciones a la incidencia del fuego durante su ciclo reproductivo, como puede ser la apertura de las piñas o el aumento de la tasa de germinación con el calor (Lloret & Vilà 1997). Los pinares constituyen las masas arboladas más extensas del bosque Mediterráneo. El pino carrasco (Pinus halepensis) es el más extendido en las costas de España, Francia, Italia, Grecia, Turquía, Marruecos, Argelia y Túnez. El pino piñonero (P. pinea), el pinaster (P. pinaster) y el laricio (P. nigra) abundan en el occidente de la cuenca, y el brutia (P. brutia) en el oriente. Dichas especies suelen tener un elevado contenido en resinas y aceites esenciales, extremadamente inflamables. Otras especies importantes son la encina (Quercus ilex) y el alcornoque (Q. suber). Fig. 1.1. Mapa de los ecosistemas forestales mediterráneos (verde, bosques mediterráneos; amarillo, otras superficies boscosas; en naranja, límite bioclimático del Mediterráneo). Capítulo 1. Introducción 5 1.2.3. Importancia del fuego El fuego es el mayor factor de degradación natural y destrucción de los bosques y las zonas arboladas de la cuenca Mediterránea. Cada año cerca de 50000 incendios queman una superficie entre 700000 y 1000000 de hectáreas de monte (FAO 2006). Los fuegos conllevan la desaparición o modificación de la cobertura vegetal (Mayor et al. 2016, Tessler et al. 2016), afectan al comportamiento hidrogeomorfológico (Mayor et al. 2007) y las propiedades físicas, químicas, mineralógicas y biológicas del suelo. El grado de dichos efectos dependen principalmente de la severidad del incendio, determinada por su duración y la temperatura que alcanzan (Certini 2005). Los principales efectos en las propiedades del suelo afectado por el fuego son el incremento del pH (Lombao et al. 2015) debido a la desnaturalización de los ácidos orgánicos, cambios en los contenidos del carbono (C) y N (Caon et al. 2014b) como consecuencia de la incorporación de cenizas y material parcialmente quemado (Nocentini et al. 2010), y alteraciones en la calidad de la MOS por el aporte del PyOM. El aporte de cenizas debido a la combustión de la vegetación y hojarasca incrementa los porcentajes de: calcio, magnesio, fósforo y potasio en los suelos quemados (Bodí et al. 2014), lo que favorece la regeneración de la vegetación debido a un aumento de la fertilidad tras un incendio. En general, los fuegos que desarrollan temperaturas inferiores a 200 ºC no afectan a las características del suelo, entre 200 y 460 ºC causan la combustión de algunas sustancias orgánicas y por encima de 460 ºC y tiempo suficiente provocan la combustión del material orgánico y descomposición de los carbonatos (Certini 2005). 1.3. Causas de los incendios El porcentaje de incendios de origen natural (fenómenos meteorológicos) en la cuenca Mediterránea oscila entre un 1 y 5%, mientras que el número y frecuencia de incendios provocados por actividades antrópicas es mucho más elevada (Alexandrian et al.1999). Así, el número de incendios causados por prácticas agrícolas y forestales tienen un peso importante en la zona Mediterránea, oscilando entre el 50 y el 77%. En otros países, Croacia, Grecia y Portugal esta tasa se encuentra entre el 25 y el 47%. 6 The impact of soil disturbance on N and C in Mediterranean soils 1.4. Situación de los incendios forestales en Europa y España Los datos estadísticos del número de incendios y superficie afectada para el período comprendido entre 1980-2015 para los países del sur de Europa (Portugal, España, Francia, Italia y Grecia) se recogen en las Tablas 1.1 y 1.2. (“JRC Technical relports, Forest fire in Europe, Middle East and North Africa 2014”) (Comisión Europea) (http://effis.jrc.ec.europa.eu/media/cms_page_media/40/Forest_fires_in_Europe_Middle_east_an d_North_Africa_2015_final_pdf_JkX4Yl3.pdf). La superficie afectada se estimó en 16000000 ha para dicho período, de la cual el 37% corresponde a España, el país que más superficie tiene quemada para dicho período y el segundo por número de fuegos (31%), después de Portugal (http://forest.jrc.ec.europa.eu/effis/ reports/annual-fire-reports). Tabla 1.1. Número de fuegos en los cinco países miembro del sur de Europa en los últimos 36 años Nº de fuegos Portugal España Francia Italia Grecia* Total 2015 15851 11928 4440 5442 510 38171 % total 2015 42 31 12 14 1 100 1980-1989 7381 9515 4910 11575 1264 34645 1990-1999 22250 18152 5538 11164 1748 58851 2000-2009 24949 18369 4418 7259 1695 56690 2010-2015 18439 12994 3857 5492 1025 41536 1980-2015 18234 15953 4727 9248 1478 48641 Total 656437 538318 170177 332929 53206 1751067 * número de fuegos incompletos desde 2009 Para el caso de España los datos estadísticos han sido recogidos del Ministerio de Agricultura y Pesca, Alimentación y Medio Ambiente para el último decenio (2005-2016) y año (2016) (http://www.mapama.gob.es/es/desarrollorural/estadisticas/Incendios_default.aspx). En España, el total de fuegos forestales ha descendido un 32% en 2016 con respecto a la media del último decenio (2005-2015). En cuanto a la superficie afectada se produjo un descenso con respecto a la media del último decenio del 37% y 34% en la superficie arbolada y forestal, respectivamente. El caso particular de Andalucía para el período (2001-2010), ocupó el quinto y Capítulo 1. Introducción 7 segundo puesto con un 6 y 30% con respecto al número de fuegos y superficie afectada por el mismo para los 10 años. Tabla 1.2. Superficie quemada en los cinco países miembro del sur de Europa en los últimos 36 años Superficie quemada (ha) Portugal España Francia Italia Grecia* Total 2015 64443 103200 11160 41511 7096 227410 % total 2015 28 45 5 18 3 100 1980-1989 73484 244788 39157 147150 52417 556995 1990-1999 102203 161319 22735 118573 44108 448938 2000-2009 150101 127229 22362 83878 49238 432809 2010-2015 92377 98660 8947 59345 29609 288937 1980-2015 105893 164592 24895 107002 45424 447807 TOTAL 3812148 5925323 896216 3852072 1635277 16121036 * número de fuegos incompletos desde 2009 1.5. Zona de estudio El área de estudio se localiza en la sierra de Aznalcóllar (Sevilla), situada entre el Aljarafe sevillano y la Sierra Norte. Parte de esta sierra pertenece a las primeras estribaciones de Sierra Morena. La selección de la zona de estudio se fundamentó en diversos criterios, tales como: i) la historia reciente de incendios en la Sierra de Aznalcóllar, la cual fue afectada por el fuego en los años 1996 y 2004, lo que permitió la selección de puntos de muestreo, no quemados, quemados una vez y quemados dos veces ii) la existencia de una amplia documentación sobre estudios previos en dicha zona (Knicker et al. 2013), lo cual nos facilitaría comparar los resultados obtenidos con la información ya recopilada, transcurridos 7 años después del último incendio iii) por ser un área con diversas zonas protegidas principalmente, El Corredor Verde del Guadiamar, que es espacio natural de interés comunitario) y iv) representar al bosque mediterráneo, compuesto principalmente por encinas (Quercues ilex), alcornoques (Q. suber) y jaras (Cistus). Las muestras utilizadas en el presente estudio proceden de suelos que han sido quemados una vez (1996), dos veces (1996 y 2004) o ninguna (suelo control) bajo un bosque de pinos o Quercus. Posteriormente, se seleccionaron dos suelos (suelo control y quemado una vez bajo 8 The impact of soil disturbance on N and C in Mediterranean soils Quercus) para un experimento de incubación durante 0.5, 1, 5, 8 12 y 16 meses llevado a cabo en invernadero. Este experimento se inició, con la producción de Lolium perenne enriquecido en 15N y no enriquecido para obtener los materiales de partida: MO fresca y PyOM enriquecida con 15N y sin enriquecer. El PyOM se produjo a 350ºC durante 8 minutos en presencia de oxígeno (O). El experimento de incubación consistió en enmendar el suelo control y quemado una vez con diferentes enmiendas: i) 15N inorgánico (Ni), ii) MO enriquecido en 15N (15N-MO), iii) MO más 15Ni, iv) 15N-MO más Ni, v) PyOM enriquecido 15N (15N-PyOM), vi) PyOM más 15Ni y vii) 15N-PyOM más Ni. 1.6. Impacto del fuego en el Ciclo del Nitrógeno El N es un macronutriente que se encuentra en bajas concentraciones en el suelo Mediterráneo variando desde 0.1 hasta 0.7% (Aponte et al. 2010), López-Serrano et al. (2016) encontró valores medios de 0.3%, lo que convierte al N en un nutriente limitante para la fertilidad del suelo y, el crecimiento de biomasa. La cantidad de N que está disponible para los microorganismos es uno de los importantes factores que controlan la biodegradación de la materia orgánica en el suelo de los bosques. Es el tercer elemento más abundante en plantas (10-30 g kg-1) (Rengel & Marschner 2007), y un componente clave de los amino ácidos, ácidos nucleicos y clorofila. El N es absorbido desde la solución del suelo en mayores cantidades que otros nutrientes y es asimilado por plantas y microorganimos en forma de NO3y NH4+, urea y aminoácidos (Bacon 1995, Hodge et al. 2000). Nannipieri y Eldor (2009) afirman que los microorganismos suelen ser mejores competidores que las plantas por el N. El ciclo del N terrestre comprende al N biológicamente activo presente en el suelo, plantas y animales y al N inerte que se encuentra en la litosfera y atmósfera. Por lo tanto, la entrada de N y su pérdida del sistema suelo-planta-animal determina la dinámica de su ciclo, que se ve afectada, entre otras perturbaciones, por el fuego. Durante los incendios forestales, el combustible (restos vegetales, residuos animales, exudados de raíces, rizodepósitos, necromasa, MOS y las excretas animales) sufre un proceso de combustión por ello que parte del C y Norg son convertidos en dióxido de C (CO2), monóxido de C (CO), metano (CH4), óxidos de N (NOx) y amoniaco (NH3) que son liberados a la atmósfera. El CH4, CO2, óxido nitroso (N2O) figuran entre los principales gases de efecto invernadero (Gärdenäs et al. 2011) que contribuyen al calentamiento global. Por lo tanto la primera respuesta del N ante un incendio es la disminución debido a su volatilización (Johnson & Turner 2014). Capítulo 1. Introducción 9 Ésta pérdida de N del suelo por volatilización se puede enmascarar por la fijación atmosférica del nitrógeno (N2) a NH3 llevado a cabo mediante procesos biológicos (microorganimos especializados) (Johnson et al. 2008) o industriales (ciclo de Haber-Bosch, donde el H2 se combina con el N2 para obtener NH3 a 500 ºC y 200 bares), rayos (el monóxido de N, NO es convertido fotoquímicamente en NO2, ácido nitroso (HNO2), ácido nítrico (HNO3) y otros subproductos y mediante el proceso de deposición (seca o húmeda) del amonio (NH4+), NH3 y nitrato (NO3-) procedentes de partículas de la atmósfera en la superficie terrestre. Ante un fuego de intensidad moderada a alta la MOS puede ser consumida por volatilización y el contenido de C y N disminuir (Heydari et al. 2016). Sin embargo, en un incendio de baja intensidad, el contenido de NH4+, NO3y N total aumenta (Alexis et al. 2010, Augustine et al. 2014), tanto por el aporte de formas de N inorgánicas como por la incorporación del PyOM (Knicker 2004). El NH4+ es un producto directo de la combustión y es absorbido en los sitios de intercambio catiónico del suelo (minerales de arcilla, MOS), mientras que el NO3formado en la nitrificación después del incendio puede ser lixiviado (Certini 2005). Durante un incendio la mineralización aumenta (Johnson et al. 2008), aunque también se ha descrito la tendencia contraria debido a la disminución de la actividad microbiana por la disminución de nutrientes y disponibilidad del C (Wang et al. 2012). En la misma línea Rivas et al. (2012) afirman que tanto la amonificación, como la nitrificación y la mineralización del N fueron negativos indicando inmovilización del N. Sin embargo (Fultz et al. 2016) justifican la disminución de la amonificación debido a la pérdida de Norg de la fracción libre, mientras que la nitrificación no se alteró debido a un cambio en la comunidad microbiana desde hongos heterotróficos hacia bacterias gran negativas autotróficas. El PyOM contiene una porción de nitrógeno Norg cuyo origen es peptídico (Knicker 2007; Nannipieri & Eldor 2009) y se encuentra en forma de heterociclos aromáticos, BN (Knicker 2010), el cual podría representar el principal mecanismo de secuestro del N en el suelo (Knicker 2011). Estos nuevos compuestos formados durante el fuego se esperaba que fuesen estables frente a la degradación química, biológica y térmica (Almendros et al. 1990, Knicker et al. 1996) y de ésta manera contribuyesen a la materia orgánica refractaria del suelo (Knicker 2004) y por lo tanto afectando al ciclo del N. Esta teoría es apoyada por Prieto-Fernández et al. (2004) quienes encontraron que después de un incendio forestal el porcentaje de N residual obtenido por hidrolisis aumentó un 45%, debido a un incremento en la aromaticidad de la MO, como 10 The impact of soil disturbance on N and C in Mediterranean soils consecuencia de la pérdida de los grupos oxigenados (OH and COOH), lo que hace que los compuestos sean resistentes a la hidrólisis ácida. La inaccesibilidad del BN para los microorganismos va a establecer una competición por el N entre, microorganismos, plantas y MOS, por lo que la productividad de las plantas a corto plazo estaría influenciada por la disponibilidad del N para ser absorbido por estas. Estudios recientes han puesto de manifiesto que la biodisponibilidad del BN es mayor de lo que se creía (de la Rosa & Knicker 2011, López-Martín et al. 2016), aunque, el uso de este N por las plantas sigue siendo objeto de estudio. 1.7. Cambios de la materia orgánica inducidos por el fuego La combustión completa del combustible orgánico del suelo resulta en la producción y oxidación de gases como CO2, CO, CH4 o NO2 y la formación del residuo mineral (ceniza). Sin embargo durante los incendios forestales, la disponibilidad de O es limitada y por tanto la combustión es incompleta,por lo que parte de la MOS es parcialmente quemada, dando lugar a necromasa (PyOM). Knicker et al. (2008a) definen el PyOM como una estructura derivada de lignina, furanos, anhidroazúcares y principalmente piranonas derivadas de la celulosa. En el caso de residuos quemados con alto contenido en N, el PyOM contendrá N heteroaromático (BN) derivado de péptidos. Durante un incendio en ausencia de O los polipéptidos pueden descarboxilarse, desaminarse y deshidratarse hasta dipéptidos, los cuales pueden volverse a deshidratar para formar la 2,5 dicetopiperazine. A temperaturas superiores a 200 ºC, de acuerdo con Baldock and Smernik (2002), la MOS sufre deshidratación y formación de estructuras cíclicas insaturadas a la vez que las macromoléculas son fragmentadas en compuestos de más bajo peso molecular (de la Rosa et al. 2012). Al aumentar el tiempo de quema se producen reacciones de descarboxilación y desmetilación (Almendros et al. 2003). Con el aumento de la temperatura las estructuras alquílicas con O (derivadas principalmente de carbohidratos, como celulosa) son convertidas en estructuras aromáticas o de tipo furano (Baldock & Smernik 2002). Durante un fuego se crean nuevas estructuras como resultado de las reacciones de deshidratación y ciclación de cadenas alifáticas con amino grupos o NH3 (Almendros et al. 2003) que dan lugar al BN. Estas estructuras son de tipo pirrol, piridina o indol (Fig. 1.2) (Knicker et al. 1996, Knicker et al. 2005). El BN también se propone como un indicador de la existencia de un fuego reciente ya que sus estructuras son encontradas en suelos afectados por incendios mientras Capítulo 1. Introducción 11 que en suelos donde los microorganimos degradaron la MOS no fueron observadas (Knicker & Lüdemann 1995a). Estos cambios químicos en la MOS asociados al incremento de temperatura reducen la disponibilidad microbiana para descomponerla. Sin embargo el modelo de PyOM propuesto por Knicker (2007) define el PyOM como una mezcla heterogénea de biomacromoléculas alteradas por el fuego que contiene sustituciones de N, O y S, hecho que hace que la estructura del PyOM presente zonas de ataque biótico (ej: oxidación de la lignina por hongos) y abiótico, en lugar de una estructura grafítica compuesta principalmente por pequeños anillos poliaromáticos condensados. Recientes estudios (de la Rosa & Knicker 2011) encontraron incorporación de N procedente de material quemado enriquecido (15N-PyOM) en un experimento en el que se hizo crecer césped en suelo mezclado con el material quemado enriquecido en 15N. Este resultado se confirmó con la detección de estructuras tipo pirrol en el espectro de resonancia magnética nuclear de 15N de dichas plantas, confirmándose que el BN presenta baja recalcitancia a la degradación biológica. Este hallazgo es apoyado por López-Martín et al. (2016) quienes evidenciaron la incorporación del N derivado del BN de material quemado enriquecido con 15N en la fracción peptídica de la MOS. De acuerdo con (González-Pérez et al. 2004) los cambios más importantes que sufre la MOS durante un incendios son: i) reducción de la solubilidad debido a la perdida de los grupos oxigenados externos, ii) reducción de la longitud de cadena de los ácidos graso, alcoholes y otros Fig. 1.2. Estructuras químicas de los compuestos del “Black Nitrogen” formados a partir de péptidos y proteínas formados durante el proceso de quema (Gärdenäs et al. 2011). 18 The impact of soil disturbance on N and C in Mediterranean soils Velez R (1990) Los incendios forestales en el mediterráneo: Perspectiva regional, Unysilva, 41, (162). http://www.fao.org/docrep/t9500s/t9500s00.htm Wang Q, Zhong M, Wang S (2012) A meta-analysis on the response of microbial biomass, dissolved organic matter, respiration, and N mineralization in mineral soil to fire in forest ecosystems. Forest Ecology and Management 271:91-97 Capítulo 2. Objetivos específicos de la Tesis Chapter 2. Specific objectives of the Thesis OBJETIVOS ESPECIFICOS DEL TRABAJO Después de un incendio la composición y el tamaño de las reservas de materia orgánica del suelo se ven afectadas, alterando consecuentemente el ciclo del nitrógeno (N). En términos generales, los mecanismos de estabilización del nitrógeno y su relación con el ciclo del carbono son poco conocidos, lo cual dificulta la predicción del impacto a largo plazo de la inmovilización del nitrógeno en el potencial secuestro de carbono. En base a estas consideraciones, las hipótesis de partida de la presente tesis fueron: 1. La estabilización del nitrógeno a largo plazo en el suelo, está determinado por los mecanismos de control del reparto de nitrógeno entre la biomasa, asociaciones órgano-minerales y las formas orgánicas protegidas, jugando un papel menos importante la inmovilización abiótica del nitrógeno. 2. La alteración del nitrógeno disponible en el suelo por incendios disminuye la entrada de materia orgánica o el nitrógeno añadido procedente de fertilizantes, modificando el reparto de nitrógeno entre la biomasa viva y las fracciones de materia orgánica del suelo. Este hecho afectará al tamaño y composición de los diferentes compartimentos de la materia orgánica del suelo y al potencial de secuestro de carbono y nitrógeno. Como consecuencia, se verá alterada la capacidad del suelo para actuar como sumidero de gases de efecto invernadero. 22 The impact of soil disturbance on N and C in Mediterranean soils Teniendo en cuenta estas hipótesis, el objetivo principal de este estudio es mejorar el conocimiento cuantitativo del papel que tiene el fuego como perturbación del suelo en los procesos competitivos de secuestro del N a corto y largo plazo, así como el uso del nitrógeno orgánico e inorgánico por las plantas y los microorganismos frente a la estabilización del N a largo plazo en la materia orgánica del suelo. Los objetivos secundarios son: 1. Muchas de las áreas en la region Mediterránea afectadas por el fuego son montañosas, por lo que la influencia del relieve en las propiedades de suelos y en la composición de la materia orgánica del suelo (MOS) debe ser estudiado. Además, teniendo en cuenta que, la distribución del combustible puede no ser homogéneo, se pueden producir variaciones considerables en la calidad y cantidad del material pirogénico (PyOM). Por lo tanto, los enfoques estadísticos (Li 2010) deben tener en cuenta dicha inohomegeidad. Por otro lado, muchos de estos análisis requieren un número elevado de réplicas, lo cual raramente es factible debido a la intensa labor o el alto coste de las técnicas analíticas avanzadas, aunque podrían aportar información necesaria para el mejor conocimiento de la dinámica ecológica y de las medidas a largo plazo para la recuperación de zonas alteradas. Una de estas técnicas es la espectroscopía de RMN en estado sólido de 13C. Para estudiar las propiedades de MOS de nuestra área de estudio y evaluar la fiabilidad de nuestros datos, el primer paso de nuestro trabajo fue evaluar el efecto del relieve y la distribución previa del combustible en las propiedades y composición de la MOS en suelos afectados y no por el fuego en una zona montañosa de Sierra de Aznalcóllar, Andalucía (España). (Capítulo 3). 2. Los estudios de los últimos años indicaron que las condiciones medioambientales incuestionablemente determinan si el PyOM permanece en el suelo o es transformado rápidamente en residuos orgánicos, los cuales no pueden ser diferenciados de la MOS formada sin el impacto del fuego (Knicker 2011). Sin embargo, los mecanismos implicados aún están lejos de ser comprendidos. E incluso menor es el conocimiento con respecto al Black Nitrogen (BN). A parte de las propiedades protectoras de la estructura molecular y los mecanismos químicos, los compartimentos de MOS están protegidos de la descomposición vía mecanismos físicos. Esto último limita la accesibilidad de la materia orgánica (MO) debido a su interacción con la fase mineral. Por lo tanto, el fraccionamiento por densidad y de tamaño de partícula de la MOS fue usado para evaluar el papel de la asociación MOS-mineral en la supervivencia del PyOM en los suelos. (Capítulo 4). 23 Capítulo 2. Objetivos del trabajo 3. Experimentos previos de corto período de tiempo usando PyOM enriquecido en 15N como enmienda del suelo, mostraron que el 15N derivó del BN (Hilscher & Knicker 2011) y por tanto que puede ser usado para el crecimiento de la planta (de la Rosa & Knicker 2011). Para obtener más información de la biodisponibilidad, fue llevado a cabo un experimento con macetas con suelos de zonas afectadas y no por el fuego en Aznalcóllar, que fueron enmendadas con PyOM u MO enriquecida en 15N tras plantar semillas de Lolium perenne. Se controló la incorporación de 15N en el suelo y en material de planta, con el objetivo de obtener más información acerca el papel del BN dentro del ciclo del N. (Capítulo 5). 4. Es ampliamente conocido que el N es un nutriente limitante para el crecimiento de las plantas y los microorganismos, de ahí que un cambio en la disponibilidad del N para los organismos del suelo debido a la alteración de la calidad del material que se incorpora puede tener un gran impacto en la productividad biológica de un suelo. Sin embargo, a pesar de la importancia del N en el ecosistema del suelo poco se sabe sobre como la transformación de las proteínas en BN afecta el ciclo del N, en suelos afectados por fuego. Por ello, se intentó incrementar nuestro conocimiento sobre el papel del BN en la producción de la biomasa microbiana, analizando la calidad y cantidad de los aminoácidos extraíbles procedentes del experimento en macetas enmendados con restos de plantas enriquecida en 15N y su PyOM correspondiente. El control del 15N incorporado en los aminoácidos en presencia y ausencia de N inorgánico fue diseñado para revelar los posibles efectos negativos o positivos relacionados con el uso del N para la producción de la biomasa (Capitulo 6). Referencias de la Rosa JM, Knicker H (2011): Bioavailability of N released from N-rich pyrogenic organic matter: An incubation study. Soil Biology and Biochemistry 43, 2368-2373 Hilscher A, Knicker H (2011): Carbon and nitrogen degradation on molecular scale of grassderived pyrogenic organic material during 28 months of incubation in soil. Soil Biology and Biochemistry 43, 261-270 Knicker H (2011): Pyrogenic organic matter in soil: Its origin and occurrence, its chemistry and survival in soil environments. Quaternary International 243, 251-263 Li Y (2010): Can the spatial prediction of soil organic matter contents at various sampling scales be improved by using regression kriging with auxiliary information?, 159. Elsevier, Amsterdam, PAYS-BAS, 13 pp Publicaciones científicas Capítulo 3. Capítulo 4. Capítulo 5. Capítulo 6. Papers Chapter 3. Chapter 4. Chapter 5. Chapter 6. Capítulo 3. Variabilidad de la calidad y la cantidad de la materia orgánica en suelo afectado por múltiples incendios Chapter 3. Variability of the quality and quantity of organic matter in soil affected by multiple wildfires Table 3.1. Localization and vegetation cover of the sampled Cambisols in fire-affected areas of the Sierra de Aznalcóllar, Southern Spain with the indication of the respective sampling method. For the first, five spots within a square of 15 m side length were probed and compared. For the second approach composite samples were obtained for each site Sample Pre-fire cover Post-fire cover Nº fires Coordinates Altitude (m) Slope (º) Sampling method AZPB1 Pinus Pinaster Pinus Pinaster 1 N 37º 34´ - W 06º 19´ 267 30 1 AZQB1 Quercus suber Quercus suber 1 N 37º 36´- W 06º 20´ 330 30 1 AZPDB1 Pinus Pinaster Pinus Pinaster 2 30 1 AZQU2-1 Quercus suber - 0 N 37º 30´ - W 6º 19´ 158 30 2 AZQU2-2 Quercus ilex - 0 N 37º 29´ - W 6º 20´ 104 30 2 AZQU2-3 Q. suber – Q. ilex - 0 N 37º 32´ - W 6º 15´ 184 30 2 AZQB2-1 Quercus ilex Q. ilex - P. Pinaster 1 N 37º 30´ - W 6º 19´ 308 30 2 AZQB2-2 Pinus Pinaster Pinus Pinaster 1 N 37º 34´- W 6º 20´ 258 30 2 AZQB2-3 Quercus suber Quercus suber 1 N 37º 33´ - W 6º 19´ 260 30 2 AZQDB2-1 Q. suber – P. Pinaster Cistaceas 2 N 37º 36´ - W 6º 20´ 396 30 2 AZQDB2-2 Quercus suber Quercus suber 2 N 37º 35´ - W 6º 22´ 300 30 2 AZQDB2-3 Quercus suber Quercus suber 2 N 37º 35´ - W 6º 23´ 357 30 2 35 Chapter 3. Variability of organic matter in fire affected soils The second sampling approach was used during a sampling campaign in December 2011.Here, two sites under Quercus suber and one under Pinus pinaster and maquia (Cistaceae, Genista scorpius, Arbutus unedo), which were burnt only in 2004 (AZQB2-1, AZPB2-2, AZQB2-3) and three under the same vegetation but burnt both in 1996 and 2004 (AZQDB2-1, AZQDB2-2, AZQDB2-3) were probed. Unburnt soils under Quercus (AZQU2-1, AZQU2-2, AZQU2-3) were chosen as reference. At all sites, the litter layer was removed manually and soil material was taken from the first 2 cm of the A horizon. After drying at 40 ºC and removal of the fine roots by hand-picking, the soils were sieved through a 2 mm mesh and stored for further analysis. 3.2.2. Determination of chemical and physical properties of the soils The pH of the soils was determined in the aqueous phase of a soil suspension (2.5 g of dry fine earth with 10 ml of deionized water) after the solid phase had settled for 30 min. The electrical conductivity was measured in the filtered supernatant. Total C and total N contents were measured in duplicate via dry combustion using an elemental analyzer (Carlo-Erba EA-1108-CHNS). Due to the acidity of the soils, the presence of inorganic C can be excluded. Inorganic nitrogen form (NO3-) was extracted with 2M of KCl and determined calorimetrically (Keeney & Nelson 1982) using an Jenway 6315 Spectrophotometer at 410 nm. During the determination of the ammonium contents, volatilization of the ammonia was indicated. Therefore, the obtained values were not uses for further analysis. 3.2.3. Demineralization of soil In order to remove paramagnetic ions and concentrate the SOM prior to solid-state NMR analysis, samples were treated with 10% (v/v) hydrofluoric acid (HF) according to Gonçalves et al. (2003). Earlier analysis of the impact of this treatment on the SOM composition in soils of the same area clearly evidenced that no preferential extraction is induced by this demineralization technique (Knicker 2011b). Differences between the intensity distribution of the solid-state 13C NMR spectra before and after the demineralization were shown to be caused by the selective impact of paramagnetics on the cross-polarization efficiency of certain C groups in the untreated samples. 36 The impact of soil disturbance on N and C in Mediterranean soils For the demineralization, approximately 10 g of the dried soil were weighed into polyethylene bottles to which 40 ml of the HF solution was added consecutively. Then, the closed bottles were shaken for 2 h at 250 rpm and centrifuged for at 3000 rpm for 10 min. Subsequently, the supernatant was carefully removed and discarded. This procedure was repeated four times. After that, the sediments were cleaned by rinsing with 50 ml of deionized water until a pH > 5 was yielded. Finally, the samples were freeze-dried. Based on the fact that, on average, only 10% of the organic C (Corg) was lost during the treatment (Table 3.2), significant alteration of the organic matter composition due to selective removal of specific C groups can be excluded. Table 3.2. Carbon loss after 10% HF treatment of fire-affected and unaffected Cambisols obtained from the Sierra de Aznalcóllar Sampling site C loss (%) AZPB1 6.7 ± 3.1 AZQB1 4.9 ± 2.4 AZPDB1 8.8 ± 3.5 AZQU2-1/2/3 25.6 ± 2.7 AZQB2-1/2/3 8.9 ± 2.3 AZQDB2-1/2/3 5.5 ± 3.7 3.2.4. Chemical oxidation with acid potassium dichromate To estimate the PyOM contents of the samples, approximately 0.6 mg of demineralized samples was oxidized according to Knicker et al. (2007) with 20 ml of 0.1 M K2Cr2O7 and 2 M H2SO4 solution but in an oven during 6 h at 60ºC. The dichromate solution was changed as soon as the color of the solution turned to greenish, which indicated its complete reduction. After 6 h, the samples were centrifuged, the supernatant was discarded, and the residues were rinsed until the solution reached a pH value between 5 and 6. The organic carbon remaining in the oxidation residue (chemical oxidation resistant carbon, COREC) was quantified by elemental analysis and further characterized by solid-state 13C NMR spectroscopy. The aromatic C content of COREC (CORECarom) was assigned as PyOM. Its contribution to the total organic C was determined by multiplying the relative intensity of the chemical shift region of aryl C of the solid-state 13C NMR spectra of the COREC samples with the percentage of C surviving this treatment. Earlier studies demonstrated that not all PyOM constituents resist this harsh oxidation method and that the chemical recalcitrance of PyOM constituents resist this hard oxidation method and that the 37 Chapter 3. Variability of organic matter in fire affected soils chemical recalcitrance of PyOM depends on its humification state. Bearing in mind that, in the study area, the last fires were within the last 20 years before sampling, we considered the present PyOM as young. Accordingly, we multiplied the amount of CORECarom with the correction factor fyoung = 2.4 which was suggested by for weakly humified PyOM (Knicker et al. 2008b). 3.2.5. Solid-state 13C NMR spectroscopy The solid-state 13C NMR spectra of HF-treated soil samples and the chemically oxidized residues were acquired with the cross-polarization (CP) magic-angle spinning (MAS) technique on a Varian 7.05 T Unity Inova NMR Spectrometer (300 MHz) and a Bruker Avance HD III (400 MHz) using spinning speeds of 8 kHz and 14 kHz, respectively. All spectra were obtained with a ramped 1H pulse during the Hartmann-Hahn contact of 1 ms, and delay times between 0.3 and 0.5 s. The chemical shift scale was referenced to tetramethylsilane using glycine (COOH:176.04 ppm). Depending on C content of the sample, between 2300 and 55500 scans were accumulated for material treated with HF and between 46100 and 118500 scans for the COREC samples. After Fourier transformation, the quantification of the spectra was performed by integration of signal intensity in the following chemical shift regions: alkyl C (45 to 0 ppm), O/N-alkyl C (110 to 45 ppm), aryl C (160 to 110 ppm), carbonyl C (245 to 160 ppm) (Knicker 2011b). For the determination of CORECarom, the aryl C region was expanded to 160 to 90 ppm since signals of anomeric C commonly attributed to the region between 110 and 90 ppm should be absent after efficient oxidation of labile C with acid dichromate solution. Owing to insufficient averaging of the chemical shift anisotropy at a spinning speed of 8 kHz and 14 kHz at magnetic fields of 7.05 T and 9.40 T, spinning side bands of the aromatic C signal (140 to 110 ppm) occurred at a distance corresponding to the frequency of the spinning speed at both sides of the parent signal (300 to 275 ppm and 0 to -50 ppm). They were considered by adding their intensities to that of the parent signal as described in Knicker et al. (2005b). One spinning side band of the carboxyl C signal is found in the region 325 to 300 ppm. Assuming that the second side band for carboxyl C between 0 and 45 ppm is equal in size, the integral of the side band between 325 and 275 ppm was doubled and added to the carboxyl signal (160 to 220 ppm), but subtracted from the intensity of the alkyl C region (0 to 45 ppm). 38 The impact of soil disturbance on N and C in Mediterranean soils 3.2.6. Statistical analysis All statistical analyses were performed using STATGRAPHICS plus 5.1. For the comparison of the chemical soil properties, the SOM quality and the PyOM contents of the soils probed with method 1 and 2 one-way analysis of variance (ANOVA) were used .Only the N and NO3contents of the composite samples were treated with the Kruskal Wallis test since they did not have normal distribution. The differences between means were tested by applying the Tukey´s honestly significant difference (HSD) test at 95% confidence level. 3.3. Results and discussion 3.3.1. pH The pH of the samples from method 1 varies between 4.7 and 6.3 and in the case of method 2 from 4.3 to 6.1, with standard errors between 0.1 and 0.3 and 0.1 to 0.2, respectively (Table 3.3). Those values are in the range typically reported for soils of this area (Núñez & Recio 2007). Note that the sampling method had no noticeable impact on the size of the error. Using the Tukey test, the samples AZPB1 and AZQB1 were statistically different from AZPDB1. Also, unburnt AZQU2-1/2/3 can be statistically distinguished from the group, AZQB2-1/2/3 and that includes AZQDB2-1/2/3. However, comparing the two methods, no clear relationship between the number of fire events and pH can be deduced, although it is expected that fire leads to an increase of the soil pH (Gómez-Rey & González-Prieto 2014, Xu et al. 2012). It seems that 7 years after the last fire the liming effect was neutralized and local heterogeneity with respect to parent rock and vegetation cover had a greater impact on soil pH than the input of charcoal. 3.3.2. Electrical Conductivity (EC) The electrical conductivity of a soil can be used as an indicator of the availability of plant micronutrients. After fires, this value commonly increases due to the accumulation of ash (Inbar et al. 2014). At our study sites, 7 years after the last fire, electrical conductivity (EC) values between 414 and 464 μS cm-1 for the samples of method 1 and between 350 and 467 μS cm-1 for those of method 2 with a standard error varying from 26 to 45 μS cm-1 and 9 to 52 µS cm-1 were determined (Table 3.3). Whereas the sites of method 1 cannot be distinguished with respect to EC, the unburnt sites of the sample set of method 2 shows lower values than all burnt soils. 