Models for estimating biomass and carbon in biomass and soils in Pinus radiata (D. Don), Eucalyptus globulus (Labill) and Eucalyptus nitens (Deane & Maiden) Maiden plantations established in former agricultural lands in northwestern Spain

MODELS FOR ESTIMATING BIOMASS AND CARBON IN BIOMASS AND SOILS IN Pinus radiata (D. Don), Eucalyptus globulus (Labill) AND Eucalyptus nitens (Deane & Maiden) Maiden PLANTATIONS ESTABLISHED IN FORMER AGRICULTURAL LANDS IN NORTHWESTERN SPAIN Departamento de Producción Vegetal Escuela Politécnica Superior César Pérez-Cruzado Doctoral Thesis June 2011 UNIVERSIDAD DE SANTIAGO DE COMPOSTELA ESCUELA POLITÉCNICA SUPERIOR DEPARTAMENTO DE PRODUCCIÓN VEGETAL MODELS FOR ESTIMATING BIOMASS AND CARBON IN BIOMASS AND SOILS IN Pinus radiata (D. Don), Eucalyptus globulus (Labill) AND Eucalyptus nitens (Deane & Maiden) Maiden PLANTATIONS ESTABLISHED IN FORMER AGRICULTURAL LANDS IN NORTHWESTERN SPAIN TESIS DOCTORAL A UTOR: CÉSAR PÉREZ-CRUZADO DIRECTORES: DR. D. ROQUE RODRÍGUEZ SOALLEIRO DR. D. AGUSTÍN MERINO GARCÍA Lugo, Junio 2011 UNIVERSIDAD DE SANTIAGO DE COMPOSTELA ESCUELA POLITÉCNICA SUPERIOR DEPARTAMENTO DE PRODUCCIÓN VEGETAL MODELS FOR ESTIMATING BIOMASS AND CARBON IN BIOMASS AND SOILS IN Pinus radiata (D. Don), Eucalyptus globulus (Labill) AND Eucalyptus nitens (Deane & Maiden) Maiden PLANTATIONS ESTABLISHED IN FORMER AGRICULTURAL LANDS IN NORTHWESTERN SPAIN CÉSAR PÉREZ-CRUZADO INGENIERO DE MONTES Memoria para optar al grado de Doctor realizada bajo la dirección de los Doctores del Departamento de Producción Vegetal y Edafología y Química Agrícola de la Universidad de Santiago de Compostela: DR. D. ROQUE RODRÍGUEZ SOALLEIRO DR. D. AGUSTÍN MERINO GARCÍA Vº Bº El director Vº Bº El director DR. D. ROQUE RODRÍGUEZ SOALLEIRO DR. D. A GUSTÍN MERINO G A RCÍ A El autor D. CÉSAR PÉREZ-CRUZADO Lugo, Junio 2011 El Dr. D. Roque Rodríguez Soalleiro, Profesor Titular del Departamento de Producción Vegetal de la Universidad de Santiago de Compostela y Dr. D. Agustín Merino García, Profesor Titular del Departamento de Edafología y Química Agrícola de la Universidad de Santiago de Compostela, informan: Que la memoria titulada “Models for estimating biomass and carbon in biomass and soils in Pinus radiata (D. Don), Eucalyptus globulus (Labill) and Eucalyptus nitens (Deane & Maiden) Maiden plantations established in former agricultural lands in northwestern Spain”, que para obtener el grado de Doctor Ingeniero de Montes presenta D. César Pérez Cruzado, ha sido realizado bajo nuestra dirección. Considerando que el trabajo está finalizado, y es materia de tesis, autorizamos su presentación. Y para que así conste a los efectos oportunos, firmamos la presente en Lugo a 22 de Junio de 2011. LOS DIRECTORES Fdo. DR. D. ROQUE RODRÍGUEZ SOALLEIRO Fdo. DR. D. AGUSTÍN MERINO GARCÍA Los trabajos llevados a cabo para la realización de esta Tesis Doctoral han sido financiados mediante los siguientes proyectos de Investigación:  “FORSEE: Gestión sostenible de los bosques: Una red de zonas piloto para la puesta en marcha operativa” Proyecto financiado por Iniciativa Comunitaria de la Unión Europea (Ref. 020-FORSEE) mediante el Fondo Europeo de Desarrollo Regional (FEDER - Interreg IIIB Atlantic Area). www.iefc.net  “Captura de carbono en terrenos agrícolas repoblados con plantaciones forestales para uso maderero y energético” Proyecto financiado mediante el Programa Nacional de Recursos y Tecnologías Agroalimentarias (Ref. SUM2006-00006-00-00) por el Ministerio de Educación y Ciencia y el Instituto Nacional de Investigación Agraria y Alimentaria (INIA), dentro de la acción movilizadora “Sumideros agroforestales de efecto invernadero”.  “Estudio de la estabilización del carbono en suelos agrícolas reforestados” Coordinado dentro del proyecto “CRONOCARB”, financiado mediante el Plan Nacional de I+D+I 2008-2011 (Ref. AGL2009-13400-C05-04) por el Ministerio de Ciencia e Innovación y el Fondo Europeo de Desarrollo Regional. Para la realización de la presente Tesis Doctoral, el autor ha contado con una beca predoctoral del Programa de Formación de Profesorado Universitario (FPU-MEC, AP2007-04367) del Ministerio de Educación (BOE 183, 30/07/2008), bajo la supervisión del Profesor Dr. D. Roque Rodríguez Soalleiro, dentro del Grupo de Investigación Unidade de Xestión Forestal Sostible (www.usc.es/uxfs). AGRADECIMIENTOS: Quiero expresar mi sentimiento de gratitud a todas aquellas personas que, de alguna u otra forma, han colaborado en la elaboración del trabajo que aquí concluye. Debido a que son muchas las personas que me han ayudado durante estos años, espero sinceramente no haberme olvidado de nadie. Muy especialmente a mis directores, Roque Rodríguez Soalleiro y Agustín Merino, por la dedicación y el trato recibido en todo momento, así como por haberme contagiado vuestra pasión por la investigación. Espero no haber agotado vuestras ganas de seguir dirigiendo tesis. A Nieves y Pepi del Depto. de Física Aplicada por la inestimable ayuda con los análisis térmicos y NMR, contado por vosotras todo parece más fácil. A los profesores y profesoras del Grupo de Investigación Unidade de Xestión Forestal Sostible Juan Gabriel Álvarez, Guillermo Riesco, Alberto Rojo, Ulises Diéguez, Federico Sánchez, Fina Lombardero, Ana Daría, Carlos López y Manuel Guaita. También a todos los profesores y profesoras del Departamento de Producción Vexetal y Edafoloxía e Química Agrícola de la Escuela Politécnica Superior, a los que he recurrido en numerosas ocasiones. A Almudena Pérez por estar siempre ahí, solucionando sobre la marcha nuestros problemas, muchas gracias por todo. A Frits Mohren, Frans Bongers, Stefan Schnitzer, Ellen Wilderink, Ute Sass-Klaassen, Lourens Poorter, Marielos Peña, Joke Jansen, Jan den Ouden, Frank Sterck, Lars Markesteijn, Jose Quero, Patrick Jansen, Gustavo Schwartz, Corneille Ewango, Meenakshi Kaul, Marisol Toledo, Esron Munyanziza, Hans Polman, Leo Goudzwaard, Lucy Amissah, Geovana Carreño, Abeje Eshete, Motuma Tolera, Neeltje van Hulten y al resto del personal del Forest Ecology and Forest Management Group (Wageningen University), por la dedicación y el buen trato recibido durante mi estancia en Holanda. Heel erg bedankt voor alles! A Fernando Basurco y Roberto Astorga (ENCE), Carlos Tejedor (SNIACE), Gabriel Toval (CIFALourizán), Miguel Balboa (Norvento), Fernando Solla-Gullón (ISEMPA), Gustavo Iglesias (GIT), Ana Cabaneiro e Irene Fernández (CSIC) por las contribuciones como expertos en vuestras respectivas materias. A Edelmiro, Roberto, Pacín, Arias, Jose Luis, Manuel, Fernado, Roberto Vila, Román, Manuel Vázquez, Julio, José Manuel y el resto de propietarios de las parcelas. A Montserrat Gómez, y Verónica Piñeiro por su dedicación incondicional para cumplir los plazos con los análisis de muestras. A Horacio, Samuel, Pablo Mansilla y Juan Daniel por la inestimable ayuda en los trabajos de campo. A Luís, Benja, Elena, Nohe, Noelia, Alberto, Moisés y Hugo por haberos atrevido a hacer el proyecto fin de carrera con nosotros, sin vosotros nada hubiera sido posible. A todos los compañeros y amigos que han estado a mi lado durante toda la etapa de formación, muy especialmente a: Luis Rodríguez DaCosta, Dori, Javi Pereira,Tino, Edu González, Laura, Ismael Castro, Marta Illera, Andrea Ferreiro, David Castañe, Sandra, Fernando Solla, Bea Omil, Cristina Eimil, Esteban, Andrea Hevia, Felipe Crecente, Fernando Pérez, Eva, Iban, Ivonne, Pablo Guindos, Emmanuel, Rafa, Noelia, Vanesa, Susi, Natalia. Muchas gracias por tantos buenos momentos. Y, por supuesto, a toda mi familia, especialmente a mis padres: Arturo, Mercedes, Begoña y Alfonso; a mi hermano David, a mis abuelos y abuelas: Vidalina, Anita, Facundo, Jesusa y Rosa. A mi abuelo Facundo†. Por último y muy especialmente a Bea, por tu paciencia, comprensión y ayuda. Espero vivir lo suficiente para devolverte todo el tiempo que te debo. Muchas gracias a tod@s. INDEX AND ABSTRACT v 0.2. General abstract Climate change is one of the most serious environmental problems nowadays. This has been brought about by the so-called greenhouse effect, caused by the huge release of greenhouse gases, particularly CO2, from the burning of fossil fuels. Energy sectors contribute most to climate change, although land use and land use change also contribute greatly to the greenhouse effect. Despite the observed trend towards deforestation in tropical zones, land use changes in Europe and North America are tending towards an increase in forest cover. This was accelerated in Europe by the implementation of a policy encouraging afforestation of former agricultural land (EEC 2080/92), which led to afforestation of large areas in northern Spain between 1992 and 2006. This usually involved planting fast growing species (Eucalyptus globulus Labill., Eucalyptus nitens (Deane & Maiden) Maiden and Pinus radiata (D. Don)) on former pasture land. International agreements on global C emissions allow the countries involved to compensate for the release of CO2 by providing C sinks. However, there is a large degree of uncertainty associated with the methods used to estimate the amounts of carbon sequestered, mainly when evaluating some compartments of C sinks, even in steady state systems. The process is more complicated when the land use under study has not reached equilibrium, and additional studies are required to evaluate the effects on C stocks. Although estimation of the carbon density in tree biomass is simple, it is much more difficult to estimate changes in carbon density in soils. Most carbon dynamics studies are based on modelling approaches, and studies based on empirical data are required for more consistent assessments, and to provide information for validating eco-physiological model predictions or for developing empirical models. However, direct measuring is time consuming and the results are highly variable and do not always enable significant conclusions to be reached. Changes in soil carbon are difficult to estimate because they occur relatively slowly, and some external variables may also affect the final estimation. Chronosequence sampling combined with pairwise comparison of plots may be a useful technique for translating spatial differences into temporal differences, while also correcting local tendencies in the measured plots, thus enabling estimation of the changes in the variable of interest at landscape level. Carbon in trees is usually estimated from biomass equations, which are supposed to be more useful for biomass estimation than biomass expansion factors. However, estimation of the dry mass of the sample trees used in developing models may not be very accurate. Moreover, the increasing interest in harvesting crown fractions and in carrying out ecological and nutritional studies, leads to the need for more accurate models. The study of crown variables as explanatory variables in biomass equations may lead to the development of more accurate models than those based exclusively on stem variables. Product yield is usually expressed in terms of harvested or produced wood volume, whereas bioenergy substitution and carbon stocks are expressed in terms of energy, or tonnes of oil CHAPTER 0 vi equivalent. This depends on the transformation procedure, which finally leads to estimations from stand volume or biomass by the application of bioenergy production factors. Relating each of these forms of energy or carbon directly by the use of specific models may avoid concatenating errors in the estimation, and at the same time enable comparison of several management alternatives or the use of particular species for energy purposes. The first step in estimating changes in soil organic carbon (SOC) after land use change is to examine changes in total SOC. However, the stability of SOC compounds varies greatly, and traditional techniques for describing these processes are time consuming and expensive. The development of inexpensive and rapid alternative methods for assessing changes in soil stability is a key factor for large scale or high temporal resolution studies. When evaluating the whole forest sector in terms of climate change mitigation, carbon accumulation in wood products and the associated dynamics must be considered, although it is still not clear how this pool will be accredited in international emissions agreements. The most commonly used techniques are those based on mechanistic models, e.g. the CO2Fix model, which enables estimation of the mitigatory effect of several management alternatives by considering all compartments in which carbon can be stored. However, few studies have been carried out to validate the results obtained with this model in southern Europe. The aim of this doctoral thesis is to discuss modelling and estimation of C accumulation in forest systems, particularly forest plantations. The study focused on the three levels at which carbon can be estimated: tree, stand and landscape level. The study involved evaluation of a specific dynamic process, i.e. the afforestation of former pasture land in northern Spain. The main research questions addressed were thus related to this issue, with the aim of improving the tools and models used to quantify carbon stocks and carbon changes resulting from land use change. The study was carried in the temperate forest of Atlantic southern Europe, one of the most productive timber production areas in Europe. The experimental design and methods used to collect data are summarised in the following points, corresponding to different chapters of the thesis.  A network of 120 paired plots (former pasture land-new plantations of different ages) was established to construct three well-replicated chronosequences of the most common tree species in Atlantic temperate areas. The paired plots represented the original land use (pasture) and forest plantations established on this land. In each of the chronosequences, carbon density was estimated in aboveground biomass, litter and mineral soil to a depth of 30 cm. Changes in carbon in mineral soil and litter were evaluated by non parametric analysis, whereas in living biomass, carbon was calculated by regression analysis. The effect of rotation age on C stock for the average trends in the measured plots was also assessed.  Forty specimens of E. nitens were felled and completely fresh weighed. The tree biomass was divided into the following components: wood, bark, thick branches, thin branches, twigs, leaves and dead branches along the stem. Intensive sampling was carried out along INDEX AND ABSTRACT vii the stem to evaluate the effect of sampling intensity on dry mass estimation, by ratio type estimators. Two different methods were compared: complete fresh weight and dry mass estimation using disks sampled along the stem (CW), and complete stem cubication and partial fresh weight for dry mass and average basic density estimation (PW). Moreover, the usefulness of crown variables for describing biomass crown biomass fractions was assessed. The utility of the models developed for estimating changes in biomass fractions proportion in aboveground biomass was assessed for different dimensional classes.  Measurements made in of 15 additional plots for E. globulus and 36 for E. nitens established on forest land were used to develop static growth models for these species. The threshold density limit derived from self-thinning and harvesting were considered for model development, and enabled evaluation of technical and silvicultural limitations for the single stem rotation. The models enabled evaluation of the effect of two different initial stocking on bioenergy production in several different ways, as well as carbon accumulation for the two eucalypt species studied.  Several soil samples were collected in the E. globulus and P. radiata chronosequence plots, to represent the average trends for these species, and were analysed by calorimetry and thermal analysis, as novel techniques to elucidate how afforestation affects the nature of soil organic matter (SOM) and soil microbial metabolism. The results were compared with those obtained by solid state nuclear magnetic resonance, often used to study SOM. The application of both techniques enabled estimation of the nature of changes in SOC following land use change, and highlighted the usefulness of calorimetry and thermal analysis for distinguishing soil organic compounds.  The CO2Fix mechanistic model was parameterized with data obtained in the abovementioned plots, and with additional measurements regarding forest industry sector products fluxes and performance. Detailed measurements of all compartments estimated by the CO2Fix model in the 120 paired plots (see above) enabled validation of the CO2Fix model estimations for the conditions under study. Detailed wood industrial sector data were obtain for parameterizing the wood products module. This, together with previously developed yield models and hypotheses regarding the lifespan of product, was used to evaluate the effects of two different management alternatives for the species considered, in relation to climate change mitigation. Regarding the changes in C after land use change from pasture to forest plantations (Chapter II), the mean rates of C sequestration (biomass and soil) estimated throughout the rotation ranged between 8.7 and 14.6 Mg C ha-1 yr-1 (Eucalyptus nitens > Eucalyptus globulus > Pinus radiata), and the contribution of the soil (litter plus mineral soil) ranged from 8 to 18% (Eucalyptus nitens > Pinus radiata > Eucalyptus globulus). The humid temperate climate and the sandy loam texture of the soils favoured large losses of SOC from the uppermost mineral soils during the 10 years following afforestation. The losses were derived from the large content of SOM in the pasture soils in the region. The higher losses of SOC from the pine soil (26% of initial SOC compared with loss CHAPTER 0 viii of 19.5% of initial the SOC from soil under eucalypts) were attributed to the lower transfer of organic C to the mineral soil, as a result of the lower litter decomposition rate and the lower belowground litter input from associated vegetation. The latter factor was attributed to differences in shading provided by the different species, which finally leads to different microclimatic conditions in the plantations. The climate also favoured the rapid development of tree biomass and subsequent C sequestration in biomass and soils. In the study of the effects controlling the SOC accumulation, soil carbon losses from the upper mineral soil layers in the plots with the highest site indexes were lower and recovery of the initial contents took place faster than in the other plots. The initial carbon content during the previous land use was a significant factor determining the changes in SOC following land use change. The initial carbon losses and later recovery of initial carbon contents were highest in plots in which the initial soil carbon content was highest. This effect was significant for both species of Eucalyptus, but not for P. radiata, possibly because of the death of herbaceous vegetation in Pinus plantations in the first years after land use change. The results of non-parametric analysis of average trends enabled evaluation of the effect of different rotation ages on C stock and accumulation rate for the three species studied and the compartments considered. The C sink capacity of forest plantations can be maximized by lengthening the rotation and adopting suitable management strategies for each species. This is especially important in plantations in which the high intensity of harvesting may prevent accumulation of SOC in the long term. In Chapter III, the intensive sampling considered for the stem component enabled estimation of the effect of sampling intensity and initial sampling point in the stem on the accuracy and bias in estimation of stem wood dry mass by ratio type estimators and systematic sampling. For both methodologies considered (CW and PW), the residues were plotted against sampling intensity, and thresholds were established to keep the relative error below 10%. The increases in moisture content and basic density along the stem explained the serious risk of dry mass overestimation when systematic subsamples were considered. Of the methods considered here, the CW approach produced better results for the largest dimensional class than the PW method for a similar sampling intensity. This is important because the PW method is usually used for large trees in which complete weighing is time-consuming. The results clearly show the trends in relative errors derived from measurement of the bottom disk or the bottom log, considered by default as the first section that should be measured. This resulted in overestimations in the CW method and underestimations in the PW method. If systematic sampling is used, it is advisable to establish the sampling intensity before randomizing the position along the stem of the first disk or log to be measured. In the case of the PW method, it is not recommended to take only one sample log per tree, and the sample should be split along stem in an attempt to represent the average basic density, which usually occurs at a relative height of 30-35% along the stem. INDEX AND ABSTRACT ix The biomass equations were fitted by seemingly unrelated regression, with corrections for heteroscedasticity carried out by weighted fitting. Diameter at breast height was the best explanatory variable, and the inclusion of height did not improve the accuracy of estimation, except for the wood component. The inclusion of crown variables improved the predictive ability for crown fractions, increasing the accuracy of estimating thick branches (by 10.8%), twigs (by 19.1%) and leaves (by 17.3%). The biomass of each fraction decreased in the following order: wood > bark > thick branches > dead branches along the stem > leaves > thin branches > twigs. The changes in these percentages with diameter classes and the predictive ability of the fitted equations were also studied. Development of stand level models for estimating biomass yield, total energy and carbon sequestration in Eucalyptus globulus and Eucalyptus nitens plantations is described in Chapter IV for first rotation stands established at the usual range of initial forest densities in south-western Europe. The timber volume, total aboveground biomass, logging residue biomass, crown biomass, carbon in aboveground biomass and soil organic layer, energy in aboveground biomass, energy in logging residue biomass and usable cellulose yield were represented in the form of isolines (taking mortality into account) and plotted against dominant height. These variables were calculated and compared with previously published data on two silvicultural options for short rotation forestry, one destined for bioenergy production and the other consisting of the standard silviculture regime applied to both species in southern Europe, considering the average site index for each. Yield levels were higher in Eucalyptus nitens than in Eucalyptus globulus for all variables, because of faster diameter increment at similar densities. The total yield in terms of biomass was 13.9-14.6 Mg ha-1 yr-1 for Eucalyptus globulus and 20.4-21.5 Mg ha-1 yr-1 for Eucalyptus nitens. The energy in aboveground biomass ranged between 233-245 GJ ha-1 yr-1 for Eucalyptus globulus and 345-364 GJ ha-1 yr-1 for Eucalyptus nitens; the carbon accumulation rate in aboveground biomass and soil organic layer was 6.9-7.2 Mg ha-1 yr-1 for Eucalyptus globulus and 12.7-13.5 Mg ha-1 yr-1 for Eucalyptus nitens, and usable cellulose was 5.7-5.9 Mg ha-1 yr-1 for Eucalyptus globulus and 9.0-10.1 Mg ha-1 yr-1 for Eucalyptus nitens. Simulated rotations were longer than the average lengths considered for short rotation woody crops. This has some positive effects, such as higher wood:bark ratios and larger average tree sizes. It was found that 50% increments in the initial density result in only marginal increments in biomass and usable cellulose yields. The density management diagrams constructed can be used as a powerful tool for management of eucalypt plantations with a multipurpose objective (pulp or solid timber production, fossil fuel substitution or carbon sequestration). The application of calorimetry and thermal analysis as novel techniques to elucidate how afforestation affects the nature of SOM and soil microbial metabolism is described in Chapter V. The techniques were applied to study the SOM dynamics in two stands afforested with Pinus radiata and Eucalyptus globulus, established on pastures in a humid temperate region. Results of thermal analysis and calorimetry were compared with those obtained by solid state nuclear magnetic resonance, which is often used to examine changes in chemical SOC stability. CHAPTER 0 x The application of differential scanning calorimetry and solid state nuclear magnetic resonance revealed that the SOM comprised carbohydrates, carbonyl/carboxyl groups, aliphatic components and aromatic carbon in the first years of the rotation. All these fractions became degraded after afforestation. The degradation was monitored by calorespirometry, which provided the calorespirometric ratio of the soil basal metabolism, together with the active biomass and the metabolic quotient. These indexes proved to be sensitive parameters that provided information about changes in the pattern of microbial metabolism in response to changes in the nature and redox state of the carbon substrates, demonstrating degradation of the aromatic and aliphatic organic matter fraction. The techniques were able to distinguish differences in SOM dynamics in the two types of stands, attributable to the different development of understory vegetation and litter composition. A comprehensive study was carried out considering the carbon sink effect in biomass, soil and wood products, the substitutive effect of bioenergy, and particular conditions of the forest industry in southern Europe, as described in Chapter VI. The CO2Fix model was parameterized by using the yield models, leaf and branch turnover ratios derived from the plots established and regionspecific parameters for products and biomass. Validation was carried out for the biomass and the soil module with plot-specific climate data obtained from a network of 120 plots. Results showed strong bias in SOC estimation, which was attributed to overestimation of the decomposition rates in soil compartments. Slight bias in the carbon biomass estimation was also observed when yield models specific for forest land were used to simulate afforestation on former pasture land. As regards the sensitivity, the Yasso model was found to be strongly robust to leaf, root and branch turnover. The woodchip production alternative yielded higher carbon stock in biomass and products, as well as in bioenergy substitution effect, than the sawn-wood production alternative. Nevertheless, the sawn-wood alternative was the most effective as regards the carbon stock in the soil. Site index had an important effect for all species, alternatives and compartments, and the mitigating effects on climate change increased with site index. Harvesting clearcutting and thinning slash for bioenergy use led to a slight decrease in the soil carbon equilibrium, but significantly increased the mitigating effect through bionergy use. INDEX AND ABSTRACT xi 0.3. Resumen El cambio climático es hoy en día uno de los problemas ambientales más preocupantes a nivel mundial. Es causado por el denominado efecto invernadero, debido en último término a la liberación de cantidades ingentes de gases de efecto invernadero, particularmente CO2. El sector energético es el que actualmente más contribuye al efecto invernadero, aunque el cambio de uso del suelo debido a la deforestación también tiene una elevada importancia. A pesar de las tendencias observadas en las zonas tropicales, donde grandes superficies de bosque nativo continúan siendo deforestadas, los cambios de uso del suelo en Europa y America del norte van en la dirección contraria. Este efecto ha sido acelerado en parte debido a la puesta en marcha de la Directiva EU 2080/92 de reforestación de tierras agrarias, la cual derivó en una gran superficie de terrenos reforestados en el norte de España entre 1992 y 2006. Estas reforestaciones fueron realizadas mayoritariamente con especies de crecimiento rápido, fundamentalmente Eucalyptus globulus Labill., Eucalyptus nitens (Deane & Maiden) Maiden y Pinus radiata (D. Don), y establecidas en su mayoría sobre praderas. Los acuerdos internacionales sobre reducción de emisiones de carbono permiten a los países firmantes la compensación de emisiones de CO2 mediante acumulación en sumideros reconocidos como tal. Sin embargo, hay una elevada incertidumbre asociada a la estimación de la cantidad de carbono secuestrada en los sistemas forestales, principalmente en algunos compartimentos concretos como por ejemplo el suelo mineral. La incertidumbre es aún mayor en sistemas que, tras un cambio de uso, aún no han alcanzado el equilibrio, por lo que el estudio de la evolución del carbono mediante la metodología habitual, suponiendo cambios de stock en el estado de equilibrio, podría acarrear importantes errores en la estimación. Aunque la estimación del carbono en la biomasa arbórea es relativamente sencilla, la estimación en otros compartimentos, como por ejemplo el suelo, es mucho más complicada y tiene asociada una mayor incertidumbre. Este es el motivo de que la mayoría de los estudios de la dinámica del carbono en el suelo están basados en modelos. Sin embargo, la evaluación empírica es necesaria para estudios más consistentes, y proporciona información valiosa para validación de modelos ecofisiológicos así como para el desarrollo de modelos empíricos. Sin embargo, la medida directa del carbono en el suelo es laboriosa y los resultados son altamente variables, provocando que a menudo no sea posible obtener resultados concluyentes debido a la alta variabilidad. Además, los cambios en el carbono en el suelo son difíciles de estimar debido a que los procesos son relativamente lentos, y a que hay otras variables externas que afectan a su evolución. En este sentido, el diseño experimental mediante cronosecuencias combinadas con parcelas pareadas puede ser útil para convertir diferentas espaciales en diferencias temporales, al mismo tiempo que permiten corregir tendencias locales debido a la alta variabilidad espacial de la variable de interés, permitiendo evaluar la evolución del carbono a escala comarcal. CHAPTER 0 xii El carbono en la biomasa arbórea suele estimarse mediante ecuaciones de biomasa, las cuales se les supone una mayor precisión que los factores de expansión de biomasa. Sin embargo, la estimación de la masa seca de los pies tipo que se emplearán en el desarrollo de las ecuaciones de biomasa es posible que presente algún problema mediante las metodologías en uso. Además, el interés creciente en aprovechar fracciones de biomasa de copa para usos energéticos, así como la necesidad de su estimación para estudios nutricionales y ecológicos, lleva a la necesidad de disponer de modelos precisos para la estimación de las fracciones de copa. La inclusión de variables de copa como explicativas en las ecuaciones de biomasa puede llevar al desarrollo de modelos más precisos que los basados únicamente en variables de fuste. La producción energética de una plantación está habitualmente expresada en términos de volumen o peso seco de madera, mientras que el efecto de substitución de carbono debido al uso bioenergético de los productos forestales suele estar expresado en términos de energía, o toneladas equivalentes de petróleo. La cantidad de energía obtenida por cada unidad de biomasa depende del proceso de transformación, lo que al final lleva a que la estimación de la cantidad de energía se haga en función del volumen en pie o de la biomasa y la aplicación de los factores correspondientes. La existencia de modelos específicos para la estimación de la producción energética en función del estado de desarrollo de la masa puede evitar la concatenación de errores en la estimación, y al mismo tiempo permitir la comparación directa de la producción energética de varias alternativas silvícolas o de diferentes especies. El primer paso en el estudio la evolución del carbono en el suelo tras un cambio de uso es la evaluación del cambio de stock. Sin embargo, la estabilidad de los compuestos orgánicos del suelo mineral varía tras el cambio de uso. Debido a que las técnicas tradicionales para su estimación son laboriosas y caras, la puesta a punto de técnicas que permitan trabajar con un gran volumen de muestras, y a un coste razonable, será de gran ayuda para estudios de modelización de la estabilidad de la materia orgánica del suelo. Ello cobra mayor importancia cuando se requiere una elevada resolución temporal o cuando se pretende hacer estimaciones a gran escala, para lo cual es necesario analizar una elevada cantidad de muestras. Cuando se evalúa el efecto de mitigación conjunto del sector forestal, se han de considerar todos los compartimentos en los que el carbono se puede encontrar retenido. Ello incluyen los productos que se generan el los bosques debido a las cortas de madera u otros productos, aunque no está claro aún como estos van a ser tenidos en cuenta en los acuerdos internacionales de reducción de emisiones. Las técnicas más habituales en estos estudios son los modelos mecanísticos, como por ejemplo el CO2Fix, el cual permite la estimación del efecto de mitigación de varias alternativas silvícolas considerando todos los compartimentos en los que el carbono se puede encontrar retenido en los sistemas forestales. Sin embargo, son escasos los estudios en los que se han evaluado los resultados obtenidos con este tipo de modelos en las condiciones del sur de Europa. El principal objetivo de esta Tesis Doctoral es la disertación acerca de la modelización y la estimación de la acumulación de carbono en sistemas forestales. Para ello se han considerado INDEX AND ABSTRACT xiii todos los niveles de estimación: árbol individual, rodal y paisaje. El estudio ha centrado en el estudio de un proceso dinámico, como es la aforestación de pastizales en el norte de España. En los distintos capítulos de la presente Tesis se han desarrollado y mejorado herramientas para la estimación de carbono en sistemas forestales. El estudio se ha centrado en plantaciones forestales establecidas sobre antiguos pastizales sobre clima atlántico del sur de Europa, uno de las zonas más productivas de Europa. El diseño experimental y los métodos empleados en la recolección de datos se resumen en los siguientes puntos, correspondientes a distintos capítulos de la Tesis.  Se estableció una red de 120 parcelas pareadas (cada parcela consta de una antigua pradera y plantación forestal establecida sobre la anterior, de diferentes edades), establecidas a modo de cronosecuencia para las tres especies mas habituales en aforestación de terrenos agrícolas en el norte de España. Las parcelas pareadas representaron el uso que originariamente cubría la totalidad de la parcela (pradera), y la plantación forestal establecida en parte de la parcela sobre el anterior uso. En cada cronosecuencia, la densidad de carbono por hectárea fue evaluada en biomasa arbórea, mantillo y suelo mineral hasta 30 cm de profundidad. La evolución del carbono en el suelo mineral y en el mantillo debida a la edad transcurrida desde la aforestación fue evaluada mediante análisis no paramétrico, mientras que el carbono en la biomasa aérea fue evaluado mediante análisis de regresión. El efecto del turno de corta en el stock de carbono para las evolución media de las parcelas medidas también fue evaluado.  Se apearon un total de 40 pies de E. nitens, los cuales fueron sometidos a análisis destructivo para evaluar su peso seco. Los pies fueron divididos en los siguientes componentes: madera, corteza, ramas gruesas, ramas finas, ramillos, hojas y ramas secas en el fuste. Para evaluar la masa seca de madera se empleó una intensidad de muestreo muy elevada, que permitió evaluar el efecto de ésta en la estimación de la masa seca de madera del árbol tipo cuando se aplica la metodología de ratio type estimators. Se evaluaron dos metodologías diferentes: pesado completo del árbol en verde y estimación de la masa seca mediante discos tomados sistemáticamente a lo largo del fuste para determinar la humedad (CW), y cubicado completo del fuste y estimación de la masa seca mediante la estimación de la densidad básica (PW). Además, se probó el ajuste de algunas variables de copa como predictivas en las ecuaciones de biomasa. La capacidad de las ecuaciones desarrolladas para la estimación de la proporción de las distintas fracciones de biomasa fue evaluada para distintas clases dimensionales.  Se desarrollaron modelos estáticos de crecimiento para las dos especies de eucalipto estudiadas, para lo que se midieron 15 parcelas adicionales para E. globulus y 36 para E. nitens, todas ellas establecidas sobre suelo forestal. Para el desarrollo de los modelos, se tuvieron en cuenta los límites máximos de densidad debido al autoclareo así como otras limitaciones dimensionales derivadas de la posibilidad de cosecha mediante maquinaria convencional. Los modelos permitieron evaluar el efecto de dos densidades iniciales en la CHAPTER 0 xiv producción energética evaluada mediante varios indicadores, así como la acumulación de carbono en distintos compartimentos para las dos especies de eucalipto consideradas.  Varias muestras consideradas como representativas de la evolución media de las parcelas de E. globulus y P. radiata fueron analizadas mediante calorimetría y análisis térmico, como técnicas recientes para la evaluación del efecto de la aforestación de terrenos agrícolas abandonados sobre la naturaleza de la materia orgánica del suelo y del metabolismo microbiano. Los resultados fueron comparados con los obtenidos mediante una técnica bien contrastada para la evaluación de la naturaleza de la materia orgánica, la resonancia magnética nuclear en estado sólido. La aplicación de las dos técnicas permitió la estimación de la naturaleza de los cambios en la composición de la materia orgánica de los suelos tras el cambio de uso de pastizal a plantación forestal, y puso de manifiesto la capacidad de la calorimetría y el análisis térmico para diferenciar grupos de compuestos de la materia orgánica del suelo.  El modelo mecanístico CO2Fix fue parametrizado con los datos obtenidos en los apartados anteriores y con datos adicionales recopilados del sector forestal en el norte de España respecto al destino industrial de los productos por especie y rendimiento en la transformación. La parametrización del modelo CO2Fix fue validada para las condiciones descritas en el presente estudio con los resultados obtenidos en las 120 parcelas de los apartados anteriores, donde el carbono en biomasa, mantillo y suelo mineral fue evaluado mediante medición directa. Una vez parametrizado el modelo, se evaluó el efecto de mitigación de carbono en distintos compartimentos para dos alternativas selvícolas y cada una de las especies estudiadas. Respecto a la evolución del carbono tras el cambio de uso de pradera a plantación forestal (Capítulo II), la acumulación media de carbono en biomasa y suelos para el total de las parcelas de cada especie osciló entre 8.7-14.6 Mg C ha-1 año-1 (Eucalyptus nitens > Eucalyptus globulus > Pinus radiata), y la contribución del suelo (mantillo + suelo mineral) sobre el total osciló entre 8-18% (Eucalyptus nitens > Pinus radiata> Eucalyptus globulus). El clima templado-húmedo y la textura del suelo franco-arenosa favorecieron las elevadas pérdidas de carbono en el suelo en los horizontes superficiales durante los primeros 10 años tras el cambio de uso. Las pérdidas fueron debidas en parte a los elevados contenidos de materia orgánica en el uso del suelo previo. Las elevadas pérdidas de carbono observadas en las parcelas de P. radiata (26% del contenido inicial de carbono, frente al 19,5% de pérdida en las parcelas de ambos eucaliptos) fueron atribuidas a la menor tasa de transferencia de materia orgánica fresca al suelo mineral, debido a una menor tasa de descomposición del mantillo y a una menor incorporación de litter procedente de las raíces de las especies herbáceas. Este último factor fue debido a diferencias en la sombra proyectada por ambos grupos de especies, lo que al final derivó en diferencias microclimáticas bajo cubierta. Las condiciones climáticas favorecieron el rápido crecimiento de la biomasa arbórea y consecuentemente del aporte de matera orgánica al suelo mineral. GENERAL INTRODUCTION AND OBJECTIVES 3 1. General introduction and objectives 1.1. Climate change and the forest sector 1.1.1. Climate change, a global problem Almost the entire scientific community now accepts the evidence that the climate is changing. This effect has been reported in relation to the trend in annual temperatures in the last 30 years (Trenberth & Josey, 2007), and on a longer timescale in the last century (Jones & Moberg, 2003), and in the last 1000 years (Jones et al., 2001). Although there have been examples of natural climate oscillations throughout the Earth’s history, the current effect is attributed to anthropogenic activity (IPPC, 2001; Grace, 2004). The two main reasons for climate change are increasing emissions of greenhouse gases (GHGs) and aerosols, and land use change, both of which cause changes in atmospheric temperature (Houghton et al., 1990) as a result of the so called greenhouse effect (GE). The main GHGs are water vapour, CO2, CH4 and N2O. Emissions of GHGs used to be expressed in terms of CO2-equivalent because CO2 alone contributes more than 30% to the total greenhouse effect in the atmosphere, 56% of the anthropogenic radioactive forcing (Hansen & Lacis, 1990), more than 82% to total emissions at European level (IPPC, 2007), and also because it can be captured through biological processes. Plants, as autotrophic organisms, transform inorganic compounds into organic tissues, consuming CO2 in the process. Releases via respiration and sequestration by biomass growth were in equilibrium until the beginning of the industrial age, when huge amounts of the system carbon (C) was released as a result of increased burning of fossil fuels carried out in response to increased energy demands (Foley et al., 2005). Some scientists think that climate change is irreversible (Solomon et al., 2009). Nonetheless, even greater increases in the average annual air temperature will occur unless mitigation policies are adopted (Meehl et al., 2007). The development of adaptation strategies and evaluation of the predictable impacts on each sector as regards the new situation are two of the most active research lines at the moment (Campioli et al., 2009). There are several ways of reducing CO2 emissions by the amounts required by international agreements, but these can be summarized in two sets of actions: increasing the C sink, and reducing C emissions. Worldwide, GHGs emissions are mainly related to energy use (63%), predominated by electricity power plants, industry and transport (Fig. 1.1). Reduction in GHGs emissions therefore CHAPTER I 4 requires a change in energy supply to alternative renewable energy sources. Although there are various CO2-free alternatives for power and industry energy supplies (i.e. nuclear, wind, hydraulic), as regards transport, the use of biofuels has received extensive attention because at present there are no other CO2-free alternatives available with the technology in use (Fulton et al., 2004). Some authors indicate that for a significant mitigation of fossil fuel emissions, very large areas of cropland would be required to produce sufficient amounts of biofuel for transport needs (Righelato & Spracklen, 2007), which may lead to displacement of agricultural production and cause additional land-use change, finally leading to net increases in GHGs emissions (Searchinger et al., 2008; Melillo et al., 2009). In this sense, only conversion of woody biomass may be compatible with retention of forest carbon stocks (Kirschbaum, 2003; Tilman et al., 2006; Righelato & Spracklen, 2007). In Spain, 80% of GHGs emissions are derived from the energy sector, and renewable energies have been considered as an important alternative for reducing emissions, with biomass having a neutral emission effect. The 2005-2010 Renewable Energy Plan (PER) in Spain aimed at a 12% contribution of renewable resources for primary energy consumption by 2010. However, the current situation clearly shows that biomass is the least well developed among the renewable energies in Spain. The share of forest biomass is difficult to calculate, although it is clear that recent initiatives for new power plants are mainly linked to the use of forest and agricultural industry residues, and that forest energy crops (short rotation coppice) would be important within the group of energy crops. One of the key elements for the previous points is Royal Decree RD 661/2007, which establishes the grant system for alternative energy production. The Decree allocates a large portion of the incentives to power plants that utilise biomass from energy crops. Industry Energy supply Industry Waste and wastewater Buildings Deforestation Agriculture Global emissions by sector NON-ENERGY EMISSIONS Industry Energy supply Industry Waste and wastewater Buildings Deforestation Agriculture Global emissions by sector Industry Energy supply Industry Waste and wastewater Buildings Deforestation Agriculture Global emissions by sector NON-ENERGY EMISSIONS Figure 1.1. Global greenhouse gases emissions by economic sector. Source: IPPC (2007). Increase of terrestrial sinks also lead to decreases in the concentration of atmospheric CO2, although the mitigating effect of this option is limited. Non energy emissions account for up to 37% of global emissions (Fig. 1.1), and a change in land use to others with more favourable C balance may partly mitigate the greenhouse effect. Moreover, alternative management schemes for current GENERAL INTRODUCTION AND OBJECTIVES 5 land uses may provide different C balances, and must therefore be investigated (Malhi et al., 2002). 