39 Chapter 3. Variability of organic matter in fire affected soils Comparable to the measurements of the pH, we could not detect a major difference in the magnitude of the standard error resulting from the sampling methods. 3.3.3. C and N concentrations With the exception of the material from site AZPB1, the mean organic C contents of the soils range between 57.2 and 58 mg C g soil-1, indicating that the last fire event had no lasting impact on the SOM concentration of those soils. The soil of AZPB1 showed not only higher mean C contents but also higher standard errors then the others (Table 3.3). This is in line with a low homogeneity of the area after considerable post-fire restoration activities such as the removal of tree residues, terracing, and replanting of Leguminosea. Based on our ANOVA, the soils in AZPB1 had significantly higher Corg concentration than AZQB1 and AZPDB1 (p=0.045, Table 3.3). Preparing a box-plot diagram (Fig. 3.2a) or a media and standard error plot (Fig. 3.2b) revealed that two subsamples from AZPB1, with 35 and 180 mg C g soil-1, are extreme values, most likely caused by the inhomogeneity of the relief. The sample with the lowest C content was collected at a slope where almost no vegetation grew, whereas the other one was obtained in an organic matter accumulation zone at the end of the slope. In the case of composite samples obtained with method 2, no differences were found. Differences with respect to C distributions have been reported for other hilly regions (Novara et al. 2011, Rumpel et al. 2006). Table 3.3. pH (H2O), electrical conductivity (EC), Corg and N contents and nitrate concentration in fire affected and unaffected Cambisols of the Sierra de Aznalcóllar and the respective standard errors Sampling site Nº fires pH EC (μS cm-1) Corg (mg g soil-1) N (mg g soil-1) NO3- (mg kg soil-1) Method 1 AZPB1 1 6.3 ± 0.3 a 442 ± 26 ns 130.4 ± 17.4 a 2.9 ± 0.7 ns 4.5 ± 0.5 ns AZQB1 1 5.8 ± 0.1 a 414 ± 34 ns 57.2 ± 11.7 b 3.6 ± 0.3 ns 5.7 ± 0.6 ns AZPDB1 2 4.7 ± 0.1 b 464 ± 45 ns 58.0 ± 4.6 b 2.8 ± 0.3 ns 6.2 ± 0.4 ns Method 2 AZQU21/2/3 0 6.1 ± 0.2 a 350 ± 15 b 55.1 ± 5.0 ns 3.8 ± 0.1 a 2.2 ± 0.0 ns AZQB21/2/3 1 4.3 ± 0.2 b 467 ± 52 ab 57.0 ± 2.1 ns 2.4 ± 0.2 b 2.2 ± 0.2 ns AZQDB21/2/3 2 4.6 ± 0.1 b 463 ± 9 a 45.4 ± 8.1 ns 1.8 ± 0.3 b 2.9 ± 0.2 ns Values followed with different letters within the same column are significantly different at p < 0.05 by Tukey´s Honestly Significant Difference (HSD) test. ns: no significant differences. 40 The impact of soil disturbance on N and C in Mediterranean soils Considering the size of the error for AZPB1, it can be confirmed that topographical inhomogeneity greatly affects Corg contents, which are better averaged by the sampling method 2. However, this method may be more appropriate if general patterns are needed and sampling method 1 for studies focusing on the variability of the soil in the same area. Total N values showed no statistical difference base on ANOVA, yielding a p-value of 0.555, for the samples obtained with method 1 (Table 3.3). For the set collected with method 2, we were able to identify two groups discerning the unburnt soils from the single and double burnt sites (Table 3.3). After and during low-intensity burning, biological and non-biological processes transform organic N into more available ammonium and nitrate N (Prieto-Fernández et al. 2004). Whereas ammonium is a direct product of combustion, nitrate forms after nitrification. Destruction of the vegetation and combustion of allelopathic compounds, such as terpenes and phenols, promotes the increase of nitrifiers and the transformation of available ammonium to nitrate (Andersson et al. 2004). However, in the soils from the Sierra de Aznalcóllar, higher nitrate levels were not maintained after the 7 years of post-fire recovery. This is in line with previous findings of Carter and Darwin Foster (2004) who highlighted that just after the fire the concentration of available N forms (NO3-, NH4+) increased, but that it decreased again with prolonged post-fire recovery time. a) b) Fig.3.2. a) Box-plot of the values measured for the C contents in fire-affected and unaffected Cambisols from the Sierra de Aznalcóllar, by the method 1 and 2 (single and composite sample). The box represents the interquartile range (IQR= Q3-Q1), the horizontal line inside of the box is the median and the whiskers are extended until 1.5*IQR. The outlier was no included in the statistical analysis (in the diagram is represented with a filled square). b) Medium and standard error of C content for fire-affected and unaffected area and the method used during sampling. 41 Chapter 3. Variability of organic matter in fire affected soils According to the statistical analysis, neither the samples obtained in June 2011 nor those collected in December 2011 showed differences which could be associated with the number of fire events. The standard errors for methods 1 and 2 range between 0.4 to 0.6 mg kg soil-1and from 0.0 to 0.2 mg kg soil-1respectively, showing that the second approach yields less variability. 3.3.4. Chemical characterization of soil organic matter With the exception of those of the HF-treated soils from site AZPB1, all solid-state 13C NMR spectra show their main intensity in the chemical shift region of O/N-alkyl C (110 to 45 ppm) (Fig. 3.3). The averaged contributions of each chemical shift region to the total spectrum are listed in Table 3.4. Among all soils, the SOM of unburnt soil AZQU2-1/2/3 contains the highest percentage of O/N-alkyl C (45% ± 2%) and the lowest content of aryl C (16 % ± 0%) which is in the typical range for forest soils unaffected by fire (Baldock et al. 1990, Knicker 2011b, Wilson 1981) and consistent with the reported absence of PyOM in soils (Ignacio Villar Movellan, CEDEFO, Sevilla, personal communication). The clear signal in the chemical shift region of Oaryl C (160 to 140 ppm) points to contributions of lignin residues, the methoxyl C of which contributes adds to the intensity between 60 and 45 ppm. Analyzing the set obtained with method 2 allowed the differentiation between the unburnt sites (AZQU2) and the group of burnt soils composed of AZQB21 and AZPDB2. No discrimination was achieved with respect to the number of fire events in the area. Within the samples obtained with method 1, the chemical composition of those from AZPB1 is significantly different from that of the other two sites. The high aryl C concentration of 46 % ± 2% of t AZPB1 together with minor intensity contributions to the O-aryl C region confirms a considerable abundance of charred residues. We can only speculate if this is due to a recent local fire or caused by the input of charcoal residues in the time frame of the recuperation measures. All other burnt sites reveal aryl C contents between 23 and 26 % of the total organic C. These amounts are clearly below those expected for fire-affected soils (Certini et al. 2011), but their aryl C contents are significantly higher than that of the unburnt site AZQU2-1/2/3. Thus, although the fire had no lasting effect on the quantity of the SOM, its quality still showed the typical pattern of elevated charcoal input. However, no major difference was revealed between the SOM composition of the single and double burnt sites for AZQB1 and AZPDB1 collected with method 1 and between AZQB2-1/2/3 and AZQDB-1/2/3 sampled with method 2. This suggests that the second fire did not lead to an extra gain of charred material into the SOM. The statistical analysis of the intensity in the chemical shift region attributed to O/N-alkyl C joins the site AZQU2-1/2/3 (unburnt) and AZQB1 into one group 42 The impact of soil disturbance on N and C in Mediterranean soils Fig. 3.3. Solid-state 13C NMR spectra of HF-treated fire-affected and unaffected Cambisols obtained with sampling method 1 (a) and 2 (b) from the Sierra de Aznalcóllar. b) AZQU2-1 AZQU2-2 AZQU2-3 200 100 ppm 0300 -100 29 55 73 104 130 152 175 O/N Alkyl C Alkyl C Carboxyl C Aryl C 200 100 ppm 0300 -100 AZQB2-1 AZQB2-2 AZQB2-3 29 55 73 104 130 152 174 200 100 ppm 0300 -100 AZQDB2-1 AZQDB2-2 AZQDB2-3 29 55 73 104 130 152 173 144 O/N Alkyl C Alkyl C Carboxyl C Aryl C O/N Alkyl C Alkyl C Carboxyl C Aryl C AZPB1b AZPB1c AZPB1d AZPB1e 200 100 ppm 0300 -100 AZPB1a 30 55 73 104 127 151 173 O/N Alkyl C Alkyl C Carboxyl C Aryl C * * AZPDB1b AZPDB1c AZPDB1e AZPDB1a 200 100 ppm 0300 -100 AZQB1a AZQB1c AZQB1d AZQB1e AZQB1b 30 55 73 104 130 151 172 O/N Alkyl C Alkyl C Carboxyl C Aryl C 29 55 73 103 131 150 174 200 100 ppm 0300 -100 AZPDB1d O/N Alkyl C Alkyl C Carboxyl C Aryl C a) Figure3 Table 3.4. Relative intensity distribution of the solid-state 13C NMR spectra of HF-treated fire-affected and unaffected Cambisols of the Sierra de Aznalcóllar, Southern Spain, the alkyl C/O/N-alkyl C ratios and the contribution of CORECarom to the initial C of the samples Sampling site Carboxyl C Aryl C O/N-Alkyl C Alkyl C Alkyl C/ O/NAlkyl C CORECarom % of initial C Method 1 AZPB1 9.1 ± 0.4 ns 46.0 ± 1.5 a 28.8 ± 1.1 b 16.1 ±1.0 b 0.56 ± 0.04 36.4 ± 2.2, a AZQB1 10.0 ± 0.7 ns 22.6 ± 0.5 b 43.4 ± 1.6 a 24.1 ± 1.9 a 0.56 ± 0.07 5.2 ± 1.9, b AZPDB1 9.9 ± 0.4 ns 23.5 ± 2.5 b 37.8 ± 2.3 a 28.9± 1.7 a 0.78 ± 0.07 10.2 ± 3.1, b Method 2 AZQU21/2/3 12.0 ± 1.0 ns 15.7 ± 0.1 b 44.6 ± 2.0 a 27.8 ± 1.6,ns 0.63 ± 0.06 0.7 ± 0.4, b AZQB21/2/3 11.5 ± 0.7 ns 25.8 ± 1.8 a 32.5 ± 1.5 b 30.2 ± 2.8,ns 0.94 ± 0.11 5.5 ± 1.1, a AZQDB21/2/3 10.1 ± 0.6 ns 26.3 ± 0.9 a 34.4 ± 1.7 b 29.3 ± 2.1,ns 0.86 ± 0.10 6.2 ± 0.3, a Letters within the same group are significantly different at p < 0.05 by HSD of Tukey test. ns: no significant differences. 50 The impact of soil disturbance on N and C in Mediterranean soils Knicker H, González-Vila FJ, Polvillo O, González JA, Almendros G (2005a): Fire-induced transformation of Cand Nforms in different organic soil fractions from a Dystric Cambisol under a Mediterranean pine forest (Pinus pinaster). Soil Biology and Biochemistry 37, 701-718 Knicker H, Totsche K-U, Almendros G, González-Vila FJ (2005b): Condensation degree of burnt peat and plant residues and the reliability of solid-state VACP MAS 13C NMR spectra obtained from pyrogenic humic material. Organic Geochemestry 36, 1359-1377 Knicker H, Almendros G, González-Vila FJ, González-Pérez JA, Polvillo O (2006): Characteristic alterations of quantity and quality of soil organic matter caused by forest fires in continental Mediterranean ecosystems: a solid-state 13C NMR study. European Journal of Soil Science 57, 558-569 Knicker H (2007): How does fire affect the nature and stability of soil organic nitrogen and carbon? A review. Biogeochemistry 85, 91-118 Knicker H, Müller P, Hilscher A (2007): How useful is chemical oxidation with dichromate for the determination of "Black Carbon" in fire-affected soils?. Geoderma 142, 178-196 Knicker H, Hilscher A, González-Vila FJ, Almendros G (2008a): A new conceptual model for the structural properties of char produced during vegetation fires. Organic Geochemistry 39, 935-939 Knicker H, Wiesmeier M, Dick DP (2008b): A simplified method for the quantification of pyrogenic organic matter in grassland soils via chemical oxidation. Geoderma 147, 69-74 Knicker H (2010): “Black nitrogen” – an important fraction in determining the recalcitrance of charcoal. Organic Geochemistry 41, 947-950 Knicker H (2011a): Solid state CPMAS 13C and 15N NMR spectroscopy in organic geochemistry and how spin dynamics can either aggravate or improve spectra interpretation. Organic Geochemestry 42, 867-890 Knicker H (2011b): Pyrogenic organic matter in soil: Its origin and occurrence, its chemistry and survival in soil environments. Quaternary International 243, 251-263 Knicker H (2011c): Soil organic N - An under-rated player for C sequestration in soils?. Soil Biology and Biochemistry 43, 1118-1129 Knicker H, González-Vila FJ, González-Vázquez R (2013): Biodegradability of organic matter in fire-affected mineral soils of Southern Spain. Soil Biology and Biochemistry 56, 31-39 Li Y (2010a): Can the spatial prediction of soil organic matter contents at various sampling scales be improved by using regression kriging with auxiliary information?. Geoderma 159, 63-75 Li Y (2010b): Can the spatial prediction of soil organic matter contents at various sampling scales be improved by using regression kriging with auxiliary information?, 159. Elsevier, Amsterdam, PAYS-BAS, 13 pp 51 Chapter 3. Variability of organic matter in fire affected soils Lombao A, Barreiro A, Carballas T, Fontúrbel MT, Martín A, Vega JA, Fernández C, DíazRaviña M (2015): Changes in soil properties after a wildfire in Fragas do Eume Natural Park (Galicia, NW Spain). CATENA 135, 409-418 López-Martín M, Nowak KM, Milter A, Knicker H (2016): Incorporation of N from burnt and unburnt 15N grass residues into the peptidic fraction of fire affected and unaffected soils. Journal of Soils and Sediments, 1-11 López-Serrano FR, Rubio E, Dadi T, Moya D, Andrés-Abellán M, García-Morote FA, Miettinen H, Martínez-García E (2016): Influences of recovery from wildfire and thinning on soil respiration of a Mediterranean mixed forest. Science of The Total Environment 573, 1217-1231 Lloret F, Vilà M (1997): Clearing of vegetation in Mediterranean garrigue: response after a wildfire. Forest Ecology and Management 93, 227-234 Martín A, Díaz-Raviña M, Carballas T (2012): Shortand medium-term evolution of soil properties in Atlantic forest ecosystems affected by wildfires. Land Degradation & Development 23, 427-439 Mayor AG, Bautista S, Llovet J, Bellot J (2007): Post-fire hydrological and erosional responses of a Mediterranean landscpe: Seven years of catchment-scale dynamics. CATENA 71, 68-75 Mayor ÁG, Goirán SB, Vallejo VR, Bautista S (2016): Variation in soil enzyme activity as a function of vegetation amount, type, and spatial structure in fire-prone Mediterranean shrublands. Science of The Total Environment 573, 1209-1216 Nannipieri P, Eldor P (2009): The chemical and functional characterization of soil N and its biotic components. Soil Biology and Biochemistry 41, 2357-2369 Nocentini C, Certini G, Knicker H, Francioso O, Rumpel C (2010): Nature and reactivity of charcoal produced and added to soil during wildfire are particle-size dependent. Organic Geochemestry 41, 682-689 Novara A, Gristina L, Bodì MB, Cerdà A (2011): The impact of fire on redistribution of soil organic matter on a mediterranean hillslope under maquia vegetation type. Land Degradation & Development 22, 530-536 Núñez MA, Recio JM (2007): Kaolinitic paleosols in the south west of the Iberian Peninsula (Sierra Morena region, Spain). Paleoenvironmental implications. CATENA 70, 388-395 Prieto-Fernández Á, Carballas M, Carballas T (2004): Inorganic and organic N pools in soils burned or heated: immediate alterations and evolution after forest wildfires. Geoderma 121, 291-306 Rengel Z, Marschner P (2007): Nutrient cycling in terrestrial ecosystems. Springer Rivas Y, Huygens D, Knicker H, Godoy R, Matus F, Boeckx P (2012): Soil nitrogen dynamics three years after a severe Araucaria–Nothofagus forest fire. Austral Ecology 37, 153-163 52 The impact of soil disturbance on N and C in Mediterranean soils Rumpel C, Alexis M, Chabbi A, Chaplot V, Rasse DP, Valentin C, Mariotti A (2006): Black carbon contribution to soil organic matter composition in tropical sloping land under slash and burn agriculture. Geoderma 130, 35-46 Tessler N, Wittenberg L, Greenbaum N (2016): Vegetation cover and species richness after recurrent forest fires in the Eastern Mediterranean ecosystem of Mount Carmel, Israel. Science of The Total Environment 572, 1395-1402 Velasco-Molina M, Berns AE, Macías F, Knicker H (2016): Biochemically altered charcoal residues as an important source of soil organic matter in subsoils of fire-affected subtropical regions. Geoderma 262, 62-70 Vergnoux A, Di Rocco R, Domeizel M, Guiliano M, Doumenq P, Théraulaz F (2011): Effects of forest fires on water extractable organic matter and humic substances from Mediterranean soils: UV–vis and fluorescence spectroscopy approaches. Geoderma 160, 434-443 Wang Q, Zhong M, Wang S (2012): A meta-analysis on the response of microbial biomass, dissolved organic matter, respiration, and N mineralization in mineral soil to fire in forest ecosystems. Forest Ecology and Management 271, 91-97 Wilson MA (1981): Applications of nuclear magnetic resonance spectroscopy to the study of soil organic matter. Journal Soil Science 32, 167-186 Xu Y, Sun J, Lin Q, Ma J, Shi Y, Lou K (2012): Effects of a surface wildfire on soil nutrient and microbial functional diversity in a shrubbery. Acta Ecologica Sinica 32, 258-264 Acknowledgments We acknowledge the Ministerio de Economía y Competitividad de España (MINECO) and the European Regional Development Fund (FEDER) for financial support (Project CGL200910557) and providing a travel grant to H. Knicker to visit the Institute of Bioand Geosciences, Agrosphere (IBG-3), Forschungszentrum Jülich GmbH, Germany. The latter and here in particular Dr. Anne Berns is gratefully acknowledged for providing access to the NMR facilities. The first author was financially supported by MINECO/FEDER (BES-2010-42581). Further Samuel Manzano Moñino and Ignacio Villar Mavellan, from the “Centro de Plan Infoca y Centros de Defensa Forestal (CEDEFO) is greatly acknowledged for providing helpful information