1.1.2. International agreements regarding reduction of GHG emissions All administrations and policymakers now acknowledge climate change as one of the most important current environmental problems, and several attempts have been made to obtain international agreement for reducing CO2 emissions. Inclusion of carbon sinks in the global carbon budget was discussed at the United Nations Framework Convention on Climate Change (UNFCCC) celebrated in Rio de Janeiro (1992). However, it was at the Kyoto meeting (UNFCCC, 1997) that the UNFCCC limited GHGs emissions, taking as a baseline for accounting the emissions levels in 1990; it was also agreed that emissions could be compensated through sinks. This document recognizes two main alternatives for emissions compensation in the Land Use, Land Use Change and Forestry (LULUCF) sector for industrialized countries, depending on whether the actions took place within their own territory (Art. 3.3, 3.4 and 6) or in other nonindustrialized countries (Art. 12). For actions within their own border, the Kyoto Protocol (KP) allows countries included in Annex I the following activities to meet the requirements of the KP:  Article 3.3, to use “net changes in greenhouse gas emissions by sources and removals by sinks resulting from direct human-induced land-use change and forestry activities, limited to afforestation, reforestation and deforestation (ARD) since 1990” (UNFCCC, 1997).  Article 3.4: “forest management (FM), cropland management (CP), grazing land management (GM) and revegetation (RV) are eligible land-use, land-use change and forestry activities” (UNFCCC, 2001).  Article 6: “transfer to, or acquire from, any other such party reduction units resulting from projects aimed at reducing anthropogenic emissions by sources or enhancing anthropogenic removals by sinks of greenhouse gases in any sector of the economy, provided that (…) any such project provides a reduction in emissions by sources, or an enhancement of removals by sinks, that is additional to any that would otherwise occur” (UNFCCC, 1997), which is known as joint implementation (JI).  Article 12, Annex I countries are allowed to acquire from non-Annex I countries certified Emissions Reduction Units (ERUs), through the so called Clean Development Mechanism (CDM), which provides that C removal projects are “additional to any that would occur in the absence of the certified project activity” (UNFCCC, 1997), and is limited to afforestation and reforestation activities for the first commitment period (UNFCCC, 2001). JI and CDM are required to consider a baseline for the certified emission reductions, in other words the reference C level with respect to gains or losses must be estimated, and for agricultural activities (CM, GM, and RV), net-net accounting is required, which consists of considering “net emissions or removals over the commitment period less net removals in the base year, times five”. CHAPTER I 6 In subsequent meetings (UNFCCC, 2002; UNFCCC, 2003) the instructions for GHGs inventories in the LULUCF sector were defined. This arrangement defines the methodology, the default values and the pools that must be considered in forest GHGs inventories, as follows: live above and belowground biomass, litter, dead trees and soil organic matter. Carbon in forest products cannot therefore be considered in carbon accounting calculations (UNFCCC, 2002). The results of the last meetings of the United Nations Climate Change Conferences (COP) celebrated in Copenhagen (COP15, 2009) and Cancun (COP16, 2010) are ambitious in regard to GHGs reduction and deferred the force period of KP. However, the current economic crisis, the generalized dependency of fossil fuel energy, and the permissiveness with developing countries with high emissions rates, makes compliance of the international agreements on emission reductions difficult. GHGs emissions increased by a record amount in 2010, to the highest carbon output in history (30.6 Tg), according to the International Energy Agency, putting hopes of holding global warming at safe levels all but out of reach. Total emissions increased between 1990 and 2008 in Spain by on average 47.8%; inclusion of LULUCF supposes a reduction of 15.1 points as regards the latter figure (UNFCCC, 2010). For the period 2008-2012, Spain is committed to increasing emissions by 15% relative to the base year (1990), which is included in the emission allowance trading scheme within the EU, i.e. global reductions in emissions of 8%. Nonetheless, the previsions of the government (II Assignation Plan) consider an increase of 37% the most likely estimation. The Spanish Climate Change and Clean Energy Strategy (EECCEL) considers that the increasing trend of GHG emissions in the 1990-2008 period corresponded to rapid, sustained economic growth, and to an increase in population in recent years. The effort made by Spain in matters of Energy Saving and Efficiency was insufficient, but the report also shows that the per capita emissions reached an average of approximately EU-15. The EECCEL states that the target established by the Government for the five-year period 2008-2012 is that Spain’s totals do not surpass a 37% increase relative to the emissions in the base year. This represents a difference of 22 percentage points with respect to +15%, 2% of which must be obtained by means of sinks, and the remainder (20%) by means of flexible mechanisms (acquisition of carbon credits). In order to reach the said objective of +37%, the National Allocation Plan (NAP) 2008-2012 requires additional measures to obtain reductions of 27.1 Mt of CO2 eq. A Plan of Urgent Measures considers reductions of 12.091 Mt CO2 eq yr-1, so that additional measures are still necessary to provide reductions of 15.033 Mt CO2 eq yr-1. 1.1.3. The role of forests and the forest-based sector in climate change mitigation As previously reported, the main research lines regarding climate change and the forest sector are impacts, mitigation and adaptation (Campioli et al., 2009). Mitigation is the action taken to reduce the atmospheric concentration of GHGs in order to prevent dangerous climate change (IPPC, 2007). However, as GHGs mitigation alone is not sufficient to prevent climate change, we GENERAL INTRODUCTION AND OBJECTIVES 7 must be prepared for the possible impacts of increased temperatures and changes in precipitation regimes on forests. Several impacts are currently under study; most of these are related to change in forest productivity, abiotic and biotic disturbances, and species migration and extinction (Thuiller, 2003; Thomas et al., 2004; Battisti et al., 2005; Blennow & Olofsson, 2008; Lindner et al., 2010; McMahon et al., 2010). The susceptibility of forests to these impacts depends on the inherent adaptive capacity of trees and forest ecosystems (Hamrick, 2004; Thuiller et al., 2005) as well as on the intensity and direction of climatic variables. Adaptation is adjustment in natural or human systems in response to actual or expected climatic stimuli or their effects (IPPC, 2007). There are several ways in which forests and the forest-based sector can contribute to mitigating the greenhouse effect. In simplified terms, the forest-based sector can be seen to be formed by the forest ecosystem, forest products, and energy from the forests. Canadell & Raupach (2008) indicated four forest management strategies for mitigating GHGs emissions: (i) increasing the forest area through reforestation, (ii) increasing the C density per area, (iii) increasing forest product use and fossil fuel substitution through bioenergy, and (iv) reducing deforestation and degradation. Together these proposals can be summarized as increasing C stock in all compartments in which forests can contribute (biomass, soil, products), and generating a stream of avoided emissions through bioenergy use. Decreasing the forest area is one of the most important factors as regard release of CO2, N2O and CH4 to the atmosphere (Shukla et al., 1990; Malhi et al., 2008). This is still occurring in the tropics (Canadell & Raupach, 2008), whereas the forest area is generally increasing in temperate and boreal regions (FAO, 2005). One of the most evident ways of mitigating climate change through the forest sector is therefore to increase forest area or maintain the area by slowing down deforestation, although this is difficult because forest land in developing countries is being transformed to agriculture land to meet human feed requirements. In this context, the area covered by forest plantations is expected to increase by 16-32% by 2030 relative to the level in 2005 (estimated at 261 M ha, Carle & Holmgren, 2009). According to these authors, the highest absolute increase will take place in Asia, whereas the highest relative increase will occur in Southern Europe. Other ways of increasing the C stock are to optimize management schemes regarding C sequestration (Liski et al., 2001; Liski et al., 2002; Kaipainen et al., 2004; Canadell & Raupach, 2008). In traditional forestry, management schemes consider wood supply or economic criteria, and other types of management schemes that favour better C balance may not be compatible with traditional management practices (Lindner & Karjalainen, 2007). Additional research is therefore needed to evaluate the effect of actual and potential management schemes on the different C pools considered in the international agreements (UNFCCC, 2002). Finally, the last way in which the forest sector can mitigate climate change is known as the substitution effect. This involves the so-called avoided emissions, derived from consumption of harvested wood products (HWP), rather than other products that require consumption of fossil CHAPTER I 8 fuels during their fabrication. HWP are not considered in global C budgets unless a country can show that its long term existing stocks are increasing (UNFCCC, 2002), although the 26th session of the Subsidiary Body for Scientific and Technological Advice (UNFCCC/SBSTA/2007/4, paragraphs 59-61) had invited parties in a position to do so, to report voluntarily on wood products in their national inventories. This is because the strong effect of HWP on the global carbon cycle due to the avoided emissions from sources that have stored carbon for thousands of years. Therefore, even when the emissions from HWP are immediate (in the case of bioenergy), or when the wood product would have a longer lifespan (furniture, paper, etc.), the avoided emissions to obtain the same amount of energy or to manufacture the alternative product are derived from a pool that has been stored for thousands of years. Keeping forests at an intense stage of growth may help maximize the supply of forest products, and therefore maximize the mitigating effect, although this does not mean that changes from mature virgin forests to plantations should be envisaged. Some studies have shown the ability of old growth forest to capture C (Carey et al., 2001; Pregitzer & Euskirchen, 2004; Zhou et al., 2006; Luyssaert et al., 2008), and other studies have reported that change from old-growth forest to young fast growing forest does not result in a positive net C balance, even if wood products are considered (Harmon et al., 1990). Holding forests at fast growing stages implies intense management, which may finally lead to losses of soil organic carbon (SOC), although the balance is usually towards net sink effects. The fight against climate change may produce restrictions, but also opportunities: mitigation can reduce our external dependence on fossil fuels and alleviate environmental problems, such as urban contamination. Land use planning can also be improved and clean transport systems promoted. 1.2. The forestry sector in northern Spain and climate change mitigation 1.2.1. Forestry, land-use change and carbon sequestration in Spain The rural landscape underwent a change during the last century, motivated by the change in economic activity. At the beginning of the 20th century, most of the primary economic sector was based on agriculture and livestock. The normal development and modernization of the primary sector were held up by the civil war and the posterior recovery period. From the mid-1950s, depopulation of rural areas led to huge abandonment of land (Lasanta, 1996; Lavorel et al., 1998); agriculture production was concentrated in the most productive areas, and less productive land was abandoned or transformed into forest. GENERAL INTRODUCTION AND OBJECTIVES 9 During this time the Spanish Government initiated an extensive afforestation program across the country, which led to in a huge areas of land being afforested. In the period 1940-1973, the total area afforested in Spain was 2306200 ha (Pemán García et al., 2009), and 48% of this corresponded to afforestation of non forest land. The main objectives of this policy were to increase the forest cover and to reduce rural unemployment. The most commonly used species in this extensive afforestation program were coniferous species, because the land available was the least productive land (Pemán García et al., 2009). As regards land property, most of this afforestation involved public land, although some involved collectively owned land. More recently (in 1992), the European Union implemented a policy to afforest former agricultural land (EEC 2080/92), which led to afforestation of approximately 160000 ha of land in northern Spain (MAPA, 2006). Although the main aim of this policy was to reduce agricultural surpluses, the effect on C stock is still under study. Livestock predominates in the agricultural sector in the north of Spain, the most frequent land use change was therefore from pasture to forest land. Pasture land was usually managed in rotation with rape-seed, with increasing importance of the pasture as abandonment became stronger. Finally, almost 3.0 M ha of land were reforested in Spain in the last century, of which at least 1.4 M ha was non forest land, predominantly agricultural land (Pemán García et al., 2009). In 2005 the total area covered by planted forests in Spain was 1.4 M ha (Del Lungo, 2009). Because of the peculiarity of this land use change, research is required to evaluate the effect of this policy on the global C sink. The main objectives of the Spanish forest-based sector are to increase the capacity of CO2 sequestration from the atmosphere by wood stocks, and to comply with the goal of compensating 2% of the base year emissions by LULUCF. The role of Spanish forest in climate change mitigation has been investigated in specific studies (Bravo, 2007; Pardos, 2010), which reported a value of 670 M t CO2 for carbon stored in the aboveground biomass of trees, whereas Gracia et al., (2005) reported a higher value of 2050 M t CO2 for a total carbon in forests, with an annual net increase in sequestration equivalent to 40 M t. The balance between National Forest Inventories IFN2 (19861996) and IFN3 (1997-2006) has been studied in detail in some areas (Bravo, 2007) and the values for increased C (new forests, growth of existing trees, ingrowth) and decreased C (felling, natural mortality) were found to lead to a net increase of 54.6%. This same source of information (IFN) shows that in all regions Spanish forests have acted as C sinks during the 20th century, with a net increase in the sink effect in the period 1990-1998 relative to the period 1974-1987 (Rodríguez-Murillo, 1999). Between the last two inventories, the total amount of C accumulated in the aboveground biomass ranged between 4.5 Mg C ha-1 for Galicia (2.0 in the previous period) and 1.1 Mg C ha-1 in Murcia (0.27 in the previous period). This change is attributed to the reforestations carried out in Spain in the 1940s, and to more recent changes in land use, mainly in agricultural land, which led to an increase in the forest area as well as in carbon density (Rodríguez-Murillo, 1999). However, in these estimations it is assumed that CHAPTER I 10 afforestion on former agricultural land follows the same dynamics as secondary successions or reforestations, although this is still under study. 1.2.2. The tree species studied The contribution of forest plantations to the global wood supply is unquestionable. While forest plantations represented less than 7% of global forest area worldwide in 2005 (FAO, 2005), in the same year they satisfied two-thirds of global industrial roundwood needs (Evans et al., 2009). In the north of Spain, forest plantations also predominate in the forest industry supply, and fast growing tree species are the most commonly used in afforestation (Álvarez, 2004). Pines and eucalypts are the most frequently used species in forest plantations, and in plantations on former agricultural land, the most commonly used tree species in north-western are Eucalyptus globulus (Labill), Eucalyptus nitens (Dean & Maiden) Maiden and Pinus radiata (D.Don). Eucalyptus globulus (Labill) Eucalyptus globulus is native to south-eastern Australia (Victoria) and the south coast of Tasmania, where it grows in the lowest zones of the island (0-550 m); the average precipitation in the zone ranges between 600 and 1500 mm, in a temperate-type clime. This species was first introduced in Europe from Australia in 1863 (Ruiz de la Torre et al., 1979), and nowadays is distributed along the entire coast of the Atlantic Iberian Peninsula. Frost tolerance is the principal constraint for planting at elevations higher than 400-500 m. The range of rainfall in the areas where it has been introduced is very broad, and the average value for the North and Nothwestern Iberian Peninsula is higher than 1500 mm, which explains the good yields. One of the main problems of this species as regards forest growth is disease attack, which causes reduced yield through defoliation at the young stage (Mycosphaerella spp. (Otero et al., 2007)), or at the adult stage (Gonipterus scutellatus (Mansilla Vázquez, 1992)). Although several resistant clones of E. globulus Mycosphaerella are available nowadays, infection of the adult stages by Gonipterus is more difficult to manage. Initially the species was widely used for mining timber and a range of other uses. Nowadays, eucalyptus timber is mainly used to produce bleached pulp, although production of energy crops is also a promising management goal, because of the high proportion of cellulose in the timber (which can lead to high production of bio-ethanol), and its resprouting ability. Other potential uses under study are as high quality sawn-wood (Nutto & Touza Vázquez, 2004), and also structural wood (Guaita & Eiras, 2007). The so-called chip industry covers the use of eucalypt timber for pulp production, which is the main use for this species, because of its favourable specific consumption relative to other species (Cotterill & Macrae, 1997). Management of stands for this purposes starts with plantation densities between 1300 and 1400 stems ha-1, and thinning is not carried out. After the first rotation (13-16 years), several coppice rotations are expected until new plantations are GENERAL INTRODUCTION AND OBJECTIVES 11 established. In each coppice rotation, sprout selection is required to maintain the sprout density constant. Eucalyptus nitens (Dean & Maiden) Maiden Eucalyptus nitens is native to the south-east of Australia, where it is discontinuously present in the states of New South Wales and Victoria. This species grows naturally at altitudes between 600 and 1600 m in the temperate zone, with precipitations of between 750 and 1750 mm per year. Mountain blue gum tolerates absolute minimum temperatures of -10ºC in north-western Spain (González-Río et al., 1997), and of 21-23ºC in the hottest month in its region of origin (Boland et al., 1980). Eucalyptus nitens is one of the most widely used plantation species worldwide in temperate cold regions, and covered more than 340000 ha in 2004 (INFOR, 2004). Expansion of this species in northern Spain began recently; research trials were established around 1950, and commercial plantations began in 1992 in north-western Spain. The reason for the success was the possibility of expansion in areas that were too cold for E. globulus, in addition to the high yield, pest resistance, and the relatively good wood properties of the species. In contrast to the trend observed for the species in Spain, the main destination for E. nitens wood worldwide is the pulp industry (INFOR, 2004). This is partly because of the good mechanical properties of E. nitens pulp (Paz, 1999), but also because of its high yield. Nevertheless, the specific consumption of E. nitens is higher than that of E. globulus (Cotterill & Macrae, 1997), and there are some problems related to pitch, which cause pulp stain. Another interesting property of E. nitens is the calorific power of its waste and wood, which makes it suitable for energetic purposes (Pérez et al., 2008). The most common management scheme for E. nitens in northern Spain is planting at densities of 1300-1400 stem ha-1, with no thinning. No coppice regimes are applied in north-western Spain, because the most common provenance used in reforestation in the region (MacAlister), has little resprouting ability (Sims et al., 1999b; Sims et al., 2001; Little & Gardner, 2003). The possibility of coppicing of other provenances in northern Spain is still under study. There is some interest in producing sawn-wood with this species, and thinning and pruning schemes are therefore under study. Another current important line of research is energetic plantations, although the species does not grow particularly well at high densities (Sims et al., 1999a; Sims et al., 2001). Pinus radiata (D.Don) Pinus radiata is native to a small area on the west coast of North America (Swanton, Monterrey and Cambria in California, and Guadalupe and Cedros in Mexico), but is one of the most commonly used species in afforestation throughout the world, and also the most commonly used exotic conifer for reforestation in the world (Lavery, 1986) and in Spain. The species was introduced in Spain in about 1850, and henceforth its cover increased throughout temperateclimate Spain (Galicia, Asturias, Cantabria and the Basque Country). The success of expansion was due to its fast growth in temperate-climate regions, the wood quality and the plasticity. CHAPTER I 12 The main destination for P. radiata wood is as sawn-wood, but also as woodchip for the chipboard industry. Management schemes for this species in the region depend on the wood destination. For the sawn-wood industry, management consists of planting at 1200-1400 stems ha-1, two thinnings, and a third in cases where high value products are the production objective and site quality is large enough. Products from the first thinning are usually used in the chipboard industry or for bioenergy, and only a few trees (400 stems ha-1) remain standing at the end of the rotation, which in this case is 35-40 years. Two prunings are required to ensure wood quality at the end of the rotation. On the other hand, management for woodchip for use in the chipboard industry starts with higher densities, and the rotations are shorter than for sawn-wood. 1.3. Management tools and methods for carbon inventory and evaluation of mitigation effects in forests There are several ways of monitoring carbon stocks and the mitigation effect in forest ecosystems, but all can be classified in two main groups: (i) methods based on measuring the system C fluxes (FA), and (ii) methods based on measuring changes in C stock (SA) (Ney et al., 2002; Houghton, 2003; Lindner & Karjalainen, 2007; Dias et al., in press). The main differences between the different ways of estimating the stock and the mitigation effects are that C in the FA approach is evaluated from a dynamic point of view (exchange rate), whereas C in the SA methodology is evaluated from the static point of view (stock). With the first approach attention is focused on the specific process under study, more importance is given to the exchange of C in the system or between two pools in the same system (i.e. soil and living biomass, total system and the atmosphere, etc.). In the second approach, more importance is given to the final result of the physiological process, which in the end is the accumulation of C in the system. Since the FA approach is sensitive to short-term changes in the environmental conditions, erroneous conclusions could be reached from this method unless a representative temporal time series is measured. Moreover, the initial value must be given for assessment of the final stock, and the SA methodology must be applied anyway. The SA approach is therefore the most commonly used method for investigating changes in forest systems and for estimating mitigation effect. However, there is still some uncertainty as regards certain aspects of this methodology, as detailed below. Other methods are used almost exclusively for estimating the mitigation effect associated with harvesting and wood products, these are the methodologies based on measuring production (PA) (Brown et al., 1999; Dias et al., in press). Depending on the methodology, all approaches may be classified as direct measurement or estimation through previously developed models. Although direct measurement is the most reliable way of estimating the C stock in forest system, this method is time consuming and expensive when an adequate replication is applied. A modelling approach is therefore usually used to determine the GENERAL INTRODUCTION AND OBJECTIVES 19 the stem to determine the dry mass from the average moisture content and the total fresh weight, or from the stem volume and the average basic density. Studies in which biomass equations are developed at tree level do not usually provide information about the exact position of the samples taken and their distribution. Moreover, in such studies samples are usually taken systematically from stump height, because of the accessibility of this position and that fact that no commercial logs are wasted. However, the effect of this on the estimation of the dry mass is still not clear. The questions addressed in Chapter IV are:  What is the effect of sampling intensity on the dry mass estimation of sampling trees when systematic sampling and ratio type estimators are applied?  