about the study area and their assistance during the field work. Juan de Dios Franco Navarro is thanked for his support during the statistical analysis. Garbriel Molina Candau, Serena Tabanera Rodrígrez and the Company “Explotaciones Agrícolas, el Prado, S. L.” are acknowledged for allowing sampling on their property. Página | 53 Página | 53 Capítulo 4. Distribución y transformación del carbono y nitrógeno negro en las fracciones físicas de un suelo quemado después de siete años Chapter 4. Distribution and transformation of Black Carbon and Black Nitrogen in physical soil fractions from fire-affected soils after seven years of recuperation Este trabajo ha sido enviado para su publicación a Organic geochemistry Distribution and transformation of Black Carbon and Black Nitrogen in physical soil fractions from fire-affected soils after seven years of recuperation M. López-Martín, F. J. González-Vila & H. Knicker Instituto de Recursos Naturales y Agrobiología de Sevilla-CSIC, Avda. Reina Mercedes, 10, 41012, Sevilla, Spain Abstract Incorporation of black nitrogen (BN), formed during charring of plant residues and litter into soil organic matter (SOM), shifts its soil organic nitrogen composition toward N-heterocyclic constituents. To investigate the medium term impact of BN on the quality of SOM, topsoils from Cambisols and Luvisols located in the Sierra de Aznalcóllar (Southern Spain) were sampled four weeks and seven years after a severe fire. The solid-state 13C and 15N nuclear magnetic resonance (NMR) spectra of the samples obtained four weeks after the fire prompted to a quick incorporation of charred residues into SOM. Correspondingly, pyrrole-type N dominated their soil organic N fraction. Seven years of recuperation led to a considerable decrease of aryl C, although its contribution to the total organic C was still higher at the burnt sites than in the unburnt reference soil. The fast loss of pyrogenic organic matter (PyOM) during soil recuperation calls the claimed longevity of charred residues in the soils into question. The BN showed even lower biochemical recalcitrance. Density and particle size fractionation of the organic matter (OM) of the retrieved soils indicated that the remaining char accumulated mainly in the particulate organic matter fractions (POM). Whereas the soils under Quercus showed comparable distribution of aromatic C in the free POM (fPOM) and occluded POM (oPOM) fractions, those under pine revealed a preferential aryl C accumulation in the fPOM fraction. However, to what extent, the lack of interaction between the mineral fraction and the BN and black carbon (BC) contributed to their low biochemical recalcitrance needs further research. Keywords Particulate organic matter, particle-size fractionation, solid-state 15N NMR, biochemical recalcitrance of pyrogenic organic matter. 56 The impact of soil disturbance on N and C in Mediterranean soils 4.1. Introduction During a wildfire event the vegetation and its litter suffer structural changes such as condensation and cyclization of organic structures (Almendros et al. 2003). The new mixture of heat-altered biomacromolecules belong to a continuum of pyrogenic organic matter (PyOM) (Masiello 2004), the structure and properties of which depends on the combustion conditions and its source (Knicker 2010). The PyOM is commonly seen as an important C sink due to its highly aromatic nature. This was supported by Vasilyeva et al. (2011) who observed that after 55 years in fallow Chernozems the quantity and quality of PyOM did not change. On the other hand, recent laboratory incubation studies reported mean resident times of PyOM in soils in the range of decades to centuries but not around millennia as previously claimed (Bird et al. 2015, Hilscher et al. 2009, Santos et al. 2012). Incubation experiments by Nocentini et al. (2010) showed that charcoal can be mineralized by microorganisms. Further indications for biochemical degradation of PyOM were also given by Hilscher and Knicker (2011) who performed laboratory incubation experiments with grass-derived charcoal and evidenced microbial attack of the aromatic network with a concomitant increase of carboxyl C. In fact, analyzing deep horizons of Ferralsols under Cerrado, revealed that those oxidized aromatic ring structures are transported into deeper soil horizons where they accumulate (Velasco-Molina et al. 2016). Abiven et al. (2011) reported that aged charcoal (10 years) released between 40-55 times more benzene polycarboxylic acids (BPCA) than fresh charcoal, possibly due to the higher oxidation state of aged sample. Moreover, the latter contributed with more BPCA where all six rings C were substituted by carboxylic C to the extract. Evidence for PyOM degradation was also obtained in natural ecosystems such as soils from a fire-prone region in the Sierra de Aznalcóllar in Southern Spain. Here, the turnover times of PyOM were only slightly higher than those obtained for humified SOM (Knicker et al. 2013). Analysis of the soils seven years after the last fire confirmed fast recuperation of most soil parameters to the pre-fire values. Comparably, the aromaticity was only slightly higher in the fire-affected soils than in the unburnt counterparts, indicating that only few pyrogenic organic C (PyOC) survived the post fire recovery time (López-Martín et al. 2016b). Only recently, the PyOM research community recognized that nitrogen containing heterocyclic aromatic structures represent an important fraction of charred organic residues (Knicker 2010, Knicker et al. 2008). It is formed by cyclization of biogenic organic N sources such as peptides and amino sugars into mainly pyrrole-type N. Whereas early studies assumed a high recalcitrance of those structures (Knicker & Skjemstad 2000), newer incubation studies with 15N-enriched source material evidenced their biochemical degradability and the subsequent 57 Chapter 4. BC and BN in physical soil fractions use of the mobilized N for plant growth (de la Rosa & Knicker 2011). Experiments using 15Nenriched PyOM as soil amendment in pot experiments further evidenced the incorporation of BN-derived nitrogen into the peptide pool of SOM (López-Martín et al. 2016a). Whereas our understanding of how and when BC is degraded in soils is still insufficient, even less is known about the fate and stability of BN during aging of PyOM. Aside from chemical protection of organic residues due to their low biochemical accessibility, i.e, due to lack of adequate enzymes or steric hindrance (Knicker et al. 1996), survival of OM in soils and sediments was suggested to occur via physical encapsulation into hydrophobic organic networks (Knicker & Hatcher 1997). Alternatively, SOM can be part of the molecular structure and the chemical mechanisms, the SOM pools are protected from the decomposition via physical mechanism (Skjemstad et al. 1996, Sollins et al. 1996) either by adsorption, entrapment into micropores (Mayer 1994) or occlusion within aggregates. According to the latter concept, density and physical fractionation are powerful methods to distinguish different OM pools that differ in their degradability. The particle size fractionation is based on the concept that SOM attached to particles of different size and mineralogy differs in structure and function (von Lützow et al. 2007). Using solid-state nuclear magnetic resonance spectroscopy, Baldock et al. (1992) observed an increasing degradation degree of SOM from coarse to fine particles. The free particulate organic matter (fPOM) is part of the fraction which is floating on the aqueous phase of a soil suspended in a solution with a defined density between 1.6-1.8 g cm-3 (Cerli et al. 2012, Sohi et al. 2001, Sollins et al. 2006). This material is only scarcely degraded and does not interact with the mineral phase (unprotected C pool). Occluded particulate-organic matter (oPOM) shows a higher degradation degree and is protected within macroaggregates. It is liberated by disruption of the aggregates by ultrasonification with defined energy between 450 and 500 J ml-1 (Kölbl & Kögel-Knabner 2004). Cerli et al. (2012) reported that the intensity of dispersion vary with the aggregate stability. Whereas the fPOM and oPOM are labile pools with turn-over times from days to years, turnover times from years to centuries were observed for material associated with the fine fractions (clay) (Wagai et al. 2015). However, one has to keep in mind that fresh microbial material can also accumulate in the fine fractions (Clemente et al. 2011, Plaza et al. 2013). Studies on Australian soils demonstrated that PyOM is preferentially associated with the fine fractions (Skjemstad et al. 1999). Singh et al. (2014), on the other hand, found the highest PyOM 58 The impact of soil disturbance on N and C in Mediterranean soils yields in the fPOM separate 10 months after PyOM addition to the soil. Comparably, during a short term incubation experiment Hilscher and Knicker (2011) using dual-isotopic (15N, 13C) labelled PyOM, recovered most of the label in the POM fraction, although in spite of the short time span of the experiment, some of the PyOM residues were already incorporated into the mineral-associated phase. Liang et al. (2008) evidenced physical protection by superficial interaction between PyOM and minerals, but suggested that chemical recalcitrance was more important. Heckman et al. (2014) found a preferential accumulation of PyOM into the oPOM fraction. The intention of the present study was to bring some light on the stability of BN in fireaffected soils. Therefore, we characterized the chemical composition of SOM in soils sampled 4 weeks and 7 years after the last fire event, both by solid-state 13C and 15N NMR spectroscopy. In order to elucidate the impact of SOM-mineral phase interaction on PyOM stabilization, we assessed how the PyOM is distributed among the density and particle size fractions and if the interaction with the mineral phase can explain the survival of PyOM, in particular its BN constituents, 7 years after of a fire 4.2. Materials and Methods 4.2.1. Site description The study site is located in the National Park of Sierra Aznalcóllar in Southern Spain, which was affected by wildfires in 1996 and 2004. The latter destroyed around 30000 ha of forest. A more detailed description of the sites and the fire history of this region was published elsewhere (López-Martín et al. 2016b). The sampled soils are classified as Leptosols and Cambisols (IUSS Working Group WRB 2014). Before the last fire, the covering vegetation was pine (Pinus pinaster, and Pinus pinea) and Quercus (suber and ilex) forests, but the 2004 fire largely destroyed the vegetation. The first sample set, which was used to describe the short-term impact of fire on SOM decomposition of the bulk soils derived from a sampling campaign, conducted 4 weeks after the fire event in August 2004. The respective samples are marked with the code 04. The material was taken from the first 5 cm of a fire-unaffected (sample code: U) mineral soil under pine (sample code: P) (sample name: AZ04PU) and of the fire-affected (sample code: B) mineral soils under oak (sample code: Q) (AZ04QB), respectively, after removing the litter or ash layer (Knicker et al. 2013). A further composite sample derived from a double burnt (sample code: DB) Cambisol 59 Chapter 4. BC and BN in physical soil fractions close to a pine which survived the fires in 1996 and in 2004 (AZ04PDB). In June 2011, those locations were visited again and material from the first 5 cm of the mineral soils were collected (López-Martín et al. 2016b). At the site under oak, the trees which survived the fire in 2004 recovered together with a dense shrub cover. Here, two samples were selected, one of which derived from a place under a shrub (AZQBd) and the second from place scarcely covered (AZQBa). At the double burnt site, from which samples of three soils were used for the present analysis, (AZPDBc,d,e) most of the burnt trees were removed and a maquia (Cistus ladanifer, Retama sphaerocarpa) had developed. A further site formerly under pine was sampled (AZPB) remaining tree trunks and their roots were removed after the fire and the area was terraced for restoration. The four corners and the center of a randomly chosen square with a side length of 15 m were sampled. However, based on pre-experiments (López-Martín et al. 2016b), three samples (AZPBc,d,e) revealing detectable difference with respect to OM quality and quantity were chosen for further analysis in the present study. Note at this site, considerable amounts of charcoal particles were visually detected already during sampling. Most likely, they originate from an unregistered fire, which occurred between 2004 and 2011, or from charred wooden hunks that were disintegrated and incorporated into the soil during the restoration work. In addition, composite samples from a third sampling event in December 2011 were used for the present study. Here, composite samples of sites under Quercus, which were burnt only once in 2004 (AZQB1) and twice in 1996 and 2004 (AZQDB1), were probed. An unburnt soil under Quercus (AZQU3) was used as a reference. The soil textural classifications were done according the (FAO 2006) (Table 4.1). The soils of the first sampling in 2004 were air-dried whereas those taken in 2011 were dried at 40º C. Al samples were sieved through a mesh of ө < 2 mm to obtain the fine earth and subsequently stored in the dark until further analysis. 4.2.2. Soil fractionation 4.2.2.1. Density fractionation In order to reveal the medium term fate of BC and BN, the topsoils taken in 2011 were subjected to a density fractionation procedure.in order to separate fPOM, oPOM and the mineral associated organic fractions (MAF) (Golchin et al. 1994b, Sohi et al. 2001). Briefly, 30 g of fine 66 The impact of soil disturbance on N and C in Mediterranean soils 4.3.3 Distribution of Corg, Nt among the fractions and its C/N ratio For the soils derived from the sites formerly under pine, most of to the initial total organic carbon (Ci) and total nitrogen (Ni) is associated to the fPOM fraction (Table 4.3). Whereas the relative contribution of Ci and Ni between those sites is fairly comparable, lower amounts of Ci and Ni are recovered in the oPOM fraction of the double burnt site than in that of the single and recently burnt soil. The contrary is true for MAF. Note that the C/N ratios of the density fractions are clearly higher in those derived from the single burnt site than in those extracted from the double burnt soils. This may indicate that the SOM at the double burnt site was at a more developed degradation state. Alternatively, the single burnt site may have received a higher input of low-N sources. Separating the MAF into sand (2000-63 µm), coarse silt (63-20 µm) and the fine fraction (< 20 µm) showed that the latter represents the largest part with respect to contribution to the bulk soil and to total OM (Table 4.4), which is in line with (de Junet et al. 2013). The low C/N ratios (Table 4.4) of the mineral fractions are in the range expected for microbial biomass or well-humified SOM. Compared to the soils under pine, in the soils under Quercus less C and N was recovered with fPOM, however a clear trend with respect to fire history was not revealed. Whereas the different fractions obtained from AZQBa,d show comparable C/N ratios to those of the unburnt composite soil AZQU3, those of AZQB1 and AZQDB1 were clearly higher. In line with the observation for the pine stand, after particle size fractionation, between 75 and 92% of the total C and N was recovered in the fine fraction. Thus, this fraction mainly determines the composition of MAF. The course silt fractions comprise less than 4% and 8% of the Nt and Corg, respectively of the bulk samples. Comparably, the sand fraction contains less than 17% of the Corg and between 8 and 16% of the Nt with respect to the bulk soil. A trend with fire history or stand was not observed. 