Do crown variables improve the accuracy of crown fraction biomass equations?  What is the ability of biomass equations to predict the proportion of biomass fractions?  Is there a systematic bias inherent in the bole sampling procedure? 1.4.3. Are Eucalyptus plantations suitable for carbon sequestration and bioenergetic purposes at the stand densities conventionally used in southern Europe? Chapter IV The effects of forests as regards climate change mitigation may be derived from product yield, bioenergy substitution and C stocks in biomass and soils (Canadell & Raupach, 2008). Product yield is usually expressed in terms of harvested or produced volume, whereas bioenergy substitution is usually expressed in terms of energy, or tonnes oil equivalent (TOE). This depends on the transformation procedure of the biomass in question. When energy is transformed by combustion it is the high heating value (HHV) of each biomass compartment that determines the transformation performance. When the biomass is transformed to second generation biofuels (i.e. bioethanol), it is the cellulose content that determines the transformation rate to this type of energy. The TOE in each case depends on the source of energy substituted, while the same amount of energy can be obtained for combustion of several types of combustible; as a convention, this must be expressed in terms of the energy released by the combustion of one Mg of oil (4·1010 J), which is equivalent to a 1.4 Mg of coal, 4.5 Mg of lignite and 10·103 m3 of natural gas. The traditional way of estimating this type of bioenergy production is from volume or biomass (Fig. 1.2). This implies a second estimation phase in which the energetic yield value is estimated from volume or biomass. A direct relation of each one of these forms of energy or carbon by specific models may avoid concatenating errors. Moreover, this would allow more accurate comparison of several management alternatives or the use of particular species for energy purposes. CHAPTER I 20 BEFs ENERGY STANDTREE VARIABLES VE ENERGY EQUATIONS CARBON CARBON EQUATIONS USABLE CELLULOSE CC LHV SPECIFIC COMPSUMPTION USABLE CELLULOSE EQUATIONS BIOMASS EQUATIONS BIOMASS VOLUME BEFs ENERGY STANDTREE VARIABLES VE ENERGY EQUATIONS CARBON CARBON EQUATIONS USABLE CELLULOSE CC LHV SPECIFIC COMPSUMPTION USABLE CELLULOSE EQUATIONS BIOMASS EQUATIONS BIOMASS VOLUME Figure 1.2. Process of biomass, carbon, energy and usable cellulose estimation from stand or tree variables. VE: volume equations; BEFs: biomass expansion factors; CC carbon concentration in biomass; LHV low heating value. Most studies that compare species or management alternatives for short rotation woody crops (SRWC) are based on a small number of plots (Mitchell et al., 1999; Sims et al., 1999b; Dickmann, 2006), sometimes a few trees, and usually established at very high densities. As thinning is not applied in SRWC management, static growth models may be suitable for yield and bioenergy production modelization, although such studies are still scarce and focus on species used for SRWC in northern Europe. Because of the large installation costs of these high density crops, the requirements of some dimensional constraints for harvesting and production of second generation combustibles, and the low implementation of this type of crop in the north of Spain, the profitability of the current management schemes for bioenergy production is being investigated. This is important because it represents the actual potential to produce biomass without substantial changes to existing management schemes. The questions addressed in Chapter IV are:  How well are standard low density plantations of Eucalyptus adapted for bioenergy purposes and carbon sequestration in northern Spain?  How do harvesting limitations as regards tree size affect stand management in short rotation woody crops of Eucalyptus?  How wide is the management window as regards self thinning for each initial stocking density? 1.4.4. How does soil organic matter nature change during afforestation of former pasture land? Chapter V Although the first step in describing SOC dynamics after land use change is to examine changes in total SOC, there are large differences among SOM compounds regarding stability, GENERAL INTRODUCTION AND OBJECTIVES 21 which must therefore be studied in greater detail. Although there are some studies about the effect of climate change on the chemical stability of soil organic matter compounds (Bellamy et al., 2005; Kirschbaum, 2006), there is an important gap in the knowledge about the chemical nature of the SOC following land use change. While information about total SOC may suggest that all C released after land use change occur in the most labile fractions, it is known that degradation of recalcitrant organic matter readily takes place in upper horizons in which some stabilization processes are less active than in deeper horizons (von Lützow et al., 2006). Moreover, more detailed work is necessary to examine changes in this type of process, since the traditional classification of SOC into labile and recalcitrant is not sufficient for describing the SOC stability and dynamics. Traditional techniques for examining changes in chemical SOC stability include 13C CP-MAS NMR, Fourier transform infrared spectroscopy and pyrolysis/GC-MS. Although these methods are highly reproducible and accurate, they are also expensive and time consuming. However, some studies on SOM dynamics, especially those focused on modelling requires analysis of a large number of samples, while at the same time offering a high degree of reproducibility and accuracy. In this sense, calorimetry and thermal analysis are rapid, inexpensive techniques that provide information about the stability of SOC and have already been tested in steady state systems. Nevertheless, accurate methods of describing dynamic processes are being investigated. The questions addressed in Chapter V are:  What fractions of SOC are affected by the losses when large losses of SOC occur in the soil?  How suitable are calorimetry and thermal analysis for detecting SOC stability changes in forest soils? 1.4.5. How do management practices affect the mitigation effect of fast growing plantations established on former pasture land? Chapter VI Prediction through empirical models is restricted to unchanged conditions rather than to adjusted conditions (IPPC, 2006). This also implies that management practices are considered to be the same as those observed in the population modelled. The equilibrium between management intensity, C stock and mitigation effect is difficult to assess. Although excessive harvesting schedules can lead to severe damage to forest ecosystems, under-harvesting can lead to under use of the stand potential, and maintenance of the forest at more unstable stages of development, as regards e.g. fires, storm events and diseases. This must be considered for maximizing the potential of forests as regards climate change mitigation. The effect of several management practices on changes in forest C accumulation must be assessed with models that are sensitive to this type of change. In this sense, the CO2Fix model (Nabuurs et al., 2002; Masera et al., 2003; Schelhaas et al., 2004) enables estimation of the changes in C stocks and fluxes by use of the full carbon CHAPTER I 22 accounting approach. This model is divided into six modules: biomass, soil, products, bioenergy, financial and carbon accounting. Changes in C in all modules are driven by current annual increments in the species considered, and also by the climatic data and the initial conditions regarding C stock. The usefulness of the model has already been demonstrated for even and uneven-aged forest stands worldwide (Schelhaas et al., 2004; Groen et al., 2006), enabling comparison between different management practices and also initial conditions. Since there are no either purely eco-physiological or empirical models (Landsberg & Sands, 2011), estimations for some modules of the model used are obtained from empirical data. Although previous studies of the sensitivity of the chosen model in regard to the input data indicates that yield tables are one of the most important parameters (Nabuurs et al., 2008), different models are sometimes available and direct measurements are not always available for validation. Three different models were therefore used to produce yield tables and were validated with measured data. This same process was carried out with the soil module, since there was not sufficient information about its usefulness for southern European conditions. Moreover, there is not enough information about the mitigation effect of forest products available on the northern Spain market. The questions addressed in Chapter VI are:  How important are accurate yield tables for describing C accumulation on biomass by using the CO2Fix model?  How well does the soil module of the CO2Fix model (Yasso model) enable estimation of the changes in C in forest soils in northern Spain, and how sensitive is the module to turnover rate parameters?  How do site conditions and management alternatives affect the C accumulation effect in fast growing plantations established on former pasture land?  What is the contribution of forest products and bioenergy substitution to the mitigating effect of the forest plantations in northern Spain? 1.5. Objectives The overall objective of this thesis was to develop methods and tools for estimating carbon stocks in biomass and soil in Pinus radiata (D.Don), Eucalyptus globulus (Labill) and Eucalyptus nitens (Dean & Maiden) Maiden plantations over former agricultural land in north-western Spain. The specific objectives were:  To evaluate carbon accumulation in biomass, litter and soil on former agricultural land afforested with different species (Chapter II). GENERAL INTRODUCTION AND OBJECTIVES 23  To develop biomass equations at tree level for E. nitens plantations, and to evaluate the effects on tree dry mass estimation of the sampling intensity, sampling methodology and independent variables considered (Chapter III).  To evaluate the energetic and carbon sequestration ability of E. globulus and E. nitens plantations in short rotation woody crops (Chapter IV).  To evaluate the stability of the soil organic matter after land use change from pasture to forest plantation with fast growing tree species (Chapter V).  To evaluate the mitigation effect of several management regimes for the tree species studied, considering the whole product cycle and the stock effect on soil and biomass (Chapter VI). 1.6. References Álvarez, P. 2004. Viveros forestales y uso de planta en repoblación en Galicia. Tesis Doctoral. Lugo, Spain, University of Santiago de Comostela. António, N.; Tomé, M.; Tomé, J.; Soares, P. & Fontes, L. 2007. Effect of tree, stand, and site variables on the allometry of Eucalyptus globulus tree biomass. Canadian Journal of Forest Research. 37 (5): 895-906. Baldocchi, D. 2008. 'Breathing'of the terrestrial biosphere: lessons learned from a global network of carbon dioxide flux measurement systems. Australian Journal of Botany. 56 (1): 1-26. Baldocchi, D.; Falge, E.; Gu, L.; Olson, R.; Hollinger, D.; Running, S.; Anthoni, P.; Bernhofer, C.; Davis, K. & Evans, R. 2001. FLUXNET: A new tool to study the temporal and spatial variability of ecosystem-scale carbon dioxide, water vapor, and energy flux densities. Bulletin of the American Meteorological Society. 82 (11): 2415-2434. Bateson, L.; Vellico, M.; Beaubien, S.E.; Pearce, J.M.; Annunziatellis, A.; Ciotoli, G.; Coren, F.; Lombardi, S. & Marsh, S. 2008. The application of remote-sensing techniques to monitor CO2-storage sites for surface leakage: method development and testing at Latera (Italy) where naturally produced CO2 is leaking to the atmosphere. International Journal of Greenhouse Gas Control. 2 (3): 388-400. Battisti, A.; Stastny, M.; Netherer, S.; Robinet, C.; Schopf, A.; Roques, A. & Larsson, S. 2005. Expansion of geographic range in the pine processionary moth caused by increased winter temperatures. Ecological Applications. 15 (6): 2084-2096. Bellamy, P.H.; Loveland, P.J.; Bradley, R.I.; Lark, R.M. & Kirk, G.J.D. 2005. Carbon losses from all soils across England and Wales 1978-2003. Nature. 437 (7056): 245-248. Blennow, K. & Olofsson, E. 2008. The probability of wind damage in forestry under a changed wind climate. Climatic Change. 87 (3): 347-360. Boland, D.J.; Brooker, M.I.H. & Turnbull, J.W. 1980. Eucalyptus Seed Canberra, Australia. 191 pp. Bravo, F. 2007. El papel de los bosques españoles en la mitigación del cambio climático Fundación Gas Natural, Barcelona, Spain. 315 pp. Briggs, R.D.; Cunia, T.; White, E.H. & Yawney, H.W. 1987. Estimating sample tree biomass by subsampling: some empirical results. Estimating Tree Biomass Regressions and Their Error.Proceedings of the Workshop on Tree Biomass Regression Functions and their Contribution to the Error of Forest Inventory Estimates.Compiled by EH Wharton and T.Cunia.USDA For.Serv.Gen.Tech.Rep.NE-117. : 119-127. CHAPTER I 24 Brown, S.; Lim, B. & Schlamadinger, B. 1999. Evaluating approaches for estimating net emissions of carbon dioxide from forest harvestng and wood products. Meeting report, Dakar, 5-7 May 1998 Edn., IPCC/OECD/IDEA Programme on National Greenhouse Gas Inventories, Paris. Campioli, M.; Härkönen, S.; Lombardi, F.; Lucas-Borja, M.E.M., D.; Pérez-Cruzado, C.; Räty, T. & Sitkova, Z. 2009. Role of Forest Research in Fighting Climate Change-Forest Research Work Group Report. En: Fighting Climate Change: Adapting Forest Policy and Forest Management in Europe. Eds. T. Karjalainen, M. Lindner, A. Niskanen & M. Lier, Joensuu Forestry Networking Week 2009 Edn., Finnish Forest Research Institute (METLA), Joensuu, Finland. Canadell, J.G. & Raupach, M.R. 2008. Managing forests for climate change mitigation. Science. 320 (5882): 1456-1457. Carey, E.V.; Sala, A.; Keane, R. & Callaway, R.M. 2001. Are old forests underestimated as global carbon sinks? Global Change Biology. 7 (4): 339-344. Carle, J.B. & Holmgren, L.P.B. 2009. Wood from Planted Forests: Global Outlook to 2030. En: Planted Forests. Uses, Impacts & Sustainability. Ed. J. Evans, Food and Agriculture Organization of the United Nations (FAO) and CAB International, Rome, Italy. 47-60 pp. Carvalho, J.P. & Parresol, B.R. 2003. Additivity in tree biomass components of Pyrenean oak (Quercus pyrenaica Willd.). Forest Ecology and Management. 179 (1-3): 269-276. Clark, A. 1982. Predicting biomass production in the South. En: Predicting Growth and Yield in the Mid-South 119-139 pp. Conen, F.; Zerva, A.; Arrouays, D.; Jolivet, C.; Jarvis, P.G.; Grace, J. & Mencuccini, M. 2004. The carbon balance of forest soils: detectability of changes in soil carbon stocks in temperate and boreal forests. En: The Carbon Balance of Forest Biomes. Eds. H. Griffiths & P.G. Jarvis, Garland Science/BIOS Scientific Publishers, Southampton, UK. 233-247 pp. Cotterill, P. & Macrae, S. 1997. Improving eucalyptus pulp and paper quality using genetic selection and good organization. Tappi Journal. 80 (6): 82-89. Del Lungo, A. 2009. Planted Forest Areas by Country. En: Planted Forests. Uses, Impacts & Sustainability. Ed. J. Evans, Food and Agriculture Organization of the United Nations (FAO) and CAB International, Rome, Italy. 155-180 pp. Dias, A.C.; Arroja, L. & Capela, I. in press. Carbon storage in harvested wood products: implications of different methodological procedures and input data—a case study for Portugal. European Journal of Forest Research. : 1-9. Dickmann, D.I. 2006. Silviculture and biology of short-rotation woody crops in temperate regions: Then and now. Biomass and Bioenergy. 30 (8-9): 696-705. Evans, J.; Carle, J.B.; Ball, J.B.; Del Lungo, A.; Neilson, D.A. & Holmgren, L.P.B. 2009. Planted Forests. Uses, Impacts & Sustainability Food and Agriculture Organization of the United Nations (FAO) and CAB International, Rome, Italy. 213 pp. FAO 2005. Global Forest Resources Assessment 2005 Food and Agriculural Organization of the United Nations (FAO), Rome, Italy. Foley, J.A.; DeFries, R.; Asner, G.P.; Barford, C.; Bonan, G.; Carpenter, S.R.; Chapin, F.S.; Coe, M.T.; Daily, G.C. & Gibbs, H.K. 2005. Global consequences of land use. Science. 309 (5734): 570-574. Fulton, L.; Howes, T. & Hardy, J. 2004. Biofuels for transport: an international perspective OECD, International Energy Agency, Paris, France. González-Río, F.; Castellanos, A.; Fernández, O. & Gómez, C. 1997. Manual de selvicultura del Eucalipto Universidad de Santiago de Compostela, Escuela Politécnica Superior de Lugo. 92 pp. Grace, J. 2004. Understanding and managing the global carbon cycle. Journal of Ecology. 92 (2): 189-202. GENERAL INTRODUCTION AND OBJECTIVES 25 Gracia, C.; Sánchez, A.; Pla, E.; Sabaté, S.; Lexer, M.J. & Jäger, D. 2005. Adaptive forest management strategies. En: Management of European forest under changing climatic conditions. Eds. S. Kellomäki & S. Leinone, Research Notes Nº163 Edn., Faculty of Forestry, University of Joensuu, Joensuu, Finland. 160-252 pp. Groen, T.; Nabuurs, G. & Schelhaas, M. 2006. Carbon accounting and cost estimation in forestry projects using CO2Fix V. 3. Climatic Change, 74. 1 (3): 269-288. Guaita, M. & Eiras, A. 2007. Estudio comparativo entre la madera de pino y de eucalipto para diversas tipologías estructurales. Boletín del CIDEU. 4: 71-76. Hamrick, J.L. 2004. Response of forest trees to global environmental changes. Forest Ecology and Management. 197 (1-3): 323-335. Hansen, J.E. & Lacis, A.A. 1990. Sun and dust versus greenhouse gases: An assessment of their relative roles in global climate change. Nature. 346 (6286): 713-719. Harmon, M.E.; Ferrell, W.K. & Franklin, J.F. 1990. Effects on carbon storage of conversion of oldgrowth forests to young forests. Science. 247 (4943): 699. Houghton, J.T.; Filho, L.M.; Lim, B.; Treanton, K.; Mamaty, I.; Bonduki, Y.; Griggs, D. & Callender, B. 1997. Revised 1996 IPPC guidelines for national greenhouse gas inventories. IPPC/OECD/IEA UK Meteorological Office, Bracknell, UK. Houghton, J.T.; Jenkins, G.J. & Ephraums, J.J. 1990. Climate change: the IPCC scientific assessment Cambridge University Press, Cambridge, MA (United States). 310 pp. Houghton, R.A. 2003. Why are estimates of the terrestrial carbon balance so different? Global Change Biology. 9 (4): 500-509. House, J.I.; Prentice, I.C.; Ramankutty, N.; Houghton, R.A. & Heimann, M. 2003. Reconciling apparent inconsistencies in estimates of terrestrial CO2 sources and sinks. Tellus B. 55 (2): 345-363. INFOR 2004. Eucalyptus nitens en Chile: Primera monografía INFOR CORFO, Valdivia, Chile. 142 pp. IPPC 2007. Climate change 2007: the phisical science basis, (sumary for Policymakers). URL: http://ipcc-wg1.ucar.edu/wg1/docs/WG1AR4_SPM_PlenaryApproved.pdf Consultado:Diciembre/15. IPPC 2006. 2006 IPPC Guidelines for National Greenhouse Gas Inventories Intergubernamental Pannel of Climate Change, Japan. IPPC 2003. Good practice guidance for land use, land-use change and foresty. National Greenhause Gas Inventories Programme. Intergubernamental Pannel of Climate Change, Kanagawa, Japan. 295 pp. IPPC 2001. Climate change 2001: the scientific basis. En: Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change. Eds. J.T. Houghton, Y. Ding, D.J. Griggs, et al, Cambridge University Press, Cambridge, U.K. 881 pp. Jandl, R.; Lindner, M.; Vesterdal, L.; Bauwens, B.; Baritz, R.; Hagedorn, F.; Johnson, D.W.; Minkkinen, K. & Byrne, K.A. 2007. How strongly can forest management influence soil carbon sequestration? Geoderma. 137 (3-4): 253-268. Jones, P.D.; Osborn, T.J. & Briffa, K.R. 2001. The evolution of climate over the last millennium. Science. 292 (5517): 662. Jones, P. & Moberg, A. 2003. Hemispheric and large-scale surface air temperature variations: An extensive revision and an update to 2001. Journal of Climate. 16 (2): 206-223. Kaipainen, T.; Liski, J.; Pussinen, A. & Karjalainen, T. 2004. Managing carbon sinks by changing rotation length in European forests. Environmental Science & Policy. 7 (3): 205-219. Keenan, T.; Sabate, S. & Gracia, C. 2008. Forest eco-physiological models and carbon sequestration. Managing Forest Ecosystems: The Challenge of Climate change. 17 (3): 83102. CHAPTER I 26 Kirschbaum, M.U.F. 2006. The temperature dependence of organic-matter decomposition--still a topic of debate. Soil Biology and Biochemistry. 38 (9): 2510-2518. Kirschbaum, M.U.F. 2003. To sink or burn? A discussion of the potential contributions of forests to greenhouse gas balances through storing carbon or providing biofuels. Biomass and Bioenergy. 