4.3.4. Chemical characterization of SOM fractions The solid state 13C NMR spectra of the fPOM of almost all soils sampled from stands under Quercus are dominated by signals from O-alkyl C, alkyl C followed by aryl C as found by Plaza et al. (2012) and Golchin et al. (1994b) for soils with different properties and climate conditions (Fig. 4.2 and 4.3). Compared to the bulk soils, the fPOM fractions contain slightly more and the oPOM separate slightly less O-alkyl C (Fig. 4.2 and 4.3). This finding is in agreement with the fact that fPOM is composed mainly of plants debris, which are only partly degraded that fPOM Table 4.3. Recovery of organic carbon (Corg) and total N (Nt) as carbon (Cf) and nitrogen (Nf) of the free (fPOM), occluded (oPOM) particle organic matter fractions and the mineral-associated organic fraction (MAF) from unburnt (AZQU3), single (AZQB1, AZQBa,b) and double burnt (AZQDB1) Cambisols under Quercus and from double (AZPDBc,d,e) and single (AZPBc,d,e) burnt Cambisols under pine sampled seven years after a fire in the Sierra de Aznalcóllar, Southern Spain. Sample fPOM oPOM MAF Cf of Corg (%) Nf of Nt (%) C/N Cf of Corg (%) Nf of Nt (%) C/N Cf of Corg (%) Nf of Nt (%) C/N AZQU3 46.0 ± 0.3 28.7 ± 4.1 24 ± 3 33.8 ± 1.0 25.3 ± 2.5 20 ± 1 21.2 ± 1.0 46.7 ± 5.9 7 ± 1 AZQB1 31.7 ± 1.6 17.9 ± 3.2 36 ± 6 48.3 ± 7.2 28.1 ± 5.5 35 ± 3 20.0 ± 0.2 54.0 ± 6.6 7 ± 1 AZQDB1 32.7 ± 1.4 16.1 ± 0.0 39 ± 5 43.9 ±0.6 26.3 ± 1.3 34 ± 4 24.1 ± 1.5 57.6 ± 6.8 11 ± 2 AZQBa 41.8 ± 4.0 17.1 ± 2.9 26 ± 3 26.9 ± 1.5 11.4 ± 1.7 26 ± 2 31.3 ± 0.0 71.6 ± 12.1 5 ± 1 AZQBd 58.5 ± 6.4 30.4 ± 5.0 26 ± 1 11.4 ± 0.4 6.8 ± 0.9 24 ± 2 30.1 ± 3.2 62.7 ± 6.5 7 ± 1 AZPDBc 67.7 ± 1.7 40.5 ± 5.7 29 ± 4 12.9 ± 0.3 8.5 ± 0.9 27 ± 3 19.4 ± 0.3 51.1 ± 3.6 7 ± 1 AZPDBd 63.1 ± 1.1 36.5 ± 7.4 31 ± 5 15.0 ± 0.4 9.9 ± 1.1 33 ± 3 21.9 ± 1.6 55.6 ±7.5 7 ± 0 AZPDBe 71.4 ± 2.1 47.3 ± 7.2 34 ± 5 16.7 ± 0.2 12.1 ± 2.3 32 ± 5 11.9 ± 2.0 40.7 ± 3.8 7 ± 1 AZPBc 65.6 ± 1.7 50.2 ± 7.0 40 ± 5 26.5 ± 2.4 21.7 ± 5.3 38 ± 5 7.9 ±0.2 28.1 ± 3.2 9 ± 1 AZPBd 62.0 ± 1.6 39.8 ± 5.1 47 ± 5 26.9 ± 0.5 17.8 ± 2.1 45 ± 4 11.1 ± 0.6 42.4 ± 3.9 8 ± 3 AZPBe 60.5 ± 2.3 26.6 ± 4.9 47 ± 6 25.3 ± 0.8 9.5 ± 1.1 55 ± 5 14.2 ±1.0 63.9 ± 12.1 5 ± 1 68 The impact of soil disturbance on N and C in Mediterranean soils is composed mainly of plants debris, which are only partly degraded (Golchin et al. 1994a) whereas the material in oPOM underwent advanced microbial degradation (Christensen 2001, Dorodnikov et al. 2011). The alkyl C-to-O-alkyl C ratio (Table 4.5) suggested by (Baldock et al. 1997, Golchin et al. 1994a) as a degradation index is in line with this interpretation. Calculating the alkyl C-to-carboxyl C ratio points toward an increase of the average chain length from fPOM to oPOM, which is line with Plaza et al. (2012) who revealed an enrichment of unsubstitutedaliphatic material such as lipid structures (cutin and suberin) in oPOM. Note that those patterns are observed for all studied samples sets irrespective of the site and fire history. Comparably, the solid-state 15N spectra of all studied POM fractions show no major difference (Fig.4.4). The only signal which unbiasedly can be discerned from the noise is attributed to amide N, indicating that Black Nitrogen did indeed not survive to a higher extent in these fractions. The soil organic carbon in MAF at particle sizes < 20 µm is composed mainly of O-alkyl C. A comparable finding is reported by Sohi et al. (2001) and Boeni et al. (2014). Comparing the relative intensity distribution in the 13C NMR spectra of the clay fractions with those of the oPOM, a higher contribution of O-alkyl C is observed for all sampling sites. A relationship with stand or fire history is not evident. The Alkyl C-to-O-alkyl C ratio increased in the following order fPOM, < 20 µm and oPOM (Table 4.5). This is in contrast to the general assumption that the most degraded SOM is associated with the fine particle size fractions. On the other hand it is in line with Bimüller et al. (2014) who stated that in spite of the enrichment in polysaccharides, the SOM in the fine fractions are expected to be less bioavailable or/and accessible for microbial degradation due to the interaction or the OM with the mineral surfaces. In the fine particle size fractions, the carboxyl C contributes with < 10% to the Corg, which is in accordance with the assumption that their OM has not experienced extensive microbial degradation but has contribution of microbial residues (Clemente et al. 2012, Clemente et al. 2011, Six et al. 2002). Table 4.4. Recovery of organic carbon (Cf) and nitrogen (Nf) from the organic carbon (Corg) and total nitrogen (Nt) of the bulk soils in the sand-size (2000-63 mm), course silt-size (63-20 mm) and fine-size (< 20 µm) fractions from unburnt (AZQU3), single (AZQB1, AZQBa,b) and double burnt (AZQDB1) soils under Quercus and from Cambisols under pine after two (AZPDBc,d,e) and one (AZPBc,d,e) forest fires and the C/N ratios of the different fractions. The samples derive from the Sierra de Aznalcóllar, Southern Spain, and were taken seven years after the last fire. Sample 2000-63 mm 63-20 mm < 20 µm Cf of Corg (%) Nf of Nt (%) C/N Cf of Corg (%) Nf of Nt (%) C/N Cf of Corg (%) Nf of Nt (%) C/N AZQU3 6.3 ± 0.8 8.4 ± 0.0 5 ± 0 2.9 ± 0.1 3.0 ± 0.4 7 ± 1 89.2 ± 6.6 91.6 ± 2.6 7 ± 0.3 AZQB1 13.2 ± 1.7 11.7 ± 2.9 10 ± 3 8.4 ± 1.2 4.4 ± 0.5 17 ± 0 78.6 ± 3.8 83.9 ± 2.0 8 ± 0.3 AZQDB1 16.8 ± 2.8 14.6 ± 7.5 13 ± 1 7.9 ± 1.4 4.0 ± 1.7 27 ± 8 74.9 ± 1.9 86.6 ± 0.0 11 ± 0.3 AZQBa 13.8 ± 1.5 15.5 ± 3.5 5 ± 1 5.8 ±0.2 3.9 ± 0.0 8 ± 0 80.1 ± 7.1 79.1 ± 14.7 6 ± 0.9 AZQBd 11.2 ± 1.3 16.0 ± 4.6 6 ± 2 2.5 ± 0.1 2.9 ± 0.0 7 ± 0 86.6 ± 1.2 80.9 ± 3.7 9 ± 0.3 AZPDBc 9.4 ± 0.2 10.6 ± 0.0 6 ± 0 4.0 ± 0.5 1.8 ± 0.2 15 ± 1 86.1 ± 5.3 87.8 ± 5.6 7 ± 0.0 AZPDBd 10.4 ± 2.5 14.6 ± 4.3 6 ± 2 4.0 ± 0.3 1.8 ± 0.7 11 ± 3 83.2 ± 3.9 82.6 ± 9.0 8 ± 0.6 AZPDBe 26.4 ± 3.0 30.4 ± 8.5 7 ± 1 8.0 ± 0.1 5.2 ± 0.0 12 ± 0 66.1 ± 0.6 64.5 ± 2.1 8 ± 0.2 AZPBc 5.3 ± 0.6 4.8 ± 0.6 8 ± 1 3.4 ± 0.3 3.2 ± 0.3 8 ± 0 91.7 ± 5.5 92.0 ± 5.9 8 ± 0.0 AZPBd 6.9 ± 0.3 8.2 ± 1.0 7 ± 1 3.2 ± 0.3 1.5 ± 2.2 8 ± 0 89.5 ± 7.7 90.3 ± 6.4 8 ± 0.1 AZPBe 12.7 ± 3.6 8.2 ± 1.1 7 ± 3 3.1 ± 0.1 3.4 ± 0.4 4 ± 0 84.5 ± 0.0 88.4 ± 0.0 4 ± 0.0 70 The impact of soil disturbance on N and C in Mediterranean soils Fig. 4.2. Solid-state 13C NMR spectra of the HF-treated fine particle-size fractions (< 20 µm) and the free and occluded particulate organic matter (fPOM, oPOM) after density fractionation (1.8 g cm-3) from Cambisols sampled seven years after a fire event in the Sierra de Aznalcóllar, Southern Spain. The latter were located under Quercus forests which were unburnt (AZQU3), burnt once (AZQB1, AZQa,d) and double burnt (AZQDB1). Further material was obtained from soils which were under pine before they were burnt once (AZPBc,d,e) or twice (AZPDBc,d,e). 71 Chapter 4. BC and BN in physical soil fractions The fact that the intensity in the methoxyl C/N-alkyl C (60 to 45 ppm) increased slightly, whereas the clear signals assignable to O-aryl of lignin between 160 and 140 ppm almost vanished underlines an accumulation of peptides in the mineral fraction. This is supported by their narrow C/N ratio and the dominance of the signal in the amide N region between -250 to - 285 ppm of their solid-state 15N NMR spectrum (Fig. 4.4). Accordingly, Zhou et al. (2014), suggested that the clay fraction adsorbs and protects N-rich OM such as peptides. Table 4.5. Alkyl C/O-alkyl C and alkyl C/carboxyl C ratios of free (fPOM), occluded (oPOM) particle organic matter and fine fraction (< 20 µm) of the mineral associated organic matter derived from unburnt (AZQU3), single (AZQB1, AZQBa,b) and double burnt (AZQDB1) soils under Quercus and from double (AZPDBc,d,e) and single (AZPBc,d,e) burnt Cambisols under pine. All soils were sampled seven years after the last fire from the Sierra de Aznalcóllar, Southern Spain. Samples Alkyl C/O-alkyl C Alkyl C/carboxyl C fPOM oPOM < 20 µm fPOM oPOM < 20 µm AZQU3 0.8 1.0 0.7 4.2 3.7 4.0 AZQB1 0.6 1.3 0.8 3.7 4.7 4.4 AZQDB1 0.7 1.3 0.8 3.7 4.8 4.3 AZQBa 0.8 1.5 1.0 3.2 3.4 3.1 AZQBd 0.8 1.0 0.7 2.9 2.9 3.4 AZPDBc 0.7 1.1 0.8 3.6 3.4 2.5 AZPDBd 0.7 1.5 0.8 3.5 3.9 2.9 AZPDBe 1.3 1.4 1.4 2.8 1.8 2.6 AZPBc 0.6 1.1 0.8 2.0 2.4 2.6 AZPBd 0.7 1.3 0.9 2.7 3.6 3.4 AZPBe 0.7 1.1 0.9 3.3 3.8 2.7 With respect to the contribution of aryl C, a clear impact of the fire history is observed for the samples obtained from the stands under Quercus and pine (Fig. 4.3). Although for the bulk soils, the aromaticity decreased considerably during the seven years of recuperation (Fig. 4.1), the fPOM and oPOM fractions of the single and double burnt site seven years after the fire showed clearly higher aryl C contributions than the unburnt site AZQU3. With the exception of the samples from AZPBc,d,e, the aromaticity of oPOM from the soils of the burnt sites collected in 2011 was higher than that of fPOM, but no difference between double and single burning was revealed (Fig. 4.3). This is in contrast to studies of Singh et al. (2014) who found the highest Fig. 4.3. Intensity distribution of the solid-state 13C NMR spectra of the HF-treated bulk soils from Cambisols, their free and occluded particulate organic matter (fPOM, oPOM) obtained after density fractionation (1.8 g cm-3) and their fine particle-size fraction (< 20 µm). The soils were sampled in the Sierra de Aznalcóllar seven years after a severe fire event and were located under Quercus stands which were unburnt (AZQU3), single burnt (AZQB1, AZQa,d) and double burnt (AZQDB1) and from soils which were under pine before they were burnt twice (AZPDBc,d,e) or once (AZPBc,d,e). 0 10 20 30 40 50 13 C intensity (%) AZQU3 0 10 20 30 40 50 AZQB1 0 10 20 30 40 50 AZQDB1 0 10 20 30 40 50 13 C i n t e n s it y ( % ) AZQBa 0 10 20 30 40 50 AZQBd 0 10 20 30 40 50 AZPDBd 0 10 20 30 40 50 13C intensity (%) AZPDBc 0 10 20 30 40 50 AZPDBe Bulk fPOM oPOM 20 µm Bulk fPOM oPOM 20 µm Bulk fPOM oPOM 20 µm Bulk fPOM oPOM 20 µm Bulk fPOM oPOM 20 µm 0 10 20 30 40 50 13C intensity (%) AZPBc Bulk fPOM oPOM 20 µm Bulk fPOM oPOM 20 µm Bulk fPOM oPOM 20 µm Bulk fPOM oPOM 20 µm Bulk fPOM oPOM 20 µm 0 10 20 30 40 50 AZPBd Bulk fPOM oPOM 20 µm Bulk fPOM oPOM 20 µm Bulk fPOM oPOM 20 µm Bulk fPOM oPOM 20 µm Bulk fPOM oPOM 20 µm 0 10 20 30 40 50 AZPBe Carboxyl C Aryl C O-alkyl C N-alkyl C Alkyl C 73 Chapter 4: BC and BN in physical soil fractions PyOM contribution in the fPOM fraction. Between 62 and 86% of the aromatic C of the bulk samples was recovered with both POM fractions (Table 4.6). A preference for oPOM or fPOM was not evidenced for the samples derived from the sites under Quercus, whereas those from the stands under pine indicated a preferential accumulation in the fPOM fraction. A closer analysis of the aromatic region of the solid-state 13C NMR spectrum of the < 20 µm fraction of the unburnt soils under Quercus (Fig. 4.2) discloses an almost compete disappearance of the signal in the chemical shift region between 160 and 140 ppm assignable to O-aryl C as it occurs in lignin. This is in accordance to other reports showing that lignin residues rarely enters this fractions (Dorodnikov et al. 2011, Heckman et al. 2014) and the aromatic contribution must originate from other sources. Alternative compounds are aromatic amino acids or olefinic structures in unsaturated lipids or charcoal contributions. Conclusively, in the case of the burnt soils, the higher aromatic C contribution is best explained with fine PyOM residues. However, their contribution is lower than 4 % of Ci in the bulk soils and less than 20% of its aromatic C moiety (Table 4.6). Less than 30% of the aromatic C has been calculated to be associated to the course mineral-associated fractions. Fig. 4.4. Solid-state 15N NMR spectra of free and occluded particulate organic matter fractions (fPOM, oPOM) and the HF-treated fine particle-size fraction (< 20 µm) of Cambisols from the Sierra de Aznalcóllar. They were sampled in 2011, 7 years after a wildfire of a single (AZQB1, AZQBd) and double (AZQDB1) burnt Quercus forest and a pine forest (AZPBc). 74 The impact of soil disturbance on N and C in Mediterranean soils For the soils obtained from the pine stands, even lower amounts of Ci are recovered as aromatic C of the < 20 µm fractions. Note that here too, clear signal for lignin are not evidenced in the respective solid-state 13C NMR spectra and the solid-state 15N NMR spectra show most of their intensity in the amide N region (Fig. 4.4). Between 59 and 69% of the aromatic C of the bulk soils was recovered with the fPOM fraction, leaving 15 to 18% associated with oPOM. In the samples with a recent charcoal input these numbers change to 47 to 57% for fPOM and 19 to 24% for oPOM. Less than 30% of the aromatic C can be attributed to the course mineralassociated fractions. However, note that at both pine sites, 60 to 70% of Ci is recovered with fPOM. Our results are in accordance with Soucémarianadin et al. (2014) evidencing that in fire affected soil the charcoal can be found in the fPOM. Heckman et al. (2014), on the other hand, identified the highest charcoal contributions in the oPOM fraction of soils derived from a temperate forest ecosystem under a conifer stand. They concluded that here fire was the main Table 4.6. Recovery of organic C of the bulk soil as aromatic carbon (Carom) in the free (fPOM), occluded (oPOM) particle organic matter, fine fraction (< 20 µm) and course fraction (2000 – 20 µm) of unburnt (AZQU3), single (AZQB1, AZQBa,b) and double burnt (AZQDB1) soils under Quercus and from double (AZPDBc,d,e) and single (AZPBc,d,e) burnt Cambisol under pine sampled seven years after a fire in the Sierra de Aznalcóllar. Carombulk corresponds to the contribution of aromatic C in each fraction to the total aromatic C content in the bulk soil. Samples fPOM oPOM < 20 µm 2000-20 µm Carom (%) Carombul k (%) Carom (%) Carombul k (%) Carom (%) Carombul k (%) Carombulk (%) AZQU3 6.8 43.0 5.4 34.6 1.8 11.5 11 AZQB1 7.0 24.5 10.6 37.3 2.4 8.4 30 AZQDB1 6.7 26.1 9.3 36.4 2.2 8.4 29 AZQBa 8.5 39.3 6.2 28.9 4.3 20.1 12 AZQBd 12.5 54.4 2.8 12.4 4.1 17.9 15 AZPBc 23.8 47.8 9.6 19.2 1.5 3.0 30 AZPBd 23.7 57.1 9.8 23.6 2.4 5.8 14 AZPBe 21.4 47.0 10.9 24.0 2.8 6.2 23 AZPDBc 17.4 68.8 3.8 14.8 3.6 14.2 2 AZPDBd 12.8 59.2 3.4 15.8 2.2 10.1 15 AZPDBe 22.1 68.2 5.9 18.1 2.4 7.4 6 75 Chapter 4: BC and BN in physical soil fractions driver of the formation of the occluded fraction and that the selective preservation of PyOM is the main C stabilization mechanism in the oPOM fraction. Other studies report a preferential accumulation of BC in the light fraction (ρ < 2 mg m-3) of agriculturally used topsoils (Brodowski et al. 2007, Vasilyeva et al. 2011). With increasing soil depth, the BC content in the light fraction decreased whereas that of the more dense fraction remained constant, possibly because the latter experienced a better protection due to interaction with the mineral phase. Protection by means of mineral-BC interaction was also suggested by Vasilyeva et al., 2011 and Rovira et al. (2009). Liang et al. (2008) examining the stability of BC in the Anthrosol proposed further that chemical recalcitrance may have been more important than physical protection at their study site. Stabilization of PyOM via mineral-BC interactions in the fine size particles cannot be confirmed by our study (Table 4.6) since < 23% of the aromatic C of the bulk soils is recovered with the fraction < 20 µm (clay, fine and medium silt). It seems that the majority of char particles introduced into the soil after the fire were too big to enter the medium and fine silt fraction. Considering that for PyOM in these soils, relative short mean residence times in the range of centuries were determined (Knicker et al. 2013) ongoing recuperation of the ecosystem may lead to further degradation and thus disintegration of PyOM associated with POM. On the one hand, this would lead to a relative enrichment of mineral-associated PyOM, but also to the formation of smaller units, which can enter the fine particle size fractions and contain sufficient functional groups for providing adsorption sites for interaction with the mineral phase. 4.4. Conclusions The present investigation confirmed low longevity of PyOM in fire-affected soils from the Sierra de Aznalcóllar. However, after seven years of recuperation from the last fire, the presence of PyOM was still revealed by higher aromatic C contents in the SOM of burnt than of unburnt soils. The observation that most aromatic structures were recovered in the POM may evidence high chemical recalcitrance of the remaining charred residues since POM is only weakly protected by mineral interactions and assumed to be quickly degraded. On the other hand, the fast loss of aromaticity during the seven years of recuperation argues against this interpretation of high chemical recalcitrance. Possibly, the lack of mineral-PyOM association may have allowed fast PyOM degradation. However, the functional groups of some partially oxidized PyOM may interact with the mineral surface before they were completely metabolized. Physically protected by this interaction the PyOM may be able to survive on a long-term scale. Following this argumentation, the BC detected in the fine fraction of Chernozems (Vasilyeva et Este trabajo se está preparando para ser enviado para su publicación a Soil Biology and Biochemistry Can Black Nitrogen serve as a N-source for plant growth in fire-affected areas and how is its use affected by the presence of inorganic N? M. López-Martín, M.Velasc-Molina & H. Knicker Instituto de Recursos Naturales y Agrobiología de Sevilla - CSIC. Av. Reina Mercedes nº 10. 41012 Sevilla, Spain Abstract In order to obtain a more profound understanding of the impact of pyrogenic organic matter (PyOM) on the N bioavailability in fire-affected soils, an incubation experiment was performed. Therefore, Lolium perenne was grown on soil material from unburnt and burnt forest areas in the Sierra de Aznalcóllar, Southern Spain after covering it with 15N-enriched fresh grass litter (15NOM) or its PyOM (15N-OM). In parallel, pots were additionally fertilized with non-enriched KNO3 (Ni). Further pots contained unenriched PyOM or unburnt plant (OM) residues together with K15NO3 (15Ni) and only 15Ni. The amount of added N was adjusted to avoid alteration of the C/N ratio of the soil as well as N-excess or limitation. After defined incubation times, the remaining litter layer was removed and soil, roots and leaves were separated. Determination of the recovered added 15N (15Nadd) together with a statistical analysis of the results showed that fire history of the soil had no impact on the 15Nadd partitioning between plant leaves, roots and soils. Whereas 15Nadd-uptake into the roots showed no statistical difference between kind of amendment or with incubation time, in the plant leaves clearly higher 15Nadd recoveries were obtained for pots treated with 15Ni and 15N-OM than for those with 15N-PyOM. Neither positive nor negative priming was detected, since addition of Ni did not alter the percentage of 15Nadd derived from 15N-PyOM and 15N-OM found in the leaves. Comparably, addition of PyOM and OM did not alter the efficiency of 15Ni incorporation into plant residues. After 16 months of incubation, the soils with 15N-PyOM amendment showed the highest 15Nadd-sequestration. Addition of 15Ni led to a considerable 15N-loss until the end of the experiment. Solid-state 15N nuclear magnetic resonance spectroscopy (NMR) suggests that some of the 15Nadd sequestered in the soils topped with 15N-PyOM occurred as amide N, most likely of microbial biomass residues. In summary, our results allow the conclusion that addition of N-rich PyOM and ash to soils after fires has a strong short-term impact on N-cycling in the affected soils. Accordingly, directly after the fire, Ni of the ash replaces the N which otherwise had been supplied by the decaying 84 The impact of soil disturbance on N and C in Mediterranean soils plant litter. It provides a fast N-fertilization that is necessary for a quick recovery of the plant cover. Since Ni which is not directly used for biomass production can be quickly lost by volatilization or leaching, PyOM serves as an efficient N sink which only slowly releases N and thus provides additional fertilization on a medium and long-term scale. On a long-term scale, almost all of the N in PyOM will be transformed into soil organic N with comparable properties and functions as SON derived from unburnt litter. 5.1. Introduction The biggest nitrogen (N) reservoir is the atmosphere where the most abundant N form dinitrogen (N2) constitutes about 78 % of the air gas. As such, it is not directly available for most microorganisms or plants although in soils, nitrogen represents a limiting nutrient and is need for the built-up of their biomass.During the N cycle, N2 is converted by microbial fixation or lightning into reactive forms (ammonia, NH3) which can be used for biomass production. At the same time, bioavailable N forms are mobilized during the decomposition of organic matter. In soils, plants and microbes can assimilate the inorganic N forms such as nitrate (NO3-) and ammonium (NH4+), but also low molecular weight organic N, such as amino acids. In particular in nutrient-poor ecosystems, plants and microbes have established a competitive relationship, which is expected to be altered by ecosystem disturbance, as there are vegetation fires. Burning of plant debris or the humic layer is expected to alter the kind of litter refilling the SOM pool. Hence, PyOM produced during the incomplete combustion of vegetation and SOM is made up by aromatic structures and heterocycle-N rings formed by dehydration, decarboxylation, demethylation, condensation and cyclization (Almendros et al. 2003, Knicker 2007). The newly formed N-compounds known as Black Nitrogen (BN) (Knicker 2010) derive from N source of biogenic origin such as proteins and amino acids which are transformed into pyrroles, imidazoles and indoles (Almendros et al. 2003) as was demonstrated by solid-sate 13C and 15N nuclear magnetic resonance (NMR) spectroscopy. These structures were formerly considered to be hardly degradable by microorganisms and their formation from labile N-sources is expected to increase the N competition between plants, microbes and SOM. Although others detected mean residence times of BC of 300 years (Lehndorff et al. 2014). However, previous short term experiments using 15N-enriched PyOM as a soil amendment showed that 15N derived from BN (Hilscher & Knicker 2011) can be used for plant growth (de la Rosa & Knicker 2011). Subsequent analysis of the soil with solid-state 15N NMR spectroscopy revealed that most of the remaining organic 15N derived from the added BN was transformed into amide structures, which 85 Chapter 5. BN as a N-source in presence of inorganic N may accumulate either as primary degradation products of BN or as biogenic residues after the incorporation of 15N mobilized during BN degradation into microbial biomass. However, it is commonly assumed that charring of N-rich biomass releases a considerable amount of ammonium, which is accumulating in the burnt soil (Mayor et al. 2016, Rodríguez et al. 2009). Since the fire has killed most of the vegetation, the competition between nitrifiers and plants for ammonium has shifted in favor to the nitrifying microbes leading to an accumulation of nitrate (Karhu et al. 2015). However, if not used by the plants, this nitrogen may be lost for the recovering ecosystem due to leaching during post-fire rains. In order to obtain a more profound understanding of the bioavailability of N from PyOM by plants and to evaluate how the input of BN affects the N cycling in the plant-soil system an incubation experiment was performed. Therefore, Lolium perenne was grown on soil material from a fire-prone forest area in the Sierra de Aznalcóllar, Southern Spain after covering it with 15N-enriched PyOM from grass residues, to simulate the accumulation of charred material. In parallel, pots were prepared with 15N-enriched but unburnt grass residues and with inorganic 15N. In addition, by amending pots with mixtures of 15N-enriched and non-enriched N sources (fresh plant material; OM, PyOM and KNO3), we intended to obtain some insights into the importance of N from PyOM as a potential N source in fire-affected soil. After defined incubation times, the remaining litter layer was removed and soil, roots and leaves were separated. Analyzing their 15N content allowed the monitoring of the incorporation of BN-derived 15N between the different compartments. Solid-state 15N NMR spectroscopy was used to identify the organic N forms sequestered in the soil. In order to reveal the impact of previous fires on the N-cycling, we used soil material from both an unburnt and a burnt forest region. 5.2. Material and methods 5.2.1. Soils The matrix for the incubation experiments derived from Cambisols (IUSS Working Group WRB, 2014) located in the Sierra de Aznalcóllar, close to Sevilla in the Southwest of Spain. In December of 2011, two different soils were sampled, both covered with Quercus forest. The first derived from an unburnt (37° 32´N and 6° 15´W) and the second from a burnt area (37° 30´N and 6° 19´W) which suffered an intense fire in 2004, seven year prior of our sampling. The soil was sieved (2 mm) and oven-dried (40 ºC). A more detailed description of the soil characteristic and the soil organic matter (SOM) composition is given in (López-Martín et al. 2016b). 86 The impact of soil disturbance on N and C in Mediterranean soils 5.2.2. Production of 15N-enriched organic matter (15N-OM) and pyrogenic organic matter (15N-PyOM) A detailed description of the production of the plant and char residues is given in (LópezMartín et al. 2016a). In brief, Lolium perenne was grown, under stable greenhouse conditions at 24 ± 2/17 ± 2 ºC (16 h day/ 8 h night) in two different trays using material from the unburnt soil as substrate. One tray was irrigated with deionized water in order to produce unlabeled organic matter (OM) and the second tray was watered every week with a K15NO3 solution (99 atom %, 0.5 g L-1) to obtain 15N-enriched organic matter (15N-OM). The aboveground mass of both experiments was cut every two weeks and dried at 40 ºC. The harvests of every cut were homogenously mixed. Pieces of 15N-OM and OM were placed in separate preheated ceramic trays and charred in a Muffle furnace at 350 ºC for 8 minutes in the presence of oxygen in order to obtain 15N-PyOM and PyOM. The temperature was selected to produce char comparable to that remaining on soil surface after natural wildfires. After the heat treatment all 15N-PyOM were combined and homogenized. The PyOM yields were treated comparably. Both samples were ball-milled and had the following characteristics OM (6.2 ± 0.1% Nt, 0.373±0.002 atom % 15N), PyOM (2.8 ± 0.2 Nt, 0.401±0.000 atom % 15N), 15N-OM (3.0± 0.2 Nt, 54.495±0.332 atom % 15N) and 15N-PyOM (4.3± 0.4 Nt, 20.641±0.261 atom % 15N). 5.2.3. Incubation experiment For the pot experiment, 0.25 g of L. perenne seeds were added on 100 g soil material which either derived from the burnt or the unburnt area. In total sixteen plastic pots were prepared. After seeding, both the unburnt and burnt soils were amended with either (1) 30 mg of K15NO3 (15Ni), (2) 600 mg of 15N-OM, (3) 300 mg of non-enriched plant residues and 30 mg of K15NO3 (OM + 15Ni), (4) 300 mg of 15N-OM plus 30 mg of KNO3 (15N-OM + 14Ni), (5) 600 mg of 15NPyOM, (6), 300 mg of PyOM plus 30 mg K15NO3 (PyOM+15Ni), or (7) 300 mg of 15N-PyOM plus KNO3 (15N-PyOM+14Ni). As controls, soils without amendment (8) were prepared. For each variation, two series of six samples were prepared. The amount of added amendment was adjusted in such a manner that the C/N ratio of the soil + amendment remained constant. For further details of the amendments see (López-Martín et al. 2016a). The pots were watered every 3rd day with 30 mL with deionized water without extra nitrogen inputs during the incubation under controlled greenhouse conditions at 24 ± 2/17 ± 2 ºC (16 h day/ 8 h night). The pots were perforated and the leachate was recovered once per months after watering in excess. However, the yields were too low for further analysis. During the first experimental months, the plant 87 Chapter 5. BN as a N-source in presence of inorganic N shoots were harvested monthly due to the fast aboveground-plant growth but thereafter the leaves were cut every four months. After 0.5, 1, 5, 8, 12 and 16 months, two pots of each variation were taken, and the remaining litter or ash layer carefully removed. From the pots, entire plant, root and soil were separated. Roots were gently washed in order to completely remove the soil. All materials were dried in an oven at 40 °C. Afterwards they were ground and stored for further analysis. 5.2.4. Elemental composition and determination of 15N content The Corg and Nt contents of the soil matrix were determined with an elemental analyzer (EA 1108; Carlo Erba, CHNS). The Corg, Nt and δ15N of the incubated soil, the harvested plants and the roots were measured with a Flash 2000 HT combustion elemental microanalyzer and a Flash HT Plus elemental analyzer coupled to a Delta-V advantage isotopic ratio mass spectrometer (IRMS) via ConFlo IV interfase (Thermo Scientific, Bremen, Germany), with an analytical measurement error of ± 0.2‰. The % atom 15N excess of the each measured material was expressed in % of labeled amendment added (15Nadd) recovery in leaves, roots and soil. 5.2.5. Solid-state 13C and 15N nuclear magnetic resonance (NMR) spectroscopy In order to concentrate OM and remove paramagnetic ions prior to the NMR analysis, samples were treated with 10 % (v/v) hydrofluoric acid (HF) (Gonçalves et al. 2003). Briefly, 10 g of dried soil sample were weighed into a polyethylene bottle and 40 ml of HF were added. After 2 hours shaking, the samples were centrifuged and the supernatant discarded. The same procedure was repeated four times. Subsequently the concentrated OM was flushed with deionized water and freeze-dried. The solid-state cross-polarization (CP) magic angle spinning (MAS) 13C and 15N NMR spectra of the HF-treated soils and the ash/litter layer were acquired with a Bruker Avance III HD 400 MHz using a triple resonance probe for rotors with a diameter of 4mm. The rotor was rotated with 14 kHz. A ramped 1Hpulse was applied during the contact time of 1 ms, using a pulse delay time of 300 ms for 13C and 400 ms in the case of 15N. The quantification of the 13C NMR spectra was performed by the integration of the following five chemical-shift regions: alkyl C (45 to 0 ppm), N-alkyl C (60 to 45 ppm), O-alkyl C (110 to 60 ppm), H/C aryl C (110 to 140 ppm), O/C aryl C (140-160) carbonyl C (220 to 160 ppm) (Knicker et al. 2005). The spinning side bands of the aryl C region were considered by adding 88 The impact of soil disturbance on N and C in Mediterranean soils the intensity between 325 to 300 ppm and 0 to -50 ppm to that of the aryl C region. Quantification of the solid-state 15N NMR spectra was only performed when considerable signal intensity was obtained in the chemical shift region of pyrrolic N (-145 ppm to -245 ppm). In this case, the detectable intensity between -145 and -245 ppm was set to 100% and the contribution of pyrrolic N and amide N (-245 ppm to -285 ppm) were determined. 5.2.6. Statistical analysis The statistical analysis was conducted with the software SPSS Statistic 17.0. The % of 15Nadd in soil, leaves and roots was transformed by square root in order to improve homogeneity of variance and normality. The comparisons between impact of soil type, and the different treatments for each incubation months were performed using 2-way-ANOVA approaches followed by Tukey’s b Post Hoc. We used a p ≤ 0.05 for significance difference between means. 5.3. Results 5.3.1. 15Nadd recovery in leaves, roots and soil After 2 weeks and 1 months of incubation, the aboveground and belowground biomass yields were still too low for an unbiased analysis of 15Nadd with the IRMS equipment. (Fig. 5.1a/b). During the incubation experiment, no significant impact of the kind of treatment or soil on biomass production was detected. After five months, the N-contents of the harvested grass residues were between 1.0 and 3.2% of the total dry weight. After eight months the values increased to 2.1 and 3.6% s and accounted for 1.7 to 2.9 % of the dry weight at the end of the experiment after sixteen months (Table A of the appendix). After 5, 8, 12 and 16 months, the total amount of 15N, which was incorporated into all aboveground plant residues or roots of each amendment, was determined. In order to identify the contribution of 15Nadd to the total 15N of the harvested biomass, the amount of 15N, which had been accumulated in the plants of the control was subtracted and the difference is given as the percentage of 15Nadd in Table 5.1a. Here, the different small letters at each single incubation month (per column) indicate significant differences among 15Nadd recovery from the applied amendments. Even after five months, only small amounts of 15Nadd were recovered within the roots. Statistical analysis revealed that neither fire history of the used soil matrix, nor sampling date had an impact on the accumulation of 15Nadd in the root system. Statistical differences are only 89 Chapter 5. BN as a N-source in presence of inorganic N observed with respect to the used amendments (p=0.000) (Fig. 5.2a). Accordingly, the roots in the pots with 15N-PyOM+Ni and 15N-PyOM toppings showed the lowest 15N uptake accounting in both amendments with 0.3 % of 15Nadd whereas the 15N-OM+Ni and 15N amendments led to higher values of 1.1 and 1.3 % of 15Nadd. Comparable to the statistical analysis of 15Nadd uptake by the roots, that of the 15Nadd recovery in the soil unveiled no impact of the fire history of the soils but differences are observed with respect to incubation time (p=0.000) (Fig. 5.2b/c). Unbunrt soil b) Burnt soil a) Unbunrt soil Fig. 5.1.a: Mass recovery leaves in a) unburnt b) burnt soil at the end of each incubation period 90 The impact of soil disturbance on N and C in Mediterranean soils With the exception of the pots with 15N-OM+Ni and 15N-OM amendments, the recovery of 15Nadd with the soil matrix diminished with incubation time. This decrease is most clearly expressed in the experiment with 15Ni additions. This decrease of the latter is partially caused by the quick use of easily available 15N for plant growth but leaching may also contribute to the 15Nadd loss. The highest 15Nadd incorporations into the soil were reached with the 15N-PyOM+Ni and 15N-PyOM toppings whereas 15N-OM+Ni and 15N-OM amendments accounted for the smallest 15Nadd recovery until the first month. However, here, one has to bear in mind that only 15Nadd, which had already entered the soil matrix was considered and some of the 15Nadd of the burnt and unburnt OM still remained in the litter layer. Fig. 5.1.b: Mass recovery roots in a) unburnt b) burnt soil at the end of each incubation period a) Unbunrt soil b) Bunrt soil 91 Chapter 5. BN as a N-source in presence of inorganic N Table 5.1.a. Recovery (%) of 15N from15Nadd applied as labeled and unlabeled fresh (OM) and pyrogenic organic (PyOM) with and without labeled (15Ni) and unlabeled (Ni) N and its control in roots, leaves and soil at the end of each incubation time (5, 8, 12 and 16 months). Amendments Roots Months 5 8 12 16 C 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 15Ni 2.7 ± 2.4 0.9 ± 0.1 0.9 ± 0.4 1.3 ± 1.6 OM+15Ni 0.4 ± 0.2 0.5 ± 0.4 1.3 ± 0.5 0.1 ± 0.3 15N-OM 1.3 ± 0.2 1.0 ± 0.1 0.7 ± 0.2 1.0 ± 1.0 15N-OM+Ni 1.7 ± 0.4 1.0 ± 0.1 1.3 ± 0.6 0.8 ± 0.7 PyOM+15Ni 0.6 ± 0.1 0.5 ± 0.1 1.3 ± 0.2 0.4 ± 0.0 15N-PyOM 0.5 ± 0.2 0.4 ± 0.1 0.2 ± 0.1 0.4 ± 0.2 15N-PyOM+Ni 0.4 ± 0.1 0.3 ± 0.1 0.2 ± 0.3 0.3 ± 0.3 Amendments Leaves Months 5 8 12 16 C 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 15Ni 23.3 ± 13.6 26.2 ± 4.3 14.4 ± 2.2 39.6 ± 23.4 OM+15Ni 15.6 ± 9.4 32.7 ± 11.2 17.4 ± 1.7 6.1 ± 5.0 15N-OM 10.2 ± 0.4 11.0 ± 3.7 7.4 ± 0.7 17.5 ± 4.3 15N-OM+Ni 12.8 ± 0.3 14.7 ± 3.7 9.9 ± 1.5 15.6 ± 7.3 PyOM+15Ni 12.5 ± 6.5 23.1 ± 3.8 22.8 ± 1.3 11.5 ± 7.8 15N-PyOM 2.7 ± 1.1 2.2 ± 0.2 1.0 ± 1.1 2.0 ± 0.6 15N-PyOM+Ni 2.2 ± 0.9 2.0 ± 0.0 1.5 ± 1.0 3.2 ± 1.2 Table 5.1.b. Recovery (%) of 15N from 15Nadd applied as labeled and unlabeled fresh (OM) and pyrogenic organic (PyOM) with and without labeled (15Ni) and unlabeled (Ni) N and its control in soil at the end of each incubation time (0.5 and 1 months). Amendments Soil Months 0.5 1 5 8 12 16 C 0.0±0.0 0.0±0.0 0.0±0.0 0.0±0.0 0.0±0.0 0.0±0.0 15Ni 40.7±16.5 55.3±1.6 19.5±3.0 17.5±1.9 13.0±1.1 21.7±9.3 OM+15Ni 63.2±3.6 39.6±15.3 9.9±4.5 16.1±7.4 14.4±1.6 7.0±5.0 15N-OM 10.5±0.4 18.8±8.1 24.7±9.5 26.7±1.2 24.4±0.9 19.0±11.6 15N-OM+Ni 15.9±5.7 17.4±6.1 35.2±2.3 38.3±7.2 25.6±1.2 25.4±23.0 PyOM+15Ni 45.7±19.2 53.1±5.1 9.2±5.9 10.4±10.7 11.7±1.5 3.1±0.4 15N-PyOM 