24 (4-5): 297-310. Landsberg, J.J. & Sands, P.J. 2011. Phisiological Ecology of Forest Production. Principles, Process and Models. Volume 4 Edn., Elsevier Academic Press, Burlington, USA. 331 pp. Lasanta, T. 1996. El proceso de marginación de tierras en España. En: Erosión y recuperación de tierras en áreas marginales. Eds. T. Lasanta & J.M. García Ruiz, Sociedad Española de Geomorfología, Logroño, Spain. 7-31 pp. Lavery, P.B. 1986. Plantation forestry with Pinus radiata: review papers. Paper Nº 12 Edn., School of Forestry, University of Canterbury, New Zealand. Lavorel, S.; Touzard, B.; Lebreton, J.D. & Clément, B. 1998. Identifying functional groups for response to disturbance in an abandoned pasture. Acta Oecologica. 19 (3): 227-240. Lindner, M. & Karjalainen, T. 2007. Carbon inventory methods and carbon mitigation potentials of forests in Europe: a short review of recent progress. European Journal of Forest Research. 126 (2): 149-156. Lindner, M.; Maroschek, M.; Netherer, S.; Kremer, A.; Barbati, A.; Garcia-Gonzalo, J.; Seidl, R.; Delzon, S.; Corona, P. & Kolström, M. 2010. Climate change impacts, adaptive capacity, and vulnerability of European forest ecosystems. Forest Ecology and Management. 259 (4): 698709. Liski, J.; Perruchoud, D. & Karjalainen, T. 2002. Increasing carbon stocks in the forest soils of western Europe. Forest Ecology and Management. 169 (1-2): 159-175. Liski, J.; Pussinen, A.; Pingoud, K.; Mäkipää, R. & Karjalainen, T. 2001. Which rotation length is favourable to carbon sequestration? Canadian Journal of Forest Research. 31 (11): 20042013. Little, K.M. & Gardner, R.A.W. 2003. Coppicing ability of 20 Eucalyptus species grown at two highaltitude sites in South Africa. Canadian Journal of Forest Research. 33 (2): 181-189. Löwe, H.; Seufert, G. & Raes, F. 2000. Comparison of methods used within Member States for estimating CO~ 2 emissions and sinks according to UNFCCC and EU Monitoring Mechanism: forest and other wooded land. Biotechnologie Agronomie Societe et Environnement. 4 (4): 315-319. Luyssaert, S.; Schulze, E.D.; Börner, A.; Knohl, A.; Hessenmöller, D.; Law, B.E.; Ciais, P. & Grace, J. 2008. Old-growth forests as global carbon sinks. Nature. 455 (7210): 213-215. Malhi, Y.; Meir, P. & Brown, S. 2002. Forests, carbon and global climate. Philosophical Transactions of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences. 360 (1797): 1567-1591. Malhi, Y.; Roberts, J.T.; Betts, R.A.; Killeen, T.J.; Li, W. & Nobre, C.A. 2008. Climate change, deforestation, and the fate of the Amazon. Science. 319 (5860): 169-172. Mansilla Vázquez, J.P. 1992. Presencia sobre Eucalyptus globulus Labill de Gonepterus scutellatus Gyll (Col. Curculionidae) en Galicia. Boletín de Sanidad Vegetal: Plagas. 18 (3): 547-554. MAPA 2006. Forestación de Tierras Agrícolas: Análisis de su evolución y contribución a la fijación de carbono y al uso racional de la tierra Spanish Ministry of Agriculture, Fisheries and Food, Madrid (Spain). Masera, O.R.; Garza-Caligaris, J.F.; Kanninen, M.; Karjalainen, T.; Liski, J.; Nabuurs, G.J.; Pussinen, A.; De Jong, B.H.J. & Mohren, G.M.J. 2003. Modeling carbon sequestration in afforestation, agroforestry and forest management projects: the CO2FIX V. 2 approach. Ecological Modelling. 164 (2-3): 177-199. GENERAL INTRODUCTION AND OBJECTIVES 27 McMahon, S.M.; Parker, G.G. & Miller, D.R. 2010. Evidence for a recent increase in forest growth. Proceedings of the National Academy of Sciences. 107 (8): 3611-3615. Meehl, G.A.; Stocker, T.F.; Collins, W.D.; Friedlingstein, P.; Gaye, A.T.; Gregory, J.M.; Kitoh, A.; Knutti, R.; Murphy, J.M. & Noda, A. 2007. Global climate projections. En: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Eds. S. Solomon, D. Qin, M. Manning, et al, IPPC, Cambridge, U.K. 747-845 pp. Melillo, J.M.; Reilly, J.M.; Kicklighter, D.W.; Gurgel, A.C.; Cronin, T.W.; Paltsev, S.; Felzer, B.S.; Wang, X.; Sokolov, A.P. & Schlosser, C.A. 2009. Indirect emissions from biofuels: How important? Science. 326 (5958): 1397-1399. Mitchell, C.P.; Stevens, E.A. & Watters, M.P. 1999. Short-rotation forestryoperations, productivity and costs based on experience gained in the UK. Forest Ecology and Management. 121: 123136. Nabuurs, G.J.; Garza-Caligaris, J.F.; Kanninen, M.; Karjalainen, T.; Lapvetelainen, T.; Liski, J.; Masera, O.; Mohren, G.M.J.; Olgin, M.; Pussinen, A. & Schelhaas, M. 2002. CO2FIX V2.0. Manual of a model for quantifying carbon sequestration in forest ecosystems and wood products. Report 445 Edn., Alterra, Wageningen, The Netherlands. 44 pp. Nabuurs, G.J.; van Putten, B.; Knippers, T.S. & Mohren, G.M.J. 2008. Comparison of uncertainties in carbon sequestration estimates for a tropical and a temperate forest. Forest Ecology and Management. 256 (3): 237-245. Ney, R.A.; Schnoor, J.L. & Mancuso, M.A. 2002. A methodology to estimate carbon storage and flux in forestland using existing forest and soils databases. Environmental Monitoring and Assessment. 78 (3): 291-307. Nutto, L. & Touza Vázquez, M.C. 2004. Producción de Madera de Sierra de alta Calidad con Eucalyptus globulus. CIS-Madera. 12: 6-18. Oliver, G.R.; Beets, P.N.; Garrett, L.G.; Pearce, S.H.; Kimberly, M.O.; Ford-Robertson, J.B. & Robertson, K.A. 2004. Variation in soil carbon in pine plantations and implications for monitoring soil carbon stocks in relation to land-use change and forest site management in New Zealand. Forest Ecology and Management. 203 (1-3): 283-295. Otero, L.; Aguín, O.; Sainz, M. & Mansilla, J.P. 2007. El género Mycosphaerella en plantaciones de Eucalyptus en Galicia. Boletín de Sanidad Vegetal: Plagas. 33 (4): 503-516. Pardos, J.A. 2010. Los ecosistemas forestales y el secuestro de carbono ante el calentamiento global. Monografías INIA: Serie Forestal Edn., Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria, Ministerio de Ciencia e Innovación, Madrid, Spain. 253 pp. Paz, J. 1999. Análisis comparativo de las propiedades microscópicas y macroscópicas de la madera de Eucalyptus nitens y Eucalyptus globulus. En: XII Silvotecna, Realidad y potencial del Eucalyptus en Chile (Concepción, Chile). Pemán García, J.; Ruvireta Brujan, J. & Blanco Ortiz, R. 2009. Influencia de la disponibilidad de superficie en la actividad repobladora desarrollada por la administración entre 1940 y 1973. En: Actas del Congreso de Historia Forestal, III Reunión sobre Historia Forestal. Eds. F.J. Silva-Pando, E. García Gómez, J. Pereira Sieso & J. Carrobles Santos, Cuaderonos de la Sociedad Española de Ciencias Forestales (Nº 30) Edn., Sociedad Española de Ciencias Forestales, Toledo, Spain. 325-330 pp. Pérez, S.; Renedo, C.J.; Ortiz, A. & Manana, M. 2008. Energy potential of waste from 10 forest species in the North of Spain (Cantabria). Bioresource Technology. 99 (14): 6339-6345. Pregitzer, K.S. & Euskirchen, E.S. 2004. Carbon cycling and storage in world forests: biome patterns related to forest age. Global Change Biology. 10 (12): 2052-2077. Righelato, R. & Spracklen, D.V. 2007. Carbon mitigation by biofuels or by saving and restoring forests. Science. 317 (5840): 902-902. CHAPTER I 28 Rodríguez-Murillo, J.C. 1999. El ciclo mundial del carbono. Método de cálculo por los cambios de uso de la tierra. Balance de carbono en los bosques españoles. En: El calentamiento global en España. Ed. F. Hernández-Álvarez, CSIC, Madrid, Spain. 97-139 pp. Ruiz de la Torre, J.; Ceballos y Fernández de Córdoba, L.; Ceballos Jiménez, M. & Ruiz del Castillo y de Navascúes, J. 1979. Árboles y arbustos de la España peninsular Escuela Técnica Superior de Ingenieros de Montes, Madrid (Spain). 512 pp. Satoo, T. & Madgwick, H.A.I. (eds) 1982. Forest Biomass, JSTOR, The Netherlands. Saugier, B.; Roy, J. & Mooney, H.A. 2001. Estimations of global terrestrial productivity: converging toward a single number? En: Terrestrial global productivity. Eds. J. Roy, B. Saugier & H.A. Mooney, Academic Press, San Diego, CA. 543-557 pp. Schelhaas, M.J.; van Esch, P.W.; Groen, T.A.; de Jong, B.H.J.; Kanninen, M.; Liski, J.; Masera, O.; Mohren, G.M.J.; Nabuurs, G.J.; Palosuo, T.; Pedroni, L.; Vallejo, A. & Vilén, T. 2004. CO2FIX V 3.1 A modelling framework for quantifying carbon sequestration in forest ecosystems Alterra, Wageningen, (The Netherlands). 122 pp. Searchinger, T.; Heimlich, R.; Houghton, R.A.; Dong, F.; Elobeid, A.; Fabiosa, J.; Tokgoz, S.; Hayes, D. & Yu, T.H. 2008. Use of US croplands for biofuels increases greenhouse gases through emissions from land-use change. Science. 319 (5867): 1238-1240. Shukla, J.; Nobre, C. & Sellers, P. 1990. Amazon deforestation and climate change. Science. 247 (4948): 1322-1325. Sims, R.E.H.; Maiava, T.G. & Bullock, B.T. 2001. Short rotation coppice tree species selection for woody biomass production in New Zealand. Biomass and Bioenergy. 20 (5): 329-335. Sims, R.E.H.; Senelwa, K.; Maiava, T. & Bullock, B.T. 1999a. Eucalyptus species for biomass energy in New Zealand—I: Growth screening trials at first harvest. Biomass and Bioenergy. 16 (3): 199-205. Sims, R.E.H.; Senelwa, K.; Maiava, T. & Bullock, B.T. 1999b. Eucalyptus species for biomass energy in New Zealand—Part II: Coppice performance. Biomass and Bioenergy. 17 (4): 333343. Smith, J.E.; Heath, L.S. & Woodbury, P.B. 2004. How to estimate forest carbon for large areas from inventory data. Journal of Forestry. 102 (5): 25-31. Solomon, S.; Plattner, G.K.; Knutti, R. & Friedlingstein, P. 2009. Irreversible climate change due to carbon dioxide emissions. Proceedings of the National Academy of Sciences. 106 (6): 17041709. Ståhl, G.; Boström, B.; Lindkvist, H.; Lindroth, A.; Nilsson, J. & Olsson, M. 2004. Methodological options for quantifying changes in carbon pools in Swedish forests. No. 214 Edn., Swedish University of Agricultural Sciences, Uppsala, Swedish. 46 pp. Thomas, C.D.; Cameron, A.; Green, R.E.; Bakkenes, M.; Beaumont, L.J.; Collingham, Y.C.; Erasmus, B.F.N.; De Siqueira, M.F.; Grainger, A. & Hannah, L. 2004. Extinction risk from climate change. Nature. 427 (6970): 145-148. Thuiller, W. 2003. BIOMOD-optimizing predictions of species distributions and projecting potential future shifts under global change. Global Change Biology. 9 (10): 1353-1362. Thuiller, W.; Lavorel, S.; Araújo, M.B.; Sykes, M.T. & Prentice, I.C. 2005. Climate change threats to plant diversity in Europe. Proceedings of the National Academy of Sciences of the United States of America. 102 (23): 8245. Tilman, D.; Hill, J. & Lehman, C. 2006. Carbon-negative biofuels from low-input high-diversity grassland biomass. Science. 314 (5805): 1598-1600. Trenberth, K.E. & Josey, S.A. 2007. Observations: surface and atmospheric climate change. En: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Eds. S. INFLUENCE OF TREE SPECIES ON C SEQUESTRATION IN AFFORESTED PASTURES IN A HUMID TEMPERATE REGION 35 end of the rotation. The afforestation may therefore result in high temporal C losses via soil respiration in comparison with the C inputs by litter. Some authors have suggested that this is a temporary effect, in which only the labile C pool is exhausted (Cerli et al., 2008; Huang et al., 2011). However, recent evidence shows that degradation of recalcitrant SOM can also occur in the uppermost soil layers (von Lützow et al., 2006, Chapter V). It is not clear whether these initial losses are compensated in fast growing, intensively managed species. The C balance after afforestation is greatly affected by the tree species, as a result of differences in growth rates of the trees. In addition, litter production and litter quality, which are greatly influenced by tree species, have a strong influence on the SOC dynamics (Berg, 2000; Vesterdal et al., 2008). However, since most of the current knowledge is based on studies of afforestation with coniferous species (Berthrong et al., 2009), the influence of the tree species on SOC dynamics has not yet been accurately evaluated. In studies that have attempted to evaluate the effect of tree species on afforested land, the differences in SOC dynamics are attributed to the influence of the different turnover rates of the litter (Vesterdal et al., 2008), the cover and type of ground vegetation (Lemma et al., 2006), or both (Paul et al., 2002; Huang et al., 2011; Kasel et al., 2011). The influence of N-fixing tree species has been also be recognized (Nilsson & Schopfhauser, 1995; Kasel et al., 2011), as a higher yield in poor soils leads to higher organic matter inputs to the soil, and also improves litter quality and the speed of decomposition of OM (Conteh et al., 1997). Recent studies have shown that these different sources of litter inputs can even lead to changes in the SOM composition (Huang et al., 2011; Chapter V). The capacity of the soil to act as a C sink is also influenced by how rapidly the litter converts C into humus (Silver et al., 2004; e.g. Kanerva & Smolander, 2007; Prescott, 2010). The decomposability of the litter not only depends on its chemical composition, but also on microclimatic conditions determined by the different canopy development and stand structure. Moreover, the same microclimatic conditions have a direct influence on the cover and type of ground vegetation, which in turn alters the amounts, composition and type of litter (Ostertag et al., 2008; Berg et al., 2009). This is particularly important because roots incorporate more C into the soil than litter layer material (Jones et al., 2009). Thus, ground vegetation dominated by grass species incorporates C rapidly into soil organic matter because the root system develops quickly (Andrade et al., 2008; Laungani & Knops, 2009). The silvicultural parameters most closely related to SOC dynamics are: tree species, site preparation techniques, initial stocking, rotation length and other parameters more specific to each type of tree species management (pruning, clearcuting, thinning, application of fertilizer, etc.) and autoecology. These aspects have been discussed by Jandl et al. (2007) and, in the case of the species considered here, by Balboa-Murias et al. (2006). In addition to management practices, a comprehensive representation of the entire forestry sector system should be considered, taking into account the C pools and fossil fuel substitution, although the latter pool has not been considered in the international agreements on reduction of emissions (UNFCCC, 2002). All of these CHAPTER II 36 factors are particularly important in fast growing tree species, in which intensive management in short rotations, as well as harvesting of logging residues may prevent accumulation of SOC in the long term. Most studies concerning the effect of different species and types of management on C sequestration in afforested land have compared the C stocks in several pools in pasture land and forest plantations in the steady state, rather than considering the dynamic changes that take place. In many cases, the shifts in key parameters throughout the rotation, such as tree growth, stand structure, associated vegetation and litter development are not considered. Thus, the temporal dynamics not only provide an understanding of the different mechanisms of C sequestration after afforestation, but are also useful for designing the most appropriate type of management to maximize the C sink capacity. Most approaches evaluating the capacity of soils as C sinks focus on plot level, and few studies have been extended to landscape levels (Johnston et al., 1996; Turner & Lambert, 2000; Conant et al., 2003). Such studies show the high degree of variability in the SOC dynamics following afforestation, even under rather homogeneous conditions. This high variability emphasizes the risk of making erroneous conclusions about SOC dynamics when the experimental design does not take this variability into account (Berg et al., 2009). A correct methodology must ensure adequate sampling replication at plot level, and proper sample analysis to take into account most of the variability for extrapolation of the results of plot level studies to a larger scale (Goidts et al., 2009). Methodological procedures for quantifying the changes in SOC after afforestation are: (i) paired sites, (ii) chronosequence studies, (iii) multiple re-sampling, and (iv) process and modelling studies (Turner & Lambert, 2000). Chronosequence studies use a series of plots in plantations of different ages with supposedly similar management regimes and environmental conditions, and translate spatial differences between soils into temporal differences (Huggett, 1998). Although the chronosequence approach cannot replace long-term field experiments, there are certain disadvantages with the latter, such as the delay in obtaining results, workload and particular trends in external parameters (i.e. climate and local conditions), which may cause a systematic bias in the observations. The paired-plots approach is an alternative method in which one of the paired plots represents the initial conditions. Because of the high spatial variability in SOC measurements (Johnston et al., 1996; Turner & Lambert, 2000; Conant et al., 2003), the combined use of chronosequences and the paired-plots approach may provide a suitable way of detecting changes in soil pools, and of correcting local tendencies. The objectives of the present study were: (i) to examine the influence of tree species on the C dynamics in the forest system (tree biomass, litter and mineral soil) following afforestation of pasture land, (ii) to explore the relationships between tree biomass development on litter and SOC dynamics in three forest plantations established on former pasture land, and (iii) to evaluate the C sink capacity of the different types of plantations in relation to management. A specific methodology for sampling and data management was designed to restrict most of the variability on INFLUENCE OF TREE SPECIES ON C SEQUESTRATION IN AFFORESTED PASTURES IN A HUMID TEMPERATE REGION 37 a regional scale and therefore to provide accurate information on C dynamics in the different tree species planted. The design was applied to temperate forest plantations of southern Europe, one of the most productive timber production systems in Europe, where important afforestation processes may play an important role in mitigating CO2 and other GHGs emissions. The data obtained in this study will provide valuable information about the effects of such afforestation programmes on C sink capacity. 2.2. Materials and methods 2.2.1. Site description and experimental design The study was carried out in Galicia (NW Spain), in an area of 7000 km-2 representative of the Atlantic-climate zone of northern Spain. The 20 year annual average rainfall in the area is 1378 mm (range 884-2107 mm), and the temperature, 12.1ºC (range 10.3-14.8ºC). The wettest month is November, with an average rainfall of 160 mm, and the driest August, with 44 mm. The lowest mean monthly temperature 7.1ºC occurs in January, and the highest 18.2ºC, in August. The soil humidity and temperature regimes are Udic (mean period with partial drought, 1 month) and Mesic (mean frost-free period, 10 months), respectively. The soils were developed from granitic rocks, schist and shale, and classified as Humic or Distric Cambisols and Alumi-humic Umbrisols (IUSS Working Group WRB, 2006). The soil has a loam or sandy loam texture and is well drained. The average values for selected characteristics of the afforested plantations studied are shown in Table 2.1. In all plots, the site qualities were higher than in plantations established on former forest soils, probably because of the better quality of the soils (soil depth, stoniness, higher water supply) and prior fertilization. A network of 120 paired plots, made up of former pasture plots and afforested plantations, was established. The plots were distributed in three sets of 40 plots planted with the most commonly used species in the area: Eucalyptus globulus Labill, Eucalyptus nitens (Deane & Maiden) Maiden and Pinus radiata D. Don. Each set was an independent chronosequence in which the range of ages covers the usual rotation lengths applied to these plantations, thus enabling conclusions to be reached as regards the effects of the single-stem rotation following land use change: 1-23 years for E. globulus, 2-18 years for E. nitens and 2-40 years for P. radiata chronosequence respectively. In all cases the prior use was as pasture land, in which low intensive management was applied for at least 25 years (according to landowners), and some of which has recently been afforested. The plots were dominated by a mixture of Lolium multiflorum, Lolium perenne, Trifolium pratense, Trifoliun repens and Dactylis glomerata, although as the time since last perturbation increased, D. glomerata, Agrostis capillaris and Holcus lanatus became more predominant. The pastures are normally harvested for silage 1-2 times a year, and grazed once or twice a year. They are usually CHAPTER II 38 renewed every 8-10 years by rotovating to a depth of 20 cm. In all cases, afforestation was carried out after ripping at 50 cm depth and planting along the row, so that soil disturbance was considered low. No fertilization or weed control was carried out in the plantations. The pastures were selected so that time since last tillage was more than four years. The minimum size of each grassland and afforested plantation was 0.5 ha. To ensure that all sites were similar as regards soil type and land use, selection of the study sites was based on direct observation of the terrain in adjacent pastures, and consultations with landowners. The plantation age was established using an increment borer to sum the ring number for P. radiata and E. nitens, which was clearly visible and easy to assign to yearly growth periods, and considering the planting date declared by the owner to the Forestry Administration for E. globulus. Table 2.1. Main site characteristics (average and standard deviation) of the stands studied. Units E. globulus E. nitens P. radiata Number of stands (n) 40 40 40 Age interval (yr) 1-23 2-18 2-40 Stand density (stems ha-1) 1108 (309) 1011 (258) 1146 (410) Bed rock (Granitic rock /Schist-Slates) (n) 8/32 5/35 25/15 Site Index* (m) 23.3 (6.5) 15.3 (4.4) 24.8 (4.4) Altitude (m) 242 (173) 517 (63) 466 (115) Average annual Temperature (ºC) 13.3 (1.0) 11.6 (0.5) 11.5 (0.8) Accumulated annual precipitation (mm) 1488 (377) 1434 (322) 1213 (219) *Reference ages for site index were 10, 6 and 20 years for E. globulus, E. nitens and P. radiata respectively. Similar forest management regimes, in terms of site preparation, source of seedlings, and pruning and harvesting regimes, were applied in all afforested stands. Stands of E. globulus were located in coastal areas at altitudes below 300 m.a.s.l., whereas E. nitens and P. radiata stands were located in the inner area, generally between 300 and 500 m.a.s.l. As the selected stands were similar in regard to climate, soils, tree species, prior land use and stand management, the only theoretically difference assumed among plots was the age since afforestation. 