23.3±7.3 100.0±0.0 87.5±47.1 35.9±7.0 31.7±2.3 61.7±8.5 15N-PyOM+Ni 97.6±3.4 100.0±0.0 36.8±5.9 45.2±5.4 58.1±18.7 65.1±9.0 98 Chapter 5. BN as a N-source in presence of inorganic N 5.3.5. Solid state of 15N and 13C NMR The solid-state 13C NMR spectrum of unburnt soil from the control pot after 2 weeks of incubation shows its main intensities in the chemical shift region of alkyl C (45 to 0 ppm) and Oalkyl C (110 to 60 ppm) (Fig. 5.6, Table 5.2). The alkyl C-to-O-alkyl C ratio of 1.2 indicates an advanced humification stage of its OM. Comparable 13C intensity distributions were determined for the unburnt soils amended with organic and inorganic N forms after 1 month of incubation. Note that the addition of PyOM or OM was not high enough to affect the bulk chemical composition of the SOM. After eight months, a slight decrease of the relative intensity in the Oalkyl C region from an average value of 28% to 25% and slight increases in that of carboxyl C and alkyl C are evidenced. Table 5.2. Intensity distribution (%) of different C groups by 13C NMR measurement in soil at the end of the experiment for each amendment at month 1th and 8th Time (month) carboxyl C O-aryl C aryl C O-alkyl C N-alkyl C alkyl C 0.5 s Control 6 5 18 29 8 34 1 15Ni 11 6 18 29 8 28 OM+15Ni 9 5 16 29 8 34 15N-OM 9 5 17 27 9 32 15N-OM+Ni 9 5 18 27 9 33 15N-PyOM 9 5 17 27 7 34 15N-PyOM+Ni 8 6 14 30 8 35 PyOM+15Ni 12 6 16 26 7 33 8 15Ni 11 6 18 27 8 31 OM+15Ni 7 5 19 25 9 36 15N-OM 8 5 14 25 10 38 15N-OM+Ni 10 6 17 26 8 33 15N-PyOM 10 6 18 25 7 34 15N-PyOM+Ni 11 5 16 25 9 34 PyOM+15Ni 11 6 18 25 8 33 99 Chapter 5. BN as a N-source in presence of inorganic N Figure 5.7 shows the respective solid-state 15N NMR spectra of the HF-treated bulk soils after 1 month and 8 months of incubation. All are dominated by the signal in the amide-N region (-245 to -285). Signals from nitrate N (-3.5 ppm) are even not observed in the spectra fertilized with 15Ni, indicating an efficient transformation of Ni into organic N forms, mostly peptides during the experiment. However, here one has to bear in mind that the used NMR acquisition parameters were optimized for organic N and inorganic N-salts may be underestimated. The 15N NMR spectra of the soils of pots to which 15N-PyOM was applied show a pronounced shoulder in the chemical shift region between -145 and - 245 ppm. The latter is assigned to pyrrole-type N and is typical for BN. Its appearance clearly demonstrates the incorporation of BN into the soil. However, its relative contribution to the detectable 15N-intensity between -145 and -285 ppm decreases between the 1st and the 8th months of incubation, from 44% and 38% to 24% and 31% in the spectra of soils amended with 15N-PyOM and with 15N-PyOM + Ni, respectively. It seems that some of the incorporated BN has been degraded and the mobilized N was used for the builtup of peptides most tentatively of microbial biomass. The latter was recently indicated by studying the 15N-recovery in the amino acid/peptide fraction of those soils (López-Martín et al. 2016a). Fig.5.6. Solid-state 13C NMR spectra of soils obtained from pots treated with 15N-PyOM after 1 and 8 months of incubation. 100 The impact of soil disturbance on N and C in Mediterranean soils Fig. 5.7. Solid-state 15N NMR spectra of the HF-treated bulk soils after 1 month and 8 months of incubation. The numbers refer to the relative contribution of pyrrole-N and amide N to the total 15N intensity. 0 -100 -400-200 -300 ppm 15N-PyOM+Ni 15N-OM+Ni 15N-PyOM 15N-OM PyOM+15Ni OM+15Ni 15Ni 1 months 0 -100 -400-200 -300 ppm 8 months Control Pyrrole Amide Pyrrole Amide 38 62 44 56 31 69 24 76 101 Chapter 5. BN as a N-source in presence of inorganic N 5.3.6. Biochemical transformation of PyOM in the litter layer For a more detailed study of biochemical reworking of the PyOM residues, the litter layers of the pots amended with 15N-PyOM with and without additional Ni fertilization were subjected to solid-state 13C and 15N NMR spectroscopy. Whereas the 13C NMR spectrum of the fresh PyOM (Fig. 5.8, Table 5.3) revealed the typical pattern of burnt N-rich grass residues with high intensity in the aromatic C (160 to 110 ppm) and alkyl C (45 to 0 ppm), the spectra of the litter after 8 months incubation depicted a considerable increase of the O-alkyl C region (110 to 60 ppm) and a doubling of the intensity in the N-alky C region (60 to 45 ppm). The first region is assigned to carbohydrates and the latter either to N-alkyl in amino acids and peptides or amino sugars. Considering that input of plant litter was neglectable, the appearance of those signals is best explained with the development of microbial biomass, soil algae or moss. Their presence was indicated by thin green spots covering the litter layer. The contribution of the latter even increases until the end of the experiment. The respective 15N NMR spectra supports the formation of new biomass by the decrease of the relative intensity assigned to pyrrole and the enhancement of the contribution of the amide N intensity with increasing incubation time. In contrast to the carbon of the new biomass, which may have entered the litter layer by photosynthesis, the 15N contributing to the additional 15N intensity in the chemical shift of amide N must derive from BN (Fig.5. 9). Table 5.3. Intensity distribution (%) in the 13C NMR spectra of the litter layer obtained from the pots treated with 15N-PyOM and 15N-PyOM+Ni after 8 and 16 months of incubation carboxyl C O-aryl C aryl C O-alkyl C N-alkyl C alkyl C 15N-PyOM (Control) 10 12 48 6 3 22 15N-PyOM+Ni (8 months) 13 10 31 19 6 21 15N-PyOM+Ni (16 months) 11 8 23 28 7 23 15N-PyOM (8 months) 12 10 33 18 5 23 15N-PyOM (16 months) 10 9 30 20 6 26 102 The impact of soil disturbance on N and C in Mediterranean soils Fig. 5.8. Solid-state 13C NMR spectrum of the fresh PyOM and the litter after 8 and 16 months of incubation Fig. 5.9. Solid-state 15N NMR spectra of the fresh PyOM and the litter after 8 and 16 months of incubation. The numbers refer to the relative contribution of pyrrole N and amide N to the total 15N intensity. 103 Chapter 5. BN as a N-source in presence of inorganic N 5.4. Conclusion Our studies on the short-term effect of PyOM addition after vegetation fires on the distribution of N between plants and soil did not reveal notable alterations due to fire history of the soil, although an impact of the latter on N-sequestration into soil peptides were recently observed (López-Martín et al. 2016a). The fact that regardless of the kind of N-amendment the amount of produced biomass and its N content were comparable to that of the control pots confirms that our treatments did not alter the overall N cycling and that the system was not under N-limitation. Under those conditions, the availability of additional Ni did not decrease the efficiency of 15N uptake from unburnt or burnt OM. Thus, negative priming of Ni is not evidenced and it can be concluded that the kind and amount of 15N which is taken up by the plants depends mainly on the bioavailability of the 15N, which again is determined by the overall degradability of the 15N-containing organic source. Although slower than N in unburnt litter, we were able to evidence degradation of BN and thus the mobilization of its BN. In summary, our results point to the conclusion that incorporation of charred organic residues and ash into the soil after a forest fire can indeed alter N-cycling in soil, although this alteration will be mainly effective on a short and medium term. With the ash, inorganic N enters the system which refills the N-stocks which without the fire had been supplemented by mobilized N from decaying litter. This fertilization provides faster bioavailable N than only litter degradation alone and allows a quick recovery of the ground vegetation and microbial activity. Unfortunately, Ni, which has not been recycled into plant and microbial biomass, will be lost by volatilization or by leaching during upcoming raining events. On the other hand, the charred residues provide with BN an additional N source, which will be available only after efficient degradation of PyOM. Thus, this PyOM can be seen as a slow-N release fertilizers which prevents N-losses from the ecosystem and provides additional N for the next vegetation cycles. Since with each N-cycle, a certain part of the BN is converted into biomass N, on a long-term scale, almost all of the BN will be transformed into SON with comparable properties and functions within the N-cycle in soils as SON derived from unburnt litter. However, considering the importance of a fast development of a new plant cover after fire events to avoid soil erosion and desertification, our study clearly underlined the important ecological role of BN in fire-prone ecosystems, a fact that certainly asked for an improved consideration of BN in environmental research. 104 The impact of soil disturbance on N and C in Mediterranean soils 5.5. References Almendros G, Knicker H, González-Vila FJ (2003): Rearrangement of carbon and nitrogen forms in peat after progressive thermal oxidation as determined by solid-state 13Cand 15NNMR spectroscopy. Organic Geochemistry 34, 1559-1568 de la Rosa JM, Knicker H (2011): Bioavailability of N released from N-rich pyrogenic organic matter: An incubation study. Soil Biology and Biochemistry 43, 2368-2373 Gonçalves CN, Dalmolin RSD, Dick DP, Knicker H, Klamt E, Kögel-Knabner I (2003): The effect of 10% HF treatment on the resolution of CPMAS 13C NMR spectra and on the quality of organic matter in Ferralsols. Geoderma 116, 373-392 Hilscher A, Knicker H (2011): Degradation of grass-derived pyrogenic organic material, transport of the residues within a soil column and distribution in soil organic matter fractions during a 28 month microcosm experiment. Organic Geochemistry 42, 42-54 Karhu K, Dannenmann M, Kitzler B, Díaz-Pinés E, Tejedor J, Ramírez DA, Parra A, Resco de Dios V, Moreno JM, Rubio A, Guimaraes-Povoas L, Zechmeister-Boltenstern S, Butterbach-Bahl K, Ambus P (2015): Fire increases the risk of higher soil N2O emissions from Mediterranean Macchia ecosystems. Soil Biology and Biochemistry 82, 44-51 Knicker H, Totsche KU, Almendros G, González-Vila FJ (2005): Condensation degree of burnt peat and plant residues and the reliability of solid-state VACP MAS 13C NMR spectra obtained from pyrogenic humic material. Organic Geochemestry. 36, 1359-1377 Knicker H (2007): How does fire affect the nature and stability of soil organic nitrogen and carbon? A review. Biogeochemistry 85, 91-118 Knicker H (2010): “Black nitrogen” – an important fraction in determining the recalcitrance of charcoal. Organic Geochemistry 41, 947-950 Lehndorff E, Roth PJ, Cao ZH, Amelung W (2014): Black carbon accrual during 2000 years of paddy-rice and non-paddy cropping in the Yangtze River Delta, China. Global Change Biology 20, 1968-1978 López-Martín M, Nowak KM, Milter A, Knicker H (2016a): Incorporation of N from burnt and unburnt 15N grass residues into the peptidic fraction of fire affected and unaffected soils. Journal of Soils and Sediments, 1-11 López-Martín M, Velasco-Molina M, Knicker H (2016b): Variability of the quality and quantity of organic matter in soil affected by multiple wildfires. Journal of Soils and Sediments 16, 360-370 Mayor AG, Valdecantos A, Vallejo VR, Keizer JJ, Bloem J, Baeza J, González-Pelayo O, Machado AI, de Ruiter PC (2016): Fire-induced pine woodland to shrubland transitions in Southern Europe may promote shifts in soil fertility. Science of The Total Environment 573, 1232-1241 105 Chapter 5. BN as a N-source in presence of inorganic N Rodríguez A, Durán J, Fernández-Palacios JM, Gallardo A (2009): Short-term wildfire effects on the spatial pattern and scale of labile organic-N and inorganic-N and P pools. Forest Ecology and Management 257, 739-746 Capítulo 6. Incorporación del N procedente de residuos quemados y sin quemar de césped marcado con 15N en la fracción péptidica de suelos afectados y sin afectar por incendios Chapter 6. Incorporation of N from burnt and unburnt 15Ngrass residues into the peptidic fraction of fire affected and unaffected soils 114 The impact of soil disturbance on N and C in Mediterranean soils Since the pH of both soils were below 7 (6 for the unburnt, 5 for the burnt), the determined Ct corresponds to Corg. Table 6.1. Amount of 15N (15Nadd), total N (Nadd) and Corg/add, amended per pot to the unburnt (U) and burnt (B) soil. Note that Nadd includes both the 14N and 15N fraction. The Cp/Np (w/w) ratios correspond to the total organic C and N of the soils in the pots after amendment (burnt and unburnt litter or inorganic N) but before their incubation. Here, the Cp content of the pot was calculated by summing up Corg of the original soil and the organic C added with the amendment (Corg/add), and Np corresponds to the sum of the Nt of the original soil plus the N added with the amendment (Nadd) Amendment K15NO3 (15Ni) 15N-OM OM + 15Ni 15N-OM +Ni (KNO3) 15NPyOM PyOM + 15Ni 15N-PyOM + Ni 15Nadd (mg) 4.36 10.20 4.43 5.11 5.51 4.39 2.77 Nadd (mg) 4.41 17.46 22.80 12.88 24.90 12.33 10.38 Corg/add (mg) 0.0 334.8 115.5 117.4 332.2 111.6 117.4 U C/N (w/w) 17 17 16 17 17 17 17 B C/N (w/w) 22 22 21 22 21 22 22 6.2.4. Extraction and analysis of amino acids yielded after hydrolysis Approximately 1 g of incubated soil or 0.2 g of fresh and charred organic amendments were weighted into a glass bottle (15 ml) and mixed with 5 ml of 6 M HCl and 0.05 mg of Lnorleucine as an internal standard. L-norleucine was used for quantification of the AAs loss during the purification steps. The soils and the organic amendments were hydrolyzed at 110 °C, for 22 hours and under N2 atmosphere. After the hydrolysis, the samples were filtered through glass-fiber membrane filters (0.07 µm, Wicom Perfect Flow, Germany) and the hydrolysate was dried under a flow of N2 to remove the HCl. The dried hydrolysates were re-dissolved in 4 ml of 0.1 M HCl and 0.05 mg of trans-4-(aminomethyl) cyclohexane carboxylic acid was added as a second internal standard Nowak et al. (2011). The solution containing the hydrolyzed and free AAs was purified by passage over H+ exchanged DOWEX 50 W X8 resin. Prior to elution of the AAs with 2.5 M ammonium hydroxide, the impurities were washed out with 25 ml 0.1 M oxalic acid and then with 5 ml 0.01 M HCl and 5 ml of distilled water which were used to eliminate residues of oxalic acid. The carboxylic groups of AAs were esterified with isopropanol / acetylchloride (1:4 v/v; 1 h, 110 ºC) and the amino groups were trifluoroacetylated with 1 ml of trifluoroacetic anhydride/dichlormethane (1:1 v/v; 1 h, 60 ºC) (Miltner et al. 2009). After 115 Chapter 6. Incorporation of 15N-grass residues into soil peptidic fractions derivatization the impurities were extracted into the aquatic phase of a mixture of chloroform/ phosphate buffers, and the chloroform phase was dried under N2 (Ueda et al. 1989) and stored at -4 °C for subsequent analysis. Samples were reconstituted in 100 µl dichloromethane, and the derivatized AAs were identified and quantified in duplicates by means of gas chromatography-mass spectrometry (GC– MS). A BPX5 column (30 m × 250 µm × 0.25µm; SGE, TOWN, COUNTRY) and 7890A GC System (Agilent Technologies, Waldbronn, Germany) with a detector 5975C inert XL MSD (Agilent Technologies) were used. The initial temperature of 50 ºC was kept for 5 min. Then 100 ºC was reached with 30 ºC min-1 and held for 5 min. After that the temperature increased at 10 ºC min-1 to 175 ºC were it remained constant for 5 min, then further heated to 250 ºC at 10 ºC min-1, held for 5 min, and to 325 ºC at 30 ºC min-1, hold for 5 min. The injection was performed at a split ratio of 1:20 and at an injector temperature of 280 ºC. For identification and quantification of the individual AAs an external standard (200 µl) containing alanine, glycine, threonine, serine valine, leucine, isoleucine, cysteine, proline, aspartic acid, methionine, glutamic acid, phenylalanine, tyrosine, lysine, histidine and cysteine at a concentration of 2.5 µmol ml-1 for all AAs except for cysteine which had a concentration of 1.25 µmol ml-1 was used. For the calculation of the contribution of AAs, the recovery of individual AAs summed up after they have been identified by comparing the retention time and the mass spectra with the external standard, using the MSD ChemStation software (Agilent Technologies). In order to determine the content of 15N of the total AAs (15NAAs), the 15N of the individual derivatized AAs were measured by GC–combustion–isotope ratio–MS (GC–C–irMS) and summed up. Therefore, the AAs were separated with a 7890 A GC System (Agilent Technologies) equipped with a BPX5 column (50 m × 0.32 mm × 0.5 µm) using the following temperature program: Increase of the temperature with 10 ºC min-1 from 50 ºC to 80 ºC, which was kept for 7 min before 120 ºC were reached with 3 ºC min-1 and hold for 5 min; heating to 210 ºC at 3 ºC min-1 After 5 min, the temperature was increased at 20 ºC min-1 to 300 ºC for 5 min. The samples were injected in the splitless mode at an injector temperature of 250 ºC. The eluting compounds were combusted and the resulting N2 was analyzed for its isotopic composition by means of a Finnigan MAT 253 IRMS (Thermo Finnigan, Bremen, Germany). In order to determine the amount of 15N recovered from 15Nadd of each treatment in the respective AAs fractions, the 15NAAs content of the control soil with natural 15N abundance was subtracted from the 15NAAs contents of the amended soils. 