2.2.2. C determination in above ground tree biomass For determining carbon density (Mg C ha-1) in aboveground biomass, diameters at breast height (to the nearest cm) and total height (to the nearest dm) were measured in all trees in circular plots of radius 10 m. Dry weight of aboveground biomass was estimated using the equations proposed by Merino et al. (2005) for E. globulus, developed in Chapter III for E. nitens, and by Balboa-Murias et al. (2006) for P. radiata chronosequences. The carbon concentrations in INFLUENCE OF TREE SPECIES ON C SEQUESTRATION IN AFFORESTED PASTURES IN A HUMID TEMPERATE REGION 39 each biomass fraction were those reported by the latter authors, except for E. nitens, for which the values reported in Chapter IV were used. 2.2.3. Soil sampling and analysis For sampling the soil (litter and mineral soil to 30 cm depth), a 50 x 50 m plot was selected within each stand, at a distance of more than 30 m from the edge of the stand. Five samples per plot were taken from between tree rows to minimize any disturbance from the site preparation. For sampling the litter layer, 0.3 x 0.3 m squared frames were thrown at random within each plot, on 5 occasions. All aboveground soil litter was collected and dried at 40ºC until constant weight. Carbon contents in the litter were measured for E. nitens (47.9%C), whereas already published data were used for E. globulus and P. radiata (Merino et al., 2005). For mineral soil, three soil layers were considered at depths of 0-5, 5-15 and 15-30 cm. The first two correspond to the A horizon, and the deepest layer A2, AB or BA. Sub-samples of the mineral soil layer were collected with a steel corer, and were combined to form one bulk sample per plot. These samples were oven-dried at 40ºC, sieved at 2 mm and the stoniness was determined. In the fine soil fraction, total C and N were analyzed with a LECO Elemental analyzer, whereas soil particle analysis in the upper 15 cm was performed by laser diffractometry, with a Mastersizer 2000 diffractometer. At each sampling point, five density corers were collected in a 100 cm3 metal cylinder, which was oven-dried at 105ºC and weighed to determine the bulk density following the methodology of Blake and Hartge (1986). The C content in each layer was determined by expression [2.1], where CD is the carbon density in each layer (Mg ha-1), CC is the carbon concentration in each layer (as a fraction of unity), Db is the bulk density (g cm-3), T is the thickness of each layer (cm) and S is the stoniness (as a fraction of unity).   1001       STDbCCCD [2.1] 2.2.4. Evaluation of C sequestration in the forest system The amounts of C in aboveground biomass and organic layers was considered as net gain relative to pastures, and therefore only absolute values are reported or plotted against time since afforestation. For modelling of aboveground biomass C changes with time since afforestation, the Richards (1959) model was used to describe the relationship between aboveground biomass carbon (CW, Mg ha-1) and stand age (t, yr), shown in equation [2.2]. The model for each species was fitted with the MODEL procedure of the SAS/ETS® system (SAS Institute Inc, 2004). The root of mean square errors (RMSE) and adjusted determination coefficient (Adjust. R2) were calculated for each model fit. CHAPTER II 40   2 1 1 0 b tb WebC   [2.2] Carbon sequestration in each mineral soil layer was evaluated in absolute terms as the difference between the forest subplot CDF and the pasture subplot CDP (carbon absolute difference CAD, Mg ha-1), and in relative terms as the difference in percentage of carbon density (carbon relative difference, CRD, % of initial carbon density), with expression [2.3]. Both parameters were represented in each plot against time since afforestation to evaluate the changes in soil carbon with time since afforestation. 100   P PF CDCDCD CRD [2.3] To describe changes in soil C over time, in previous studies on changes in mineral soil carbon after secondary succession, an empirical modelling approach including the gamma function was used successfully (Covington, 1981; Zak et al., 1990). However, in the present study the changes are expected to follow a different pattern, as we hypothesized that the C equilibrium level is different in pasture than in afforested land, and the gamma is function biologically inconsistent in such cases. Alternative models include more parameters than the gamma function, making convergence of the parametric model fit difficult, although this will depend on the amount of data available. In this case, parametric curve fitting procedures did not converge because of the large number of parameters needed to fit a model that adequately captures the apparent shape of the data. Thus, nonparametric fitting was carried out to describe the general trend in the changes in litter and mineral soils. The LOESS procedure in the SAS/STAT (SAS Institute Inc, 2004), was used to divide residuals of the nonparametric curve into ten age intervals, and the 95% confidence levels were determined. Data were also analyzed by analysis of variance with the GLM procedure of SAS/STAT (SAS Institute Inc, 2004). The Tukey test was used to detect differences between means, considered significant p<0.05. 2.3. Results 2.3.1. Changes in the C accumulated in biomass throughout the rotation Results of non linear fit of aboveground biomass C density are shown in Table 2.2. All parameters were significant at p<0.005, and the accuracy of the statistics was adequate. The average changes in the C accumulated in the tree aboveground biomass for the three tree species studied throughout the rotation are shown in Fig. 2.1. The tree growth rates followed the order E. INFLUENCE OF TREE SPECIES ON C SEQUESTRATION IN AFFORESTED PASTURES IN A HUMID TEMPERATE REGION 41 nitens > E. globulus ≥ P. radiata. The three species continued accumulating C at high rates, even after the common rotation lengths in the region (10-15 years for both eucalypts and 30-35 years for P. radiata), which indicates the large potential of the biomass of these species as a C sink. Table 2.2. Aboveground biomass carbon model parameters and fits for age since afforestation. Parameter estimate (Std. error) Specie b0 b1 b2 RMSE Adjust. R2 E. globulus 520.3 (6.23) 0.0589 (0.0021) 2.356 (0.0325) 3.0331 0.718 E. nitens 784.1 (10.41) 0.0393 (0.0013) 1.815 (0.0218) 2.1483 0.835 P. radiata 1569.5 (13.75) 0.0126 (0.0008) 1.466 (0.0201) 4.3544 0.788 Figure 2.1. Changes in total aboveground biomass C throughout the rotation. Continuous lines indicate the fitted Richards model; dotted lines are 95% confidence levels for the mean. E. nitens red; E. globulus green; P. radiata black. 2.3.2. Changes in C accumulated in litter throughout the rotation The changes in the C accumulated in the litter layer (and mineral soil to depth 15 cm) for each of the three species are shown in Fig. 2.2. The average trends throughout the rotation were fitted by the nonparametric procedure described in point 2.2. The three species showed different patterns as regards the dynamics and the amounts of C accumulated at the end of the rotation. Thus, in accordance with the higher growth rates of biomass, litter accumulation occurred earlier (2 yr after establishment) in E. nitens than in P. radiata and E. globulus (4-5 yr after forest establishment). The litter C accumulation rates followed the order: E. nitens > P. radiata > E. globulus. Litter accumulation was lower than expected in E. globulus stands, considering the high aboveground tree growth. The correlations between the changes in crown biomass and litter layer dynamics were different for each species, and were generally rather weak (n.s. for E. globulus; R2= 0.40 for E. nitens, and R2= 0.55 for P. radiata). CHAPTER II 42 Litter accumulation stabilized 10 years after afforestation by both Eucalyptus species, whereas the litter layer continued to increase in mature P. radiata plantations. At the end of the rotation, the average amounts of C in the litter layer ranged from 10.1 (E. globulus, 10 yr), to 24.8 Mg ha-1 (E. nitens, 10 yr) and 50.9 (P. radiata, 35 yr). Figure 2.2. Changes in carbon density accumulation (Mg ha-1) in the litter layer and mineral soils throughout the first rotation after afforestation. Dark shaded area: 95% confidence limits for 0-15 cm depth mineral soil (CAD); light shaded area: 95% confidence limits organic layer (n= 40 for each of the three species). INFLUENCE OF TREE SPECIES ON C SEQUESTRATION IN AFFORESTED PASTURES IN A HUMID TEMPERATE REGION 43 2.3.3. Changes in SOC in the mineral soils after afforestation The changes in SOC density relative to that of the paired pasture (CRD) in the mineral soil for each soil depth layer, considering each species separately and together, are shown in Table 2.3. In the first 10 years after afforestation, losses of C in the 0-15 cm layer were found in all three tree species, ranging between -52.0% in P. radiata to about -0.2% in both species of Eucalyptus in the first 5 years. However, the ANOVA only revealed significant changes in the 0-5 and 5-15 cm soil layers under P. radiata, which emphasizes the high variability in the Eucalyptus stands. In the 11-15 yr period, the SOC contents were similar in afforested and pasture soils for all species. Finally, in the stands older than 20 yr (only for P. radiata), the SOC contents were significantly higher in the uppermost soil layer (0-5 cm) than in the other layers. No such trend was detected in the 15-30 cm layer. Table 2.3. Average values (and standard deviations) of relative difference in carbon density (CRD, %) considering all species jointly (n= 120), and each species separately (n=40). Significant differences for a given soil depth are indicated by different letters. Mineral soil layers (cm) T (years) 0-5 5-15 15-30 0-15 0-30 All species 0-5 -21.1 (26.7) a -6.9 (52.8) 1.3 (42.3) -14.6 (36.8) -9.5 (33.9) 6-10 -6.2 (31.2) ab 3.5 (40.6) 10.2 (45.8) -2.0 (31.7) 1.5 (33.1) 11-15 -3.3 (38.4) ab 8.7 (73.4) 6.4 (64.8) 0.8 (47.9) 2.0 (51.8) 16-20 16.6 (49.7) bc -4.1 (33.8) -3.2 (44.4) 1.2 (28.6) -1.1 (31.0) >21 30.7 (29.7) c 7.8 (29.1) 3.8 (45.9) 15.5 (23.3) 9.9 (27.0) E. globulus 0-5 -4.1 (17.8) 12.3 (43.6) 28.4 (46.4) -0.1 (22.0) 13.4 (37.6) 6-10 -9.4 (34.8) -9.0 (45.1) -4.1 (57.9) -9.2 (39.6) -8.4 (43.6) 11-15 5.4 (45.4) 19.2 (110.4) 4.7 (84.2) 6.8 (66.5) 6.9 (72.7) 16-20 8.7 (30.0) -0.1 (38.7) -2.5 (42.8) 3.2 (24.3) -1.4 (30.2) E. nitens 0-5 -3.1 (18.5) 13 (34.8) 19.6 (37.4) -0.3 (24.6) 8.0 (27.9) 6-10 -6.0 (29.8) 3.4 (41.7) -17.3 (46.8) -1.4 (26.1) -8.6 (31.9) 11-15 -8.9 (22.4) -6.3 (29.5) 5.0 (62.8) -8.9 (21.6) -4.5 (33.7) 16-20 29.9 (71.6) -12.3 (13.8) -28.2 (20.2) 6.6 (39.9) -7.4 (26.0) P. radiata 0-5 -51.3 (18.6) a -52.8 (16.5) a -5.3 (29.5) -52.0 (16.7) a -34.1 (17.3) a 6-10 -23.6 (29.4) ab -23.4 (13.1) ab -4.5 (18.8) -24.1 (10.6) ab -17.1 (9.0) ab 11-15 -14.3 (41.8) ab -16.8 (30.9) ab -13.8 (31.0) -15.6 (32.6) ab -16.0 (29.7) ab 16-20 32.5 (65.2) bc -4.4 (28.8) bc 4.6 (47.3) 4.6 (31.4) bc 3.7 (32.6) ab >21 28.4 (29.3) bc 13.4 (28.7) bc 10.1 (50.5) 18.9 (23.2) cd 14.7 (27.8) b This effect was also observed in relation to the number of plots that are in a certain group of losses or gains with respect to CRD, shown for soil depths of 0-15 and 15-30 cm, and considering all species together (Fig. 2.3). Although the data revealed a high degree of variability, both representations show a general trend for the SOC content in the 0-15 cm soil layer. Carbon was CHAPTER II 44 lost from this soil layer during the first 10 years after afforestation, although gains were observed thereafter. Net positive gains were found from 20 years onwards. 0-5 years 0 2 4 6 8 10 12 14 -80-60-40-20 0 20406080 Carbon content difference (%) Number of plots 6-10 years 0 2 4 6 8 10 12 14 -80 -60 -40 -20 0 20 40 60 80 Carbon content difference (%) Number of plots 11-15 years 0 2 4 6 8 10 12 14 -80-60-40-20020406080 Carbon content difference (5) Number of plots 16-20 years 0 2 4 6 8 10 12 14 -80-60-40-20 0 20406080 Carbon content difference (%) Number of plots >21 years 0 2 4 6 8 10 12 14 -80-60-40-20 0 20406080 Carbon content difference (%) Number of plots 0-15 cm 15-30 cm 0-15 cm 15-30 cm Figure 2.3. Changes in relative difference in carbon density (CRD, %) in each mineral soil layer grouped in age classes of 5 years and CRD classes of 20% for all species considered. The average mineral soil C density in pastures and afforested stands, grouped in age classes of 10 years and for the three mineral soil layers studied, are shown in Fig. 2.4. For direct comparison of the C densities in each soil depth, the value of each soil layer was divided by the corresponding depth (Mg ha-1 cm-1). The average SOC densities in pasture sub-plots were constant, since there were no significant changes in carbon densities over time for any of the species studied (p<0.001). Changes in CRD and CAD were therefore only due to changes in SOC densities in forest subplots, because carbon remained constant for a given soil layer in pasture land. There were no significant differences in the distribution of SOC across soil depth either between species or ages considered (p<0.001). Nevertheless, there were significant differences in P. radiata stands in the upper mineral soil layers, as previously reported. On the other hand, the mean changes in CAD and the 95% confidence levels in the 0-15 cm mineral soil (and in the litter layer) for each of the three species considered in this study throughout their respective rotations are shown in Fig. 2.2. For calculation of the average trends and confidence levels, the nonparametric procedure fitting described above was applied. This type of representation enables consideration of the variability in the data. Significant SOC losses were detected in the first 5-10 years after afforestation in P. radiata stands. The average losses amounted to -10.1 Mg ha-1 (for 95% of confidence level, between -7.1 to -13.0), which constituted an average loss of -24.1% of the initial SOC (for 95% of confidence level, between 18.8-29.4%). There were then large gains in soil C, coinciding with significant accumulation of litter, reflecting a change in the environmental equilibrium between decomposition and production. The average compensation age (the time at which the initial SOC content is recovered) was 20 yr (for 95% of confidence level, the data ranged from 14 to 25 yr) and progressive gains occurred thereafter. INFLUENCE OF TREE SPECIES ON C SEQUESTRATION IN AFFORESTED PASTURES IN A HUMID TEMPERATE REGION 51 not only affect the more labile organic SOM compounds, but also complex and recalcitrant C compounds (Chapter V). Similar patterns to those described above have been described in other studies (Table 2.6). These studies were selected to represent the same land use change as described in the present study, and to provide information enabling estimates in the short and long-term and at compensation age. In the short term (<20 yr), the average losses were similar to those observed in the present study. The average C compensation ages recorded in the present study, between 10 and 25 years, were within the most common range reported in the relevant literature (Table 2.6), although much longer compensation ages (80 yr) have been simulated in colder climates than described here. Moreover, greater (although highly variable) long-term gains than those observed here have been reported, although the time frame considered was longer than in the present study (Table 2.6). The present results show that the subsequent gains in SOC took place after canopy closure in the stands. Significant C gains were only recorded in the 15 cm upper mineral soil layer. However, it is known that root turnover from trees can incorporate organic matter deeper than 30 cm (Brown & Lugo, 1990; Trumbore et al., 1995; Jackson et al., 1996; Jobbágy & Jackson, 2000), although C accumulation in root biomass was not taken into account here. The subsoil horizons were very variable (A2, AB, B), and the SOC contents were different, which may have prevented identification of any clear trends. Furthermore, within the same species, the SOC gains were slightly higher in the stands with higher site indexes, which reflects the influence of the greater biomass production on litter production. Nevertheless, the effect of the site index was not able to be evaluated accurately, since site index was rather high in most cases. This aspect is of interest, as climate change is expected to change site index, and therefore net primary production, worldwide, and the effect on changes in SOC required further investigation. Soil texture is one of the most important factors controlling SOC dynamics, and SOC increases with clay content in afforested soils (Mendham et al., 2003). The present study, however, did not identify any changes in the SOC dynamics attributable to soil texture, probably because the soils were rather homogeneous as regards this parameter. CHAPTER II 52 This study Morris et al. (2007) Thuille & Schulze (2006) Ussiri et al. (2006) Hooker & Compton Vesterdal et al. (2002) Turner & Lambert (2000) Jug et al. (1999)(1) Ross et al. (1999) Richter et al. (1999) Bashkin & Binkley (1998) Giddens et al. (1997) Johnston et al. (1996) Zak et al. (1990) Source Pinus radiata Eucalyptus nitens Eucalyptus globulus Deciduous spp. Coniferous spp. Picea abies Robinia pseudoacacea Casuarina spp. Pinus strobus Quercus robur; Picea Eucalyptus grandis Populus spp; Salix Populus spp; Salix Pinus radiata Pinus taeda Eucalyptus saligna Pinus radiata Five forest types Quercus ellipsoidalis Species P P P A A P P P P P P A,P A, P P A A P A A Origen PL PL PL PL PL PL PL PL S PL PL PL PL PL PL PL PL S S Ref. land use PP, CH PP, CH PP, CH PP PP CH PP PP CH CH PP, CH LT LT PP LT PP PP CH CH Approach 0-15 0-15 0-15 100 100 0-50 0-10 0-10 0-20 0-5 0-10 0-5 0-5 0-10 0-7.5 0-10 0-10 0-10 0-10 Mineral soil dept (cm) 5-13 11 8 - - 15-60 10 10 - 5-10 15 7 10 19 6-10 - - 5-10 8-9 Age (yr) -26.0% (-34.4%--17.0%) -16.7% (-34.7%-+0.2%) -22.2% (-52.0%-+2.7%) - - -30% -11% -16% - -45% -40% +17% +50% -13% -18% - - -10% -31% Effect S. T. EFFECT 17 (14-25)(5) (-21.4%-+27.9%) 16(4) (-15.5%-+18.8%) 12(4) (-15.7%-+21.7%) - - 80(3) - - - - - - - - 16-18 10-13 16-24(2) 20 20 (yr) COMP. 34 18 21 65 65 93-112 - - 115 29 35 - - - 35 - - 40 40 Age (yr) A, agriculture; P, pasture; S, secondary succession, PL, plantation; CH, chronosequence; PP, paired plots; LT, long term study. (1) Fertilized short rotation plantations. (2) For 60% of the studied sites. (3) Obtained from simulation model. (4) Compensation age for the average value; (5) Compensation age fo r the 95% confidence intervals. (6) Mg C ha-1 yr-1. +23.5% (+10.6%-+37.4%) +2.2% (-15.6%-+18.8%) +10.8% (+2.0%-+24.2%) +35.6% +24.7% +0.24-+0.34(6) - - 0% -25% -40% - - - +22% - - +40% +35% Effect L. T. EFFECT Table 2.6. Changes in upper mineral soil carbon after land use change from agriculture to pasture to forest reported in several reference studies. For the values obtained in the present study: short term (S.T.) and long term (L.T.) effects show the average and the range of values provided by LOESS analysis. The range of compensation age (Pinus radiata) and the CRD compensation age (COMP.) are also shown. INFLUENCE OF TREE SPECIES ON C SEQUESTRATION IN AFFORESTED PASTURES IN A HUMID TEMPERATE REGION 53 2.4.4. Influence of species The results of the present study revealed important differences in the SOC dynamics following afforestation, attributable to the tree species and the associated vegetation. Although SOC losses were always recorded after afforestation, in the soils under the two eucalypt species, losses were generally lower and the periods of loss were shorter. In the P. radiata stands there was clear net gain of SOC from 25 years onwards, because of the longer rotation. In both eucalypt stands, the compensation ages were close to the end of the rotation, which implies no net gains prior to cutting in most cases. Longer rotation in eucalypts may lead to positive SOC gains. Moreover, the variability in SOC in the earlier period (0-10 yr) was much higher in both Eucalyptus stands than in P. radiata (Fig. 2.2). This was probably due to the death of weeds and herbaceous species in the forest subplot because of shading, which leads to loss of SOC in all plots, independently of the initial conditions. In the P. radiata plots (Table 2.4) there were no significant differences between the upper and lower part of the CAD curve (Fig. 2.2). The different patterns in the SOC dynamics may be determined by the different SOM dynamics and litter turnover for the three tree species studied. Thus, it is possible that the rapid litter turnover in both types of eucalyptus stands prevented SOC losses in the first years after afforestation, as also suggested by Vesterdal et al. (2008), and Huang et al. (2011). However, the differences in SOC may also be due to the different ground vegetation development in eucalyptus and pine stands, which affects the SOC via different mechanisms (Lugo & Brown, 1993; Silver et al., 2004). In the pine plantations, the wider crown intercepting solar radiation exerts a negative influence on the ground vegetation (very large decreases in the ground vegetation occur from the 5th year (Omil et al., 2007)) and probably also negatively affects litter decomposition. Thus, the higher losses of SOC observed in the young pine plantations may be due to the lower transfer of organic C to the mineral soil, as a consequence of the lower litter inputs from the ground vegetation and the lower decomposition rate of the litter. The opposite occurs in the eucalyptus plantations, in which the higher crown light-transmission favours higher ground vegetation cover throughout the whole rotation, and with a high presence of grass species (González-Hernández et al., 1998; Silva-Pando et al., 2002). The different ground vegetation cover probably determined the amounts and the type of litter (aerial, root) in these plantations. Thus, the presence of grass in these young plantations resulted in greater belowground C inputs such as root biomass turnover and root exudates (Jones et al., 2009). Grass material is incorporated more rapidly than litter layer material into soil organic matter (Andrade et al., 2008; Laungani & Knops, 2009). The lower losses of SOC in eucalypt soils may therefore be due to the higher inputs of litter from grasses in the ground vegetation, thus compensating for the initial losses of SOC following afforestation. Similar mechanisms has also been suggested by Lemma et al. (2006) and Huang et al. (2011) to explain the greater SOC gains in soils afforested with Pinus patula and E. nitens, respectively. CHAPTER II 54 Moreover, the C/N ratios in the mineral soil layer of the mature afforested soils under pines increased throughout the rotation. This effect has also been reported in other studies (Smethurst & Sadanandan Nambiar, 1995; Giddens et al., 1997; Jug et al., 1999; Ussiri et al., 2006), and is probably due the increased influence of forest litter on SOM quality throughout the rotation. This would reflect a shift from organic input dominated by grass litter, to forest litter containing greater amounts of recalcitrant biopolymers (resins, waxes, suberin and cutin-derived compounds (Chefetz et al., 2002; Otto & Simpson, 2006)). The production of recalcitrant compounds from this type of litter and their release to the mineral soil may therefore explain the higher C/N ratio in the soils under mature pine plantations. The higher C/N ratio in the afforested soils may also be due to a lower presence of legumes in the understory vegetation (Corbeels et al., 2002) and to higher N immobilization in trees. However, the C/N ratio was not higher in the eucalyptus stands, possibly because of the presence of more grass in the underground vegetation in these plantations. Soil analyses revealed the presence of more carbohydrates in the SOM in these mature stands, reflecting different sources of litter (possibly due to the input of root litter and root exudates) relative to the mineral soil under pine, in which more recalcitrant compounds were identified (Chapter V). 