116 The impact of soil disturbance on N and C in Mediterranean soils 6.2.5. Solid-state 15N NMR spectroscopy Prior to NMR analysis the soil samples were demineralized with 10 % (v/v) hydrofluoric acid (HF) (Gonçalves et al. 2003) in order to remove paramagnetic ions and to concentrate the OM. Briefly, 10 g of dried soil sample were weighed into a polyethylene bottle and 40 ml of HF were added. The closed bottles were shaken for 2 hours. After centrifugation the supernatant was removed and discarded. The same procedure was repeated four times. The concentrated OM was washed with deionized water and freeze-dried. The solid-state cross-polarization (CP) magic angle spinning (MAS) 15N-NMR spectra of the HF-treated soils were acquired with a Varian 7.05T Unity Inova (15N resonance frequency: 60.8 MHz) and the fresh and charred OM were obtained with a Bruker DMX 400 (15N resonance frequency: 40.6 MHz) Bruker Avance III 600 (15N resonance frequency: 60.8 MHz) using a spinning speed of 8, 4 and 15 kHz, respectively. A ramped 1Hpulse was applied during the contact time of 0.7 ms. Using a pulse delay time of 200 ms, 7500 scans were accumulated for the spectra of the labeled fresh and charred OM. For the 15N NMR spectra of the soil samples, 1,000,000 scans were acquired with a contact time of 1 ms and a delay time 0.4 s. 6.3. Statistical analysis The statistical analyses were accomplished using the software SPSS Statistic 17.0. Differences between the results obtained from different sampling events were evaluated using Mann-Whitney test. The impact of soil type and the effect of the substrate amendment on the change of the amount of newly synthesized AAs were analyzed using the Wilcoxon test. In order to reveal the impact of the fire history on NAAs extractability and 15N incorporation into NAAs, we statistically combined the results of all variations (both with amendment and with incubation time) of each soil. For more detailed information about the impact of the source material on the incorporation of 15Nadd in AAs of soils, the treatments were grouped into the following four sets, I) control (C), II) inorganic source (15Ni, OM+15Ni, PyOM+15Ni), III) organic litter (15NOM+14Ni, 15N-OM and III) charred organic matter (15N-PyOM+14Ni, 15N-PyOM). Finally, with the aim to study the effect the presence of inorganic nitrogen on the use and degradation of organic N we compared the relative contribution of 15NAAs to NAAs in the experiments with addition of 15Ni, 15N-OM or 15N-PyOM from burnt and unburnt soil. A p-value ≤ 0.05 was considered as statistically significant. 117 Chapter 6. Incorporation of 15N-grass residues into soil peptidic fractions 6.4. Results and discussion 6.4.1. Elemental composition of organic C, Nt and 15N enrichment in the soils, OM and the PyOM Table 6.2 lists the organic C, Nt and 15N contents of PyOM, OM and the burnt and unburnt soils. Recent statistical evaluation of SOM alterations due to fire in the probed area indicated that 7 years after the event, the Corg and Nt values are only slightly higher in the unburnt area. A comparable observation is reported by Alcañiz et al. (2016) for soils recovered for 9 years after a prescribe fire in a Mediterranean area. The non-enriched and 15N-enriched Lolium perenne showed C/N-values between 6 and 13, which is in the range found in other studies (de la Rosa & Knicker 2011, Hilscher & Knicker 2011, Knicker & Lüdemann 1995). Note, the respective values for the PyOM and 15N-PyOM are within this range, which indicates that in spite of N-losses during heating a considerable fraction of the organic N was incorporated as BN into the charred material. For both, unburnt and charred plant residues, the 15N content is 0.373 and 0.401 atom %, which agrees with the natural 15N abundance. The slightly higher value found for PyOM may be caused by the loss of N-volatile compounds during combustion leading to a depletion of the lighter isotope (Fraser et al. 2013). The atom % 15N for unburnt and burnt soil range between 0.355 to 0.377 which also corresponds with the natural 15N abundance in organic material (Robinson 2001). The 15N abundance in 15NOM is with 54.495 ± 0.332 atom % considerably higher than in 15N-PyOM (20.641 ± 0.261 atom %), which is best explained with the fact that the source material of those residues derived from two different growing experiments. 6.4.2. Distribution of N-forms in the starting materials The 15N NMR spectra (Fig. 6.1) of the burnt and unburnt soils are dominated by the signal between -240 and -285 ppm which is assigned to amide N (Knicker & Lüdemann 1995). Note, indole-type N and proline N may also contribute to the region between -240 and -250 ppm. A small signal appears at -346 ppm which is most likely caused by N bound to Cε in lysine and by N in other free amino groups of amino acids and amino sugars (Knicker 2011c, Witanowski et al. 1993). The solid-state 15N NMR spectrum of the burnt soil shows no major intensity in the BN-typical region of pyrrole-type and indole-type N between -145 to -250 ppm ppm (Knicker 2010). Since in other studies of the same sampling area, signals of BN were clearly dominating the solid-state 15N-spectrum of a burnt soil collected 4 weeks after an intense fire (Knicker 118 The impact of soil disturbance on N and C in Mediterranean soils 2011b), the low intensity in the chemical shift region of heterocyclic N may indicates that BN has been partially remove either by degradation, erosion or leaching. The dominance of amide N in this spectrum, most tentatively of biogenic origin, is in line with recent results obtained by 13C NMR spectroscopy, revealing a fast recovery of the SOM to its pre-fire composition (LópezMartín et al. 2016). In contrast to the solid-state 15N NMR spectrum of the fresh grass material, which is dominated by signals of peptides (de la Rosa & Knicker 2011, Knicker & Lüdemann 1995), the 15N NMR spectrum of PyOM confirms the presence of BN by clear signals peaking at -235 and - 245 ppm which are typical for pyrrole N and indole N. However, some intensity is still recovered between -250 and -285 ppm which may indicate that not all peptides were transformed into heterocyclic N. 6.4.3. Total amino acids and total N of the extracted AAs in the starting materials The contents of extracted AAs in the control soil of the unburnt area and in that of the nearby burnt region were 2.9 ± 0.2 mg AAs g soil -1 and 4.0 ± 0.6 mg AAs g soil -1, respectively (Table 6.2). These values are in the range of those found for soils studied by (Amelung and Zhang (2001) where the total AAs varied between 0.5 and 16.0 mg AAs g soil -1. In general, only a Fig. 6.1. Solid state 15N NMR spectra of an unburnt and a burnt Cambisol from the Sierra Aznalcóllar (Spain), 15N-labeled organic matter (15N-OM) and pyrogenic organic matter (15N-PyOM) produced from Lolium perenne at 350°C for 8 min. Table 6.2. Corg, Nt, atom % 15N, extractable AAs contents and contribution of the extractable AAs and its nitrogen (NAAs) to total N (Nt) in 15Nenriched and non-enriched organic matter and charred material and in an unburnt and a burnt Cambisol from Sierra de Aznalcóllar (Spain) Corg (mg g dry matter -1) Nt (mg g dry matter -1) atom % 15N Total AAs (mg g dry matter-1) AAs (mg g Nt -1) NAAs (% of the Nt) Unburnt soil 69.2 ± 1.4 (a) 4.1 ± 0.3 (a) 0.369 ± 0.002 2.9 ± 0.2 701.48 ± 39.55 (b) 8.79 ± 0.73 (b) Burnt soil 57.6 ± 1.9 (b) 2.6 ± 0.0 (b) 0.370 ± 0.001 4.0 ± 0.6 1480.75 ± 228.29 (a) 18.90 ± 3.08 (a) OM 385.4 ± 4.4 61.7 ± 1.3 0.373 ± 0.002 142.2 ± 45.6 2305.11 ± 522.99 (b) 30.45 ± 7.85 (b) 15N-OM 391.2 ± 2.5 29.6 ± 2.2 54.495 ± 0.332 112.4 ± 3.4 3990.31 ± 276.18 (a) 53.01 ± 3.53 (a) PyOM 371.2 ± 16.5 27.8 ± 1.9 0.401 ± 0.000 4.9 ± 1.1 175.06 ± 40.76 (ns) 2.51 ± 0.62 (ns) 15N-PyOM 386.9 ± 19.4 43.0 ± 3.6 20.641 ± 0.261 5.8 ± 0.0 134.25 ± 0.90 (ns) 1.93 ± 0.02 (ns) Values followed with different letters within the same column are significantly different at p < 0.05. ns: no significant differences. Table 6.3. Average contents of the extractable AAs in the burnt (B) and unburnt (U) soils (mg AAs g soil-1), which were incubated with different amendments as a function of incubation time 0.5 month 1 month 8 months 16 months B U B U B U B U C 5.7 ± 0.2 6.4 ± 0.0 4.9 ± 0.2 4.8 ± 1.0 3.0 ± 0.5 4.8 ± 0.2 2.4 ± 0.0 3.9 ± 0.0 15N-PyOM+Ni 2.6 ± 0.8 5.7 ± 0.4 2.9 ± 0.7 4.7 ± 0.1 3.7 ± 0.2 4.3 ± 0.5 2.8 ± 0.0 4.6 ± 0.1 15N-OM+Ni 2.7 ± 0.1 5.8 ± 0.8 2.5 ± 0.1 4.7 ± 0.5 3.7 ± 0.3 4.6 ± 0.6 2.8 ± 0.0 4.9 ± 0.1 15N-PyOM 3.5 ± 0.5 13.2 ± 2.4 3.5 ± 1.2 9.5 ± 6.6 3.8 ± 0.1 4.1 ± 0.2 2.7 ± 0.0 4.4 ± 0.1 15N-OM 6.5 ± 0.8 11.2 ± 0.8 6.8 ± 3.8 5.1 ± 1.2 3.6 ± 0.3 5.1 ± 0.2 3.5 ± 0.0 5.5 ± 0.2 PyOM+15Ni 4.3 ± 0.5 4.5 ± 1.6 2.2 ± 0.4 4.1 ± 0.0 3.0 ± 0.2 3.9 ± 0.3 2.5 ± 0.0 4.6 ± 0.0 OM+15Ni 4.7 ± 0.6 7.1 ± 0.8 2.5 ± 0.9 5.9 ± 1.3 3.5 ± 0.1 4.8 ± 0.2 2.4 ± 0.0 4.7 ± 0.1 15Ni 4.5 ± 1.2 7.8 ± 0.5 4.1 ± 2.3 6.0 ± 0.5 3.5 ± 0.0 4.6 ± 0.0 3.0 ± 0.1 5.0 ± 0.0 120 The impact of soil disturbance on N and C in Mediterranean soils small proportion of Nt of soils is hydrolyzed and recovered as NAAs by the used method. (Friedel and Scheller (2002) or (Amelung et al. (2006) recovered 28 to 50 % and 22 to 46 % of Nt as NAAs. In our approach, the recovery of Nt as NAAs was between 9 and 19 % for the soils (Table 6.2). Higher AAs contents and NAAs recoveries were obtained for OM and 15N-OM. Here the contribution of NAAs to Nt ranged from 31 to 53%. It seems that due to the higher humification degree of SOM, their peptides are better protected from hydrolysis than those in fresh litter. Considerably low amounts of AAs were obtained for PyOM (4.9 ± 1.1 mg g dry material-1). In PyOM and 15N-PyOM 2.51 ± 0.62 and 1.93 ± 0.02 % of Nt accounted for NAAs. Comparably, only 2% of the total 15Nadd in 15N-PyOM was amended as 15NAAs to the soil before starting the experiment. Statistical analysis of the extractable NAAs contents normalized to Nt (Table 6.2) for both soils confirmed that in the fire-affected area the percentage of Nt attributable to AAs is twice the amount determined for the unburnt soil (p = 0.01). Thus, although the N content and the dominance of peptide-N in the soils of the burnt region recovered to the status of the unburnt soil, (López-Martín et al. 2016), the quality of the present peptides seems to be still affected by the former fire. Possibly, the fire history resulted in the production of fresh peptides which are more accessible to hydrolysis than those commonly accumulated in soils during humification. 6.4.4. Extractability of total AAs and NAAs as a function of incubation time During the incubation, the mean concentration of extracted AAs in the soils treated with the different amendments varied between 2.4 to 13.2 mg g soil -1 (Table 6.3). In order to assess the relationship of the AAs extractability with either incubation time, we statistically compared the combined treatments of the unburnt with those of the burnt applying the Wilconxon test. As it can be revealed from Table 6.4, this analysis confirms decreasing recovery of AAs with incubation time. Comparing the results between burnt and unburnt soils showed further that the yields were always higher for the latter. A comparable approach was used for the statistical analysis of the relationship between NAAs extractability with either incubation time or soil type (Fig. 6.2). Here, it should be noted that with ongoing incubation, the Nt content of the soils did not change significantly, neither between unburnt and burnt material nor between the different amendments, indicating that no major N loss by volatilization occurred. NMR spectroscopic data (data not shown) confirmed that no heterocyclic N was formed during the incubation, although for both soils the amount of extractable NAAs decreased from the beginning until the end of the experiment (Fig. 6.2). Thus, the decline of extractable AAs with incubation time points toward 121 Chapter 6. Incorporation of 15N-grass residues into soil peptidic fractions ongoing AAs sequestration and transformation as it was suggested by (Miltner et al. 2009) and also agrees with findings by Creamer et al. (2012) and Nowak et al. (2013). Comparing the experiments with burnt and unburnt soils, we observed a general lower extractability of the NAAs in the burnt soils. However, this difference is only statistically confirmed for the samples analyzed after 8 (p = 0.000) and 16 (p = 0.013) months of incubation. These observations may evidence that in our soils the fire history can indeed affect the N-cycling, although the impact was only clearly discernable at the end of the experiment. 6.4.5. Incorporation of 15Nadd into NAAs Figure 6.3 shows that already two weeks after addition of the amendment, 15Nadd is incorporated into the NAAs. Its amount varied between 2 and 4 % of 15Nadd. The low values may be explained by the facts that (1) part of 15Nadd was also incorporated into the growing plant residues and that (2) not all AAs of the soil were in an extractable form. For the experiment with 15N-PyOM, the recovery of 15NAAs derived from 15Nadd varied between 0.67 and 12.62 % of 15Nadd which in average is slightly higher than the amount of 15NAAs added at the beginning of the experiment. This may allow the conclusion that that BN underwent a microbial transformation into non-heterocyclic residues. This is supported by the observation that already after 4 months approximately 2% of 15Nadd were recovered in the leaves of the freshly grown grass (data not shown). The 15Nadd incorporated into NAAs decrease with time. The assimilation of 15Nadd into NAAs was a bit higher in burnt than in unburnt soil with statistical differences at month 8 (p = 0.039) suggesting that the fire history after 7 years no affected the incorporation of organic N derived from fresh litter into peptideous material of SOM and microbial biomass. 6.4.6. Impact of the N sources on the 15N content of NAAs In order to obtain more detailed information about the impact of the source material on the incorporation of 15Nadd in AAs of soils, the contribution of 15N to the total N of the amino acids (NAAs) was determined and the results were statistically analyzed after grouping the treatments according to their 15Nadd source into the following four sets: I) control, II) inorganic source (15Ni, OM+15Ni, PyOM+15Ni), III) organic litter (15N-OM+14Ni, 15N-OM and III) charred organic matter (15N-PyOM+14Ni, 15N-PyOM). Note the value of 0.366% corresponds to the natural 15N abundance and was found for the control set I (Fig. 6.4). The fact that relative to the control set I, 122 The impact of soil disturbance on N and C in Mediterranean soils Fig. 6.3. Contribution of 15N from 15N-amendments (15Nadd) recovered as 15N of AAs (15NAAs) after incubation of soil material from an unburnt and burnt Cambisol amended with 15Nenriched burnt and unburnt organic matter and inorganic 15N. For statistical reason, the impact of the kind of amendment was not considered. Median values ± interquartile range (Mann-Whitney test, p < 0.005, n = 16). 0.5 1 8 16 0 2 4 6 8 Burnt soils Unburnt soils Incubation time (months) 15NAAs of 15Nadd (%) 0.5 1 8 16 0 10 20 30 40 Unburnt soils Burnt soils Incubation time (months) NAAsof Nt(%) Fig. 6.2. Effect of time on the percentage of Nt which is extractable with the AAs (NAAs) from burnt and unburnt soils amended with N-rich burnt and unburnt organic matter or inorganic N. For statistical reason, the impact of the kind of the amendment was not considered. Median values ± interquartile range (Mann-Whitney test, p < 0.005, n = 16). 123 Chapter 6. Incorporation of 15N-grass residues into soil peptidic fractions all other sets had higher 15NAAs recoveries indicates that 15Nadd has been incorporated into the extractable AAs fraction of the soil. However, since in average the amount of 15Nadd in NAAs in the soil is only slightly higher than the percentage of NAAs in PyOM we cannot unbiasedly differentiate if the recovered 15NAAs originate from the accumulation of PyOM-derived AAs or from AAs which were newly synthesized by microorganisms from BN. In general, the contribution of 15NAAs to the total NAAs increased continuously until the 8th month (Fig. 4) although the recovery of the latter decreased with incubation time (Fig. 6.2). At the end of the experiment time, the contribution of 15NAAs to total NAAs is approximately 46% higher than that obtained after two weeks. From this it can be concluded that peptidic pool suffered a fast turnover in which the original AAs were mobilized and replaced by new peptides with AAs containing 15Nadd. Whereas the first was most likely used for the synthesis of new biomass, the latter may have been released from biomass which already had been incorporated 15N from the amendment. Possible sources of the released material are decaying residue, exudates or as exo-enzymes. Statistical analysis confirmed that differences also can be discerned between samples analyzed after the same incubation time but with different amendments (p < 0.05) (Table 6.4). Compared with the experiments amended with 15N-PyOM those with the addition of 15N-OM always showed higher 15Nadd contribution to NAAs (Fig. 6.4). However, here 0.5 1 8 16 0.000 0.005 0.010 0.015 0.020 C 15N-PyOM 15N-OM 15Ni Incubation time (months) 15NAAsinNAAs(mg mg-1) Fig. 6.4. Contribution of 15N (15NAAs) to N of AAs (NAAs) after incubation of soil material from an unburnt and burnt Cambisol amended with 15N-enriched burnt and unburnt organic matter and inorganic 15N. In order to evaluate the impact of the source material, the impact of the fire history of the soil was not considered and the results were grouped depending on the N source: control (C), inorganic N (15Ni), fresh organic N (15N-OM) and charred organic N (15N-PyOM) Median values ± interquartile range (Wilcoxon test, p < 0.005, n = 8).