2.4.5. Increasing the C sink capacity by tree species selection and management The data obtained in the present study show that of the species studied, E. nitens has the highest C sink capacity, followed by E. globulus and, very closely by P. radiata. The mean rates of C sequestration (biomass and soil) estimated in this study for the most common rotations (Table 2.5) ranged between 8.7 and 12.6 Mg C ha-1 yr-1 (average value for the three species, 10.9 Mg C ha-1 yr-1). Considering that the afforested area in northern Spain using these three species can be estimated as 135000 ha for the period 1994-2006 (MAPA, 2006), afforestation would have resulted in a sink of 1.2-1.7 Tg C yr-1 (average 1.5 Tg C yr-1), with respect to the Spanish CO2 emissions (101 Tg C in the year 2009, (MMAMRM, 2010)). This indicates the significant contribution of afforestation to the mitigation of CO2 emissions, and also shows that selection of the tree species is a major factor influencing the C sink capacity. Prolongation of the rotation by 10 and 5 years for Eucalyptus and P. radiata respectively resulted in a C sequestration rate ranging between 10.9 and 14.2 Mg C ha-1 yr-1 (average value for the three species, 12.7 Mg C ha-1 yr-1, Table 2.5), which implies a sink of 1.6-2.3 Tg C yr-1 (average 1.7 Tg C ha-1 yr-1). These and previous results (Balboa-Murias et al., 2006; Diaz-Balteiro et al., 2009) show that in order to maximize the C sink capacity, plantations should be managed according to the optimal harvesting schedules for these species. In the present study, the data show that the C sink capacity of these plantations can be increased greatly by prolonging the rotation age. The selection of tree species and the harvest scheduling may also favour C gains in the soil. The contribution of the soil (litter plus mineral soil) to the overall C sequestration ranged from 8 to INFLUENCE OF TREE SPECIES ON C SEQUESTRATION IN AFFORESTED PASTURES IN A HUMID TEMPERATE REGION 55 18% (in average, 15%), which is similar to the inputs reported by De Vries et al. (2006) and Woodbury et al. (2006), in Europe and United States, respectively and lower than those reported by Liski et al. (2002). Nevertheless, since these intensively managed plantations are harvested repeatedly, the equilibrium will probably be determined by the specific management of each species, particularly by the intensity of the harvesting and disturbance after cutting. Thus, unlike in colder climates (Berg et al., 2009), in these intensively managed forests the C contained in the litter cannot be considered as a medium-term store of stored C because it may be rapidly lost after clear cutting or site preparation. This is particularly important in the two Eucalyptus plantations, in which management in short rotations may lead to continuous loss of SOC. On the contrary, since plantations of E. globulus are coppiced (i.e. the root system is kept without any site preparation), the subsequent rotation will grow faster and SOC will probably continue to increase. Although the site preparation techniques carried out in the region (ripping or holing), do not involve important soil disturbance, logging residues are often removed from the site for energy production. This reduces litter input and nutrient returns (Merino et al., 2005), which may reduce the C sink capacity of both biomass and soil (Maillard et al., 2010). Thus, net losses of soil C have been recorded in soils subjected to high degree of disturbance for site preparation (Pérez-Batallón et al., 2001). On the other hand, the main limiting nutrients in these plantations are P and Mg. Improvement of the nutritional status of the plantation not only implies higher biomass growth rates, but also favours accumulation of SOC (Turner et al., 2005). The application of charcoal-containing wood ash, a by-product of biomass power plants, is increasingly carried out in the area. The application of wood ash replenishes nutrients and enhances timber production (e.g. Solla-Gullon et al., 2008; Pérez-Cruzado et al., 2011). Moreover, charcoal is relatively recalcitrant and can therefore act as a long-term sink for atmospheric CO2 (e.g. Krull et al., 2006). 2.5. Conclusions In this study an intensive sampling scheme was used to assess the C sink capacity of forest stands of the three species most commonly used in afforestation programmes in northern Spain. The high spatial variability in the different compartments illustrates the risk of reaching wrong conclusions about SOC dynamics when the experimental design does not cover most of the variability. The humid temperate climate resulted in C accumulation rates as high as 9-14 Mg C ha-1 yr-1 in the first 20 yr, depending on the species and the rotation length. The results of the study show how selection of the tree species is a major factor influencing the post afforestation C sink capacity, affecting the amounts of C accumulated in both biomass and soil. CHAPTER II 56 The role of the tree species is particularly important during the first years after afforestation when the litter input from herbaceous vegetation may compensate for losses of SOC. The patterns of SOC dynamics differed greatly in relation to the different tree species used in the afforestation, and were determined by transfer of C to the soil via the roots of the ground vegetation and the turnover rate of the litter. Both of these sources were lower in the pine plantations than in the eucalyptus plantations, which may explain the higher SOC during the first years after afforestation. The humid temperate climate, along with the lack of physically protected SOM (sandy loam texture of the soils) favoured important losses of SOC in the uppermost mineral soils during the first years. The study provides accurate information on the success of these afforestation programmes as regards CO2 mitigation. To enhance the C sink capacity, plantations should be managed according to optimal harvesting schedules for the species. Elongation of the rotation length led to larger C sink capacities in all three species. This is especially important in such intensively managed plantations, in which harvesting in short rotations may lead to continuous loss of SOC. 2.6. Bibliography Alvarez, E.; Fernández Marcos, M.L.; Torrado, V. & Fernández Sanjurjo, M.J. 2008. Dynamics of macronutrients during the first stages of litter decomposition from forest species in a temperate area (Galicia, NW Spain). Nutrient Cycling in Agroecosystems. 80 (3): 243-256. Andrade, H.J.; Brook, R. & Ibrahim, M. 2008. Growth, production and carbon sequestration of silvopastoral systems with native timber species in the dry lowlands of Costa Rica. Plant and Soil. 308 (1): 11-22. Balboa-Murias, M.A.; Rodríguez-Soalleiro, R.; Merino, A. & Álvarez-González, J.G. 2006. Temporal variations and distribution of carbon stocks in aboveground biomass of radiata pine and maritime pine pure stands under different silvicultural alternatives. Forest Ecology and Management. 237 (1-3): 29-38. Balesdent, J.; Chenu, C. & Balabane, M. 2000. Relationship of soil organic matter dynamics to physical protection and tillage. Soil and Tillage Research. 53 (3-4): 215-230. Bashkin, M.A. & Binkley, D. 1998. Changes in soil carbon following afforestation in Hawaii. Ecology. 79 (3): 828-833. Bell, M.J.; Harch, G.R. & Bridge, B.J. 1995. Effects of continuous cultivation on Ferrosols in subtropical southeast Queensland. I. Site characterization, crop yields and soil chemical status. Australian Journal of Agricultural Research. 46 (1): 237-253. Berg, B. 2000. Litter decomposition and organic matter turnover in northern forest soils. Forest Ecology and Management. 133 (1-2): 13-22. Berg, B.; Johansson, M.B.; Nilsson, A.; Gundersen, P. & Norell, L. 2009. Sequestration of carbon in the humus layer of Swedish forests—direct measurements. Can J For Res. 39: 962-975. Berthrong, S.T.; Jobbágy, E.G. & Jackson, R.B. 2009. A global meta-analysis of soil exchangeable cations, pH, carbon, and nitrogen with afforestation. Ecological Applications. 19 (8): 22282241. Blake, G.R. & Hartge, K.H. 1986. Bulk density. En: Methods of soil analysis. Ed. A. Klute, 2nd Edn., ASA and SSAA, Madison, WI. 363-375 pp. INFLUENCE OF TREE SPECIES ON C SEQUESTRATION IN AFFORESTED PASTURES IN A HUMID TEMPERATE REGION 57 Brown, S. & Lugo, A.E. 1990. Effects of forest clearing and succession on the carbon and nitrogen content of soils in Puerto Rico and US Virgin Islands. Plant and Soil. 124 (1): 53-64. Carneiro, M.; Serrão, V.; Fabião, A.; Madeira, M.; Balsemão, I. & Hilário, L. 2009. Does harvest residue management influence biomass and nutrient accumulation in understory vegetation of Eucalyptus globulus Labill. plantations in a Mediterranean environment? Forest Ecology and Management. 257 (2): 527-535. Cerli, C.; Celi, L.; Kaiser, K.; Guggenberger, G.; Johansson, M.B.; Cignetti, A. & Zanini, E. 2008. Changes in humic substances along an age sequence of Norway spruce stands planted on former agricultural land. Organic Geochemistry. 39 (9): 1269-1280. Chefetz, B.S.; Salloum, M.J.; Deshmukh, A.P. & Hatcher, P.G. 2002. Structural components of humic acids as determined by chemical modifications and carbon-13 NMR, pyrolysis-, and thermochemolysis-gas chromatography/mass spectrometry. Soil Science Society of America Journal. 66 (4): 1159-1171. Conant, R.T.; Smith, G.R. & Paustian, K. 2003. Spatial variability of soil carbon in forested and cultivated sites: Implications for change detection. Journal of Environmental Quality. 32: 278286. Conteh, A.; Lefroy, R.D.B. & Blair, G.J. 1997. Dynamics of organic matter in soil as determined by variations in 13C/12C isotopic ratios and fractionation by ease of oxidation. Australian Journal of Soil Research. 35 (4): 881-890. Corbeels, M.; O’Connell, A.M.; McMurtrie, R.E.; Grove, T.S. & Mendham, D.S. 2002. Modelling changes in nitrogen mineralisation following conversion of improved pasture to eucalypt plantation. Agronomie. 22 (7-8): 801-815. Corre, M.D.; Schnabel, R.R. & Shaffer, J.A. 1999. Evaluation of soil organic carbon under forests, cool-season and warm-season grasses in the northeastern US. Soil Biology and Biochemistry,. 31 (11): 1531-1539. Covington, W.W. 1981. Changes in forest floor organic matter and nutrient content following clear cutting in northern hardwoods. Ecology. 62 (1): 41-48. Davidson, E.A. & Ackerman, I.L. 1993. Changes in Soil Carbon Inventories Following Cultivation of Previously Untilled Soils. Biogeochemistry. 20 (3): 161-193. De Vries, W.I.M.; Reinds, G.J.; Gundersen, P.E.R. & Sterba, H. 2006. The impact of nitrogen deposition on carbon sequestration in European forests and forest soils. Global Change Biology. 12 (7): 1151-1173. Detwiler, R.P. 1986. Land use change and the global carbon cycle: the role of tropical soils. Biogeochemistry. 2 (1): 67-93. Diaz-Balteiro, L.; Bertomeu, M. & Bertomeu, M. 2009. Optimal harvest scheduling in Eucalyptus plantations: A case study in Galicia (Spain). Forest Policy and Economics. 11 (8): 548-554. Don, A.; Rebmann, C.; Kolle, O.; Scherer-Lorenzen, M. & Schulze, E.D. 2009. Impact of afforestation-associated management changes on the carbon balance of grassland. Global Change Biology. 15 (8): 1990-2002. EU 1992. COUNCIL REGULATION (EEC) Nº 2080/92 of 30 June 1992 Instituting a Comunity aid Scheme for Forestry Measures in Agriculture. Fearnside, P.M. & Barbosa, R.I. 1998. Soil carbon changes from conversion of forest to pasture in Brazilian Amazonia. Forest Ecology and Management. 108 (1-2): 147-166. Fernández-Núñez, E.; Rigueiro-Rodríguez, A. & Mosquera-Losada, M.R. 2010. Carbon allocation dynamics one decade after afforestation with Pinus radiata D. Don and Betula alba L. under two stand densities in NW Spain. Ecological Engineering. 35: 876-890 Giddens, K.M.; Parfitt, R.L. & Percival, H.J. 1997. Comparison of some soil properties under Pinus radiata and improved pasture. New Zealand Journal of Agricultural Research. 40 (3): 409416. CHAPTER II 58 Glaser, B.; Turrion, M.B.; Solomon, D.; Ni, A. & Zech, W. 2000. Soil organic matter quantity and quality in mountain soils of the Alay Range, Kyrgyzia, affected by land use change. Biology and Fertility of Soils. 31 (5): 407-413. Goidts, E.; Van Wesemael, B. & Crucifix, M. 2009. Magnitude and sources of uncertainties in soil organic carbon (SOC) stock assessments at various scales. European Journal of Soil Science. 60 (5): 723-739. Golchin, A.; Oades, J.M.; Skjemstad, J.O. & Clarke, P. 1994. Soil structure and carbon cycling. Australian Journal of Soil Research. 32 (5): 1043-1068. González-Hernández, M.P.; Silva-Pando, F.J. & Jiménez, M.C. 1998. Production patterns of understory layers in several Galician (NW Spain) woodlands: Seasonality, net productivity and renewal rates. Forest Ecology and Management. 109 (1-3): 251-259. Guggenberger, G. & Zech, W. 1999. Soil organic matter composition under primary forest, pasture, and secondary forest succession, Region Huetar Norte, Costa Rica. Forest Ecology and Management. 124 (1): 93-104. Guo, L.B. & Gifford, R.M. 2002. Soil carbon stocks and land use change: a meta analysis. Global Change Biology. 8 (4): 345-360. Guo, L.B.; Wang, M. & Gifford, R.M. 2007. The change of soil carbon stocks and fine root dynamics after land use change from a native pasture to a pine plantation. Plant and Soil. 299 (1): 251-262. Hobbie, S.E.; Ogdahl, M.; Chorover, J.; Chadwick, O.A.; Oleksyn, J.; Zytkowiak, R. & Reich, P.B. 2007. Tree species effects on soil organic matter dynamics: the role of soil cation composition. Ecosystems. 10 (6): 999-1018. Hooker, T.D. & Compton, J.E. 2003. Forest ecosystem carbon and nitrogen accumulation during the first century after agricultural abandonment. Ecological Applications. 13 (2): 299-313. Huang, Z.; Davis, M.R.; Condron, L.M. & Clinton, P.W. 2011. Soil carbon pools, plant biomarkers and mean carbon residence time after afforestation of grassland with three tree species. Soil Biology and Biochemistry. in press. Huggett, R.J. 1998. Soil chronosequences, soil development, and soil evolution: a critical review. Catena. 32 (3-4): 155-172. IPPC 2007. Climate change 2007: Synthesis report. Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Core Writing Team, Pachauri, R.K. and Reisinger, A. (Eds.). http://ipccwg1.ucar.edu/wg1/docs/WG1AR4_SPM_PlenaryApproved.pdf Edn., United Nations Office, Geneva, Switzerland. 104 pp. IUSS Working Group WRB 2006. World reference base for soil resources 2006. World Soil Resources Reports No. 103 FAO, Rome. 128 pp. Jackson, R.B.; Banner, J.L.; Jobbágy, E.G.; Pockman, W.T. & Wall, D.H. 2002. Ecosystem carbon loss with woody plant invasion of grasslands. Nature. 418 (6898): 623-626. Jackson, R.B.; Canadell, J.; Ehleringer, J.R.; Mooney, H.A.; Sala, O.E. & Schulze, E.D. 1996. A global analysis of root distributions for terrestrial biomes. Oecologia. 108 (3): 389-411. Jandl, R.; Lindner, M.; Vesterdal, L.; Bauwens, B.; Baritz, R.; Hagedorn, F.; Johnson, D.W.; Minkkinen, K. & Byrne, K.A. 2007. How strongly can forest management influence soil carbon sequestration? Geoderma. 137 (3-4): 253-268. Jobbágy, E.G. & Jackson, R.B. 2000. The vertical distribution of soil organic carbon and its relation to climate and vegetation. Ecological Applications. 10 (2): 423-436. Johnson, D.W. 1992. Effects of forest management on soil carbon storage. Water, Air, & Soil Pollution. 64 (1): 83-120. INFLUENCE OF TREE SPECIES ON C SEQUESTRATION IN AFFORESTED PASTURES IN A HUMID TEMPERATE REGION 59 Johnston, M.H.; Homann, P.S.; Engstrom, J.K. & Grigal, D.F. 1996. Changes in ecosystem carbon storage over 40 years on an old-field/forest landscape in east-central Minnesota. Forest Ecology and Management. 83 (1-2): 17-26. Jones, D.L.; Nguyen, C. & Finlay, R.D. 2009. Carbon flow in the rhizosphere: carbon trading at the soil-root interface. Plant and Soil. 321 (1): 5-33. Judd, T.S.; Attiwill, P.M. & Adams, M.A. 1996. Nutrient concentrations in Eucalyptus: a synthesis in relation to differences between taxa, sites and components. En: Nutrition of Eucalyptus. Eds. P.M. Attiwill & M.A. Adams, CISRO Publishing, Collingwood, Australia. 123-154 pp. Jug, A.; Makeschin, F.; Rehfuess, K.E. & Hofmann-Schielle, C. 1999. Short-rotation plantations of balsam poplars, aspen and willows on former arable land in the Federal Republic of Germany. III. Soil ecological effects. Forest Ecology and Management. 121 (1-2): 85-99. Kanerva, S. & Smolander, A. 2007. Microbial activities in forest floor layers under silver birch, Norway spruce and Scots pine. Soil Biology and Biochemistry. 39 (7): 1459-1467. Kasel, S.; Singh, S.; Sanders, G.J. & Bennett, L.T. 2011. Species-specific effects of native trees on soil organic carbon in biodiverse plantings across north-central Victoria, Australia. Geoderma. 161 (1-2): 95-106. Kirschbaum, M.U.F.; Guo, L.B. & Gifford, R.M. 2008. Why does rainfall affect the trend in soil carbon after converting pastures to forests? A possible explanation based on nitrogen dynamics. Forest Ecology and Management. 255 (7): 2990-3000. Krull, E.S.; Swanston, C.W.; Skjemstad, J.O. & McGowan, J.A. 2006. Importance of charcoal in determining the age and chemistry of organic carbon in surface soils. Journal of Geophysical Research. 111 (G04001): 1-9. Laganière, J.; Angers, D.A. & Paré, D. 2010. Carbon accumulation in agricultural soils after afforestation: a meta‐analysis. Global Change Biology. 16 (1): 439-453. Laungani, R. & Knops, J.M.H. 2009. The impact of co‐occurring tree and grassland species on carbon sequestration and potential biofuel production. GCB Bioenergy. 1 (6): 392-403. Leirós, M.C.; Trasar-Cepeda, C.; Seoane, S. & Gil-Sotres, F. 2000. Biochemical properties of acid soils under climax vegetation (Atlantic oakwood) in an area of the European temperate-humid zone (Galicia, NW Spain): general parameters. Soil Biology and Biochemistry. 32 (6): 733745. Lemma, B.; Kleja, D.B.; Nilsson, I. & Olsson, M. 2006. Soil carbon sequestration under different exotic tree species in the southwestern highlands of Ethiopia. Geoderma. 136 (3-4): 886-898. Lemma, B.; Nilsson, I.; Kleja, D.B.; Olsson, M. & Knicker, H. 2007. Decomposition and substrate quality of leaf litters and fine roots from three exotic plantations and a native forest in the southwestern highlands of Ethiopia. Soil Biology and Biochemistry. 39 (9): 2317-2328. Liski, J.; Perruchoud, D. & Karjalainen, T. 2002. Increasing carbon stocks in the forest soils of western Europe. Forest Ecology and Management. 169 (1-2): 159-175. Lugo, A.E. & Brown, S. 1993. Management of tropical soils as sinks or sources of atmospheric carbon. Plant and Soil. 149 (1): 27-41. Maillard, É.P.; Munson, D. & Alison, D. 2010. Soil carbon stocks and carbon stability in a twentyyear-old temperate plantation. Soil Science Society of America Journal. 74 (5): 1775-1785. Mann, L.K. 1986. Changes in soil carbon storage after cultivation. Soil Science. 142 (5): 279-288. MAPA 2006. Forestación de Tierras Agrícolas: Análisis de su evolución y contribución a la fijación de carbono y al uso racional de la tierra Spanish Ministry of Agriculture, Fisheries and Food, Madrid (Spain). Marín-Spiotta, E.; Silver, W.L.; Swanston, C.W. & Ostertag, R. 2009. Soil organic matter dynamics during 80 years of reforestation of tropical pastures. Global Change Biology. 15 (6): 15841597. CHAPTER II 60 Martius, C.; Höfer, H.; Garcia, M.V.B.; Römbke, J. & Hanagarth, W. 2004b. Litter fall, litter stocks and decomposition rates in rainforest and agroforestry sites in central Amazonia. Nutrient Cycling in Agroecosystems. 68 (2): 137-154. Martius, C.; Höfer, H.; Garcia, M.V.B.; Römbke, J.; Förster, B. & Hanagarth, W. 2004a. Microclimate in agroforestry systems in central Amazonia: does canopy closure matter to soil organisms? Agroforestry Systems. 60 (3): 291-304. Mendham, D.S.; O'Connell, A.M. & Grove, T.S. 2003. Change in soil carbon after land clearing or afforestation in highly weathered lateritic and sandy soils of south-western Australia. Agriculture, Ecosystems & Environment. 95 (1): 143-156. Merino, A.; Balboa, M.A.; Rodríguez-Soalleiro, R. & González, J.G. 2005. Nutrient exports under different harvesting regimes in fast-growing forest plantations in southern Europe. Forest Ecology and Management. 207 (3): 325-339. Merino, A.; Fernández-López, A.; Solla-Gullón, F. & Edeso, J.M. 2004. Soil changes and tree growth in intensively managed Pinus radiata in northern Spain. Forest Ecology and Management. 196 (2-3): 393-404. Merino, A.; López, Á.R.; Brañas, J. & Rodríguez-Soalleiro, R. 2003. Nutrition and growth in newly established plantations of Eucalyptus globulus in northwestern Spain. Annals of Forest Science. 60 (6): 509-517. MMAMRM 2010. Inventario de gases de efecto invernadero de España. Sumario de resultados Ministerio de Medio Abiente, Medio Rural y Marino. Secretaría de Estado de Cambio Climático, Madrid (Spain). Morris, S.J.; Bohm, S.; Haile‐Mariam, S. & Paul, E.A. 2007. Evaluation of carbon accrual in afforested agricultural soils. Global Change Biology. 13 (6): 1145-1156. Murty, D.; Kirschbaum, M.U.F.; Mcmurtrie, R.E. & Mcgilvray, H. 2002. Does conversion of forest to agricultural land change soil carbon and nitrogen? A review of the literature. Global Change Biology. 8 (2): 105-123. Neill, C.; Melillo, J.M.; Steudler, P.A.; Cerri, C.C.; de Moraes, J.F.L.; Piccolo, M.C. & Brito, M. 1997. Soil carbon and nitrogen stocks following forest clearing for pasture in the southwestern Brazilian Amazon. Ecological Applications. 7 (4): 1216-1225. Nilsson, S. & Schopfhauser, W. 1995. The carbon-sequestration potential of a global afforestation program. Climatic Change. 30 (3): 267-293. Oades, J.M. 1988. The retention of organic matter in soils. Biogeochemistry. 5 (1): 35-70. Omil, B.; Mosquera-Losada, R. & Merino, A. 2007. Responses of a Non N-Limited Forest Plantation to the Application of Alkaline-Stabilized Dewatered Dairy Factory Sludge. Journal of environmental quality. 36 (6): 1765-1774. Ostertag, R.; Marín-Spiotta, E.; Silver, W.L. & Schulten, J. 2008. Litterfall and decomposition in relation to soil carbon pools along a secondary forest chronosequence in Puerto Rico. Ecosystems. 11 (5): 701-714. Otto, A. & Simpson, M.J. 2006. Sources and composition of hydrolysable aliphatic lipids and phenols in soils from western Canada. Organic Geochemistry. 37 (4): 385-407. Paul, K.I. & Polglase, P.J. 2004. Prediction of decomposition of litter under eucalypts and pines using the FullCAM model. Forest Ecology and Management. 191 (1-3): 73-92. Paul, K.I.; Polglase, P.J.; Nyakuengama, J.G. & Khanna, P.K. 2002. Change in soil carbon following afforestation. Forest Ecology and Management. 168 (1-3): 241-257. Paustian, K.; Six, J.; Elliott, E.T. & Hunt, H.W. 2000. Management options for reducing CO2 emissions from agricultural soils. Biogeochemistry. 48 (1): 147-163. Pérez-Batallón, P.; Ouro, G.; Macías, F. & Merino, A. 2001. Initial mineralization of organic matter in a forest plantation soil following different logging residue management techniques. Annals of Forest Science. 58: 807-818. IMPROVEMENT IN ACCURACY OF ABOVEGROUND BIOMASS ESTIMATION IN EUCALYPTUS NITENS PLANTATIONS: EFFECT OF BOLE SAMPLING INTENSITY AND EXPLANATORY VARIABLES 67 probability proportional to a given dimension. This was the case for Kleinn and Pelz, (1987), who chose disks with a probability of selection proportional to estimated volume. On the other hand, the PW method is preferred for large trees in sites with difficult access, as fresh weighing of the whole stem is quite laborious and time consuming (Snowdon et al., 2000). In the CW method, the distribution of moisture along the stem is the main source of error for dry weight estimation, whereas in the PW method, it is the variation in basic density along the bole height that affects that error. In both cases, sampling intensity and distribution should guarantee a suitable description of the variability in moisture content and specific density. Diameter at breast height (d) and total height (h) are the most common independent variables used in biomass regression, because of their ease of measurement and predictive capacity (Parresol, 1999; Snowdon et al., 2000). However, because of the current increasing interest in obtaining accurate predictions of crown fractions for bioenergy, nutrient stability and silvicultural or ecological studies, there is a corresponding increasing interest in crown biomass modelling. Some authors have observed that the use of crown variables as explanatory variables improves the accuracy of biomass equations (Satoo & Madgwick, 1982; António et al., 2007). In biomass studies in which high precision is required for crown fractions, and destructive sampling cannot be applied, highly accurate models are required. The objectives of the present study were: (i) to obtain biomass estimation tools for a fast growing species, Eucalyptus nitens, in northwestern Spain, considering the most complete set of aboveground components; (ii) to evaluate the bias and accuracy of wood biomass estimation for different intensities of systematic subsampling across the stem and two ratio-type estimators (dry/fresh weight and dry mass/fresh volume), (iii) to evaluate the increased accuracy derived from the inclusion of crown variables in the estimation of individual tree biomass components, and (iv) to evaluate the ability of the proposed equations to estimate the proportion of each biomass component over total aboveground biomass, for a range of diameter classes. 3.2. Material and methods 3.2.1. Study site and trees sampled This study was carried out in northwestern Spain, in an inland area located at elevations of 500 to 1000 m.a.s.l., with average precipitation of 900-1200 mm and average annual temperature of 12-13ºC (Martínez Cortizas & Pérez Alberti, 1999). Although frost occurrence limits planting of the most common Eucalyptus species in Spain (Eucalyptus globulus Labill.), Eucalyptus nitens (Deane & Maiden) Maiden was successfully introduced in the mid 1990s, providing yields of 15-50 m3 ha-1 yr-1 (Pérez-Cruzado, 2009). CHAPTER III 68 As the aim of the present study was to construct biomass models that are as representative as possible, sampling consisted of two phases: (1) study of the variability of the most commonly used independent variables in biomass equations at tree level (d and h, see below) across the distribution area, and (2) destructive sampling of trees covering the observed range (Parresol, 1999). For this purpose, 76 plots were established (see location in Fig. 3.1), covering the observed range of ages and site qualities, with a minimum plot size of 314 m2, which is generally suitable for biomass estimation procedures in plantations (Satoo & Madgwick, 1982). Figure 3.1. Location of the measured plots (dots) and the distribution of Eucalyptus nitens in north-western Spain (shaded area). A sample size of 40 trees was chosen because of the low variability in site conditions and densities of plantations, most of which were established with the MacAlister provenance. The sampled trees were chosen in two steps, two trees per diameter and height class were first selected, and 16 additional trees were then chosen, considering the relative importance of each diameter class in the population. The aim of this procedure was to cover the full range of tree size, which is shown for height and diameter in Fig. 3.2. Trees were felled in 12 plots, in which the values of the quadratic mean diameter and d of the trees sampled was similar; undamaged, healthy trees that represented the dominant and codominant strata, were chosen. The average standard deviation and range of representative stand and single tree variables, for both the population and the sample are shown in Table 3.1. The variability in crown variables was similar to that observed in stem variables, unlike in other studies (Satoo & Madgwick, 1982). IMPROVEMENT IN ACCURACY OF ABOVEGROUND BIOMASS ESTIMATION IN EUCALYPTUS NITENS PLANTATIONS: EFFECT OF BOLE SAMPLING INTENSITY AND EXPLANATORY VARIABLES 69 0 5 10 15 20 25 30 35 40 45 50 55 0 5 10 15 20 25 30 35 40 45 50 55 60 65 d (cm) h (m) c Figure 3.2. Height-diameter distribution of Eucalyptus nitens in an initial inventory in north-western Spain. Table 3.1. Statistics for stand and single tree variables in the population (76 plots, 3864 trees) and the sample plots (12 plots, 40 trees). Stand variables Individual tree variables SI (m) N (stems ha-1) Age (yr) d (cm) h (m) All plots Average (Std. dev.) 15.3 (4.4) 1089 (280) 9.5 (4.2) 18.5 (7.5) 20.2 (6.4) Range 8.8 - 20.8 446 - 1560 2 - 18 1.0 - 59.6 2.2 - 48.3 Sample plots Average (Std. dev.) 15.7 (2.7) 1101 (223) 10.2 (2.8) 19.5 (7.7) 19.3 (5.5) Range 9.8 - 18.9 446 - 1401 2 - 13 1.1 - 47.0 2.4 - 35.1 where SI is the site index (m at reference age of 6 years); N is stand density (stems ha-1), d is diameter at breast height (cm), and h is the total height (m). The following variables were measured in the sample trees while still standing: diameter at breast height (d, cm) and stump diameter at 0.15 m (dst, cm), both measured in two perpendicular directions to the nearest mm; total height (h, m) and live crown base height, defined as the height of the first live branch insertion in the stem (hcb, m), both measured to the nearest dm; crown diameter (dc, m) measured in two perpendicular directions following the cardinal points to the nearest cm. Living crown length (hc, m) was estimated as difference between total height (h) and live crown basis height (hcb, m). Crown volume (vc) was calculated from hc and dc by assimilating the crown shape to an ellipsoid [3.1]. Descriptive statistics for these variables are shown in Table 3.2.             223 42 cc c hd v  [3.1] CHAPTER III 70 3.2.2. Ratio type estimators and subsampling The felled trees were cut into 0.5 m logs to a small-end diameter of 7 cm. The logs were weighed fresh and a systematic subsample of one 5 cm-disk in the bottom part of each log was taken, also considering a further disk at the top of the stem. Sample disks were weighed fresh and transported to the laboratory in plastic bags. The over and under-bark diameters of the disks were measured in two directions and the bark and wood were then separated and weighed. For each disk, the dry wood weight was measured after oven drying at 105ºC to constant weight and the ratio of the dry/fresh weight of the wood was determined. Only one composite sample per tree was considered for the bark. Fresh bark of all disks was weighed jointly, and dried to determine dry bark weight, thus enabling the ratio of dry/fresh weight of bark to be obtained for each tree. Table 3.2. Descriptive statistics of sampled trees. Variable Average Maximum Minimum St. Dev. Independent variables d (cm) 20.84 41.55 3.95 10.04 dst (cm) 25.63 52.40 6.60 12.13 h (m) 19.94 30.80 4.40 7.27 hcb (m) 12.55 20.60 2.80 4.68 hc (m) 7.39 19.80 1.20 4.21 dc (cm) 3.50 8.55 1.25 1.66 vc (m3) 81.78 566.5 1.00 129.3 Dependent variables (kg tree-1) Wl 10.73 48.85 0.28 12.94 Wt 4.15 23.33 0.18 5.10 Wtb 4.40 18.46 0.04 4.53 WTb 13.57 75.65 1.29 18.71 Ww 168.43 599.5 0 176.9 Wb 24.59 111.3 0 28.33 Wdb 11.29 68.28 0.03 13.64 Wtot 237.2 838.2 2.54 248.1 Definitions of independent and dependent variables are given in section 3.2.2. Wtot refers to total aboveground biomass. The dry weight of wood and of bark in each log was calculated from the average ratios calculated for the delimiting disks. The total wood (Ww, to a small-end diameter over bark of 7cm) and bark (Wb, evaluated till the threshold diameter considered for wood) dry biomass in each tree was calculated as sum of the biomass of each log. Four biomass fractions were considered for the crown: thick branches (WTb, diameters over bark 2-7cm), which also include the tops of the boles, thin branches (Wtb, diameters over bark 0.52cm), twigs (Wt, diameter less than 0.5 cm) and leaves (Wl). Dead branches in the stem (Wdb) is IMPROVEMENT IN ACCURACY OF ABOVEGROUND BIOMASS ESTIMATION IN EUCALYPTUS NITENS PLANTATIONS: EFFECT OF BOLE SAMPLING INTENSITY AND EXPLANATORY VARIABLES 71 also an important fraction in E. nitens. Crown biomass was first fractioned in the field into three groups: WTb, Wdb and the sum of Wtb, Wt and Wl, and then weighed fresh, with a balance, to the nearest 10 g. A subsample of 10-15% of fresh weight of each fraction was taken to represent the top, medium and bottom part of the crown. These subsamples were weighed in the field, with scales, to the nearest 0.01 g. The composite subsample of Wtb, Wt and Wl, was fractioned and weighed in the laboratory and the proportion of each fraction was determined to enable estimation of the fresh weight of each crown fraction. The dry weight of each fraction was then estimated from the dry/fresh weight ratios. 3.2.3. Methodologies for bole mass estimation The information obtained enabled comparison of two methods of estimating bole mass or weight at a range of sampling intensities. The CW method consisted of determining the complete stem weight and estimating dry mass from disks. Disk subsampling intensity was modified considering a series of inter-disk distances which were multiples of 0.5. For each inter-disk distance tested, there were several solutions, depending on the height of the first section considered. For the logs between two disks, the dry weight estimation was calculated from the average dry weight wood ratio of each disk, and for basal and terminal logs the disks immediately above or below the log were considered. For the PW method, it was considered that only one part of the stem was weighed, and for the rest of the tree the volume was calculated from diameter under bark measured every 0.5 m along the stem and by use of the Smalian formula. The length of the weighed and cubed log was made to range between 0.5 m and the total stem height (up to a small-end diameter of 7 cm), considering a variable position of the log along the stem. The fresh weight of the log was transformed to dry weight by considering the moisture content derived from the whole set of disks taken each 0.5 m. Volume to dry weight ratios were then used to estimate the total dry mass of the stem by multiplying by the calculated volumes. Both methods and sampling intensities were compared with the results obtained by the CW method and disk equidistance of 0.5 m, considering the relative difference in the biomass estimation for each tree [3.2].   100  W WW ˆ RD [3.2] where Ŵ is the predicted bole mass value with each sampling methodology and intensity. The combinations of inter-disk distances, weighed log lengths and starting point along the bole provided a relative difference value, and these were plotted against subsampling intensity for different diameter classes. The 95% confidence intervals were obtained considering a normal CHAPTER III 72 distribution for different classes of sampling intensity. The default consideration of the bottom disk (CW method) or the bottom log (PW method) was considered separately for comparison. 3.2.4. Models and fit Models for predicting biomass of tree components are usually based on the allometric relationship [3.3] between tree biomass and tree variables. This was then used as the basic form of the models to be fitted (Zianis & Mencuccini, 2004). 1n2 b n b 11i ·...·x·xbW   [3.3] where Wi is the dry mass biomass of the fraction i and xn are the independent variables. The model fitting was carried out in two steps. First, each biomass fraction was fitted individually considering each independent variable and their combinations, by use of the minimum generalized squares in the MODEL procedure of SAS/STAT® (SAS Institute Inc, 2004). As initial parameters in the iteration process, a previous linear fit was carried out for all combinations of variables, with the linearized allometric model (Parresol, 2001), by use of the REG procedure of SAS/STAT®. In selecting the best model for each family of equations, the following statistics were calculated for each equation: bias (MRES, [3.4]), root mean square error (RMSE, [3.5]) and adjusted determination coefficient (R2Adj., [3.6]).  N W ˆ W RESM N 1i ii     [3.4]  pN W ˆ W RMSE N 1i 2 ii     [3.5]                               pN 1N · WW W ˆ W Adj.R N 1i 2 ii N 1i 2 ii 2 [3.6] where N is the number of data used in the fitting, p is the number of parameters to be estimated, i W is the average value of the dependent variable. In the second step, each family of equations was fitted simultaneously by the seemingly unrelated regressions method (SUR) to guarantee the additivity of the system (Parresol, 2001). IMPROVEMENT IN ACCURACY OF ABOVEGROUND BIOMASS ESTIMATION IN EUCALYPTUS NITENS PLANTATIONS: EFFECT OF BOLE SAMPLING INTENSITY AND EXPLANATORY VARIABLES 73 This method is based on the fit of an apparently non related equation system formed by the regression functions of the biomass fractions considered and the total biomass. The SUR method iteratively forces the sum of the components to equal the equation for total biomass, ensuring that the global solution is the best possible, although the solution for each k fraction is not necessarily the best. The independent variables in the system of equations for biomass components has to be the same as in the total biomass equation (Parresol, 2001), and in fact this equation was expressed as the sum of each component equation to ensure the additivity of the system (ÁlvarezGonzález et al., 2007). The MODEL procedure of SAS was applied to obtain the SUR estimates, considering the parameters obtained in the individual fitting as initializers. Lack of homogeneity in error variance, or heteroscedasticity, is commonly observed in biomass equations (Parresol, 1993; 2001). As with the standard errors of the parameter estimates, heteroscedasticity was detected by representing the studentized residuals against the real values, and the White (1980) and Breusch & Pagan (1979) tests were applied. Heteroscedasticity was corrected by weighted fitting (Schaegel, 1982; Clutter et al., 1983; Cunia, 1987; Parresol, 1999; 2001), by use of the inverse of the variance of the residuals (σi2) assigned at each observation as a weighting factor, and use of the potential expression [3.7] (Neter et al., 1989). k i 2 ixσ [3.7] The value of the k exponent can be calculated by the optimization method proposed by Harvey (1976), which consists of using the model errors fitted without weights ( i e ˆ) as dependent variable in the potential variance error model (Álvarez-González et al., 2007), the linearized form of which is shown in expression [3.8]. )k·ln(xa)e ˆ ln( i 2 i [3.8] For each fraction, the value of the k exponent was determined for the independent variables or the combination of these that provided the best fit. In those cases in which the statistics did not detect heteroscedasticity, the weighted fit was carried out anyway. The values of the k exponents were added to the fitting program in SAS/STAT® (SAS Institute Inc, 2004). After fitting, the models were again subjected to heteroscedasticity tests to verify their correctness. 3.3. Results 3.3.1. Estimation of bole biomass through systematic subsamplig The relative difference (RD) obtained by the CW method was plotted against systematic subsampling intensity (disks m-1 of stem), for three dimensional classes: DC1 (d<14 cm), DC2 CHAPTER III 74 (14<d<24 cm) and DC3 (d>24 cm) (Fig. 3.3). The figure shows overestimates for all the data, with a clear trend for the relative error to decrease as sampling intensity increased. Small trees are clearly biased towards overestimation (d<14 cm), but the tendency for overestimation decreases greatly with increasing tree size. A threshold of 5% relative error would mean a minimum subsampling intensity of 0.95 disks m-1 (DC1), 0.8 disks m-1 (DC2) or 0.75 disks m-1 (DC3). These values were respectively 0.7, 0.4 and 0.3 disks m-1, for a relative error threshold set at 10%. It is important to note that the default consideration of the stem bottom as the position of the first disk would mean a systematic tendency to overestimation. Figure 3.3. Relative difference for three dimensional classes: DC1 (d<14cm; n = 6 922), DC2 (14<d<24cm; n = 17075) and DC3 (d>24cm; n = 18360), plotted against sampling intensity (disks per stem meter) for the CW method. Continuous black line: average value for all data; dotted black lines: 95% confidence intervals for all data; continuous grey line: average value for alternatives that include bottom log. n is the number of simulated alternatives for each dimensional class. The RD values obtained for the PW method were plotted against the subsampling intensity, expressed as the percentage of stem height that was weighed (Fig. 3.4). Only the higher diameter classes were considered in this case, as there is no reason to avoid obtaining the complete fresh weight of small trees. In this case there was again a clear tendency for the weights to be overestimated, which was even clearer for the highest diameter class. This procedure ensured a maximum relative error of 5%, only when 90% of the stem was weighed for both dimensional classes considered. This percentage was 55% when the threshold of relative error was set at 10%. IMPROVEMENT IN ACCURACY OF ABOVEGROUND BIOMASS ESTIMATION IN EUCALYPTUS NITENS PLANTATIONS: EFFECT OF BOLE SAMPLING INTENSITY AND EXPLANATORY VARIABLES 75 In this case, consideration of the bottom log as the one that should be weighed systematically led to underestimation. Figure 3.4. Relative difference for two dimensional classes: DC2 (14<d<24 cm; n = 7712) and DC3 (d>24cm; n = 12001), plotted against sampling intensity (fraction of stem height weighed) for the PW method. Continuous black line: average value for all data; dotted black lines: 95% confidence intervals for all data; continuous grey line: average value for alternatives that include bottom log. n is the number of simulated alternatives for each dimensional class. The differences in the two methods arise from the observed trends in moisture content and basic density along the stem. The moisture content (wet basis) in relation to the relative height along the stem indicates an increasing trend that is more marked at the bottom 20%, where around 40% of bole dry matter occurs (Fig. 3.5). The increasing trend of basic density (ρ, kg m-3) along the stem, which can easily explain the underestimations derived from the default use of the bottom log in the PW method, is shown in Fig. 3.6. 0 10 20 30 40 50 60 0 102030405060708090100 Relative stem height (%) Moisture (%) 0 10 20 30 40 50 60 70 80 90 100 Acumulated dry weight (%) Figure 3.5. Changes in moisture content (%) and accumulated stem dry weight (%) plotted against relative height along the stem (until 7 cm of diameter over bark). Continuous line: average value; dotted lines: 95% confidence intervals. Such increasing trends, which have been described for different species of eucalypts, with some exceptions, such as for Eucalyptus regnans, indicate the possibility of studying the relative CHAPTER III 76 height at which the average basic density can be found. This value was plotted for the sampled trees with non-zero amounts of wood (Fig. 3.7). The relative height tended to decrease with increasing breast height diameter. These results are of great interest for defining the height along the stem that should be sampled to obtain a good estimate of basic density. 0 100 200 300 400 500 600 700 800 900 1000 0 102030405060708090100 Relative stem height (%) Basic density (Kg m -3 ) 0 10 20 30 40 50 60 70 80 90 100 Acumulated fresh volume (%) Figure 3.6. Changes in basic density (ρ, kg m-3) and accumulated stem fresh volume (%) plotted against relative stem height (until 7 cm of diameter over bark). Continuous line: average value; dotted lines: 95% confidence intervals. For the PW method, it should be considered that the errors would be cumulative if transformation of fresh weight of the log to dry weight is carried out after obtaining information derived from disks with equidistance greater than 0.5 m. Moreover, the volumes were calculated considering 0.5 m logs, which could provide volume estimates close to those obtained from water displacement of fresh samples (Brown et al., 1995). 0 5 10 15 20 25 30 35 40 45 50 0 5 10 15 20 25 30 35 40 45 d(cm) Relative stem height (%) for average ρ Figure 3.7. Changes in relative stem height (%) where composite average basic density (ρ) is found in relation to breast height diameter of sampled trees. Continuous line: average value; dotted lines: 95% confidence intervals. IMPROVEMENT IN ACCURACY OF ABOVEGROUND BIOMASS ESTIMATION IN EUCALYPTUS NITENS PLANTATIONS: EFFECT OF BOLE SAMPLING INTENSITY AND EXPLANATORY VARIABLES 83 percentages shown for diameters less than 12 cm. As a result, the equations presented here would provide reasonable estimates of biomass component percentages for diameters larger than 12 cm. For smaller diameters, exclusive use of the equation predicting total biomass is recommended. 0 10 20 30 40 50 60 70 80 90 0 10203040 d (cm) Proportion of aboveground biomass (%) Ww WTb Pred. Ww Pred, WTb 0 2 4 6 8 10 12 14 16 0 10203040 d (cm) Proportion of aboveground biomass (%) Wdb Wtb Pred. Wdb Pred. Wtb 0 5 10 15 20 25 010203040 d (cm) Proportion of aboveground biomass (%) Wb Pred. Wb 0 5 10 15 20 25 30 35 40 45 0 10203040 d (cm) Proportion of aboveground biomass (%) Wl Wt Pred. Wl Pred. Wt Figure 3.11. Proportion of each biomass fraction over total aboveground biomass. Open figures: trees used in developing biomass equations; filled figures: additional small trees; lines: prediction of biomass equations. 3.4. Discussion 3.4.1. Stem biomass estimation The results of this study show that the error may be important, and will depend on the intensity of subsampling, when ratio-type estimators are used to estimate dry mass. The error also depends on the method used (complete fresh weight or partial fresh weight) and on the average tree size. Other authors have observed that ratio-type estimators provide biased estimates (Cunia, 1979; Valentine et al., 1984; Briggs et al., 1987). These overestimates are as large as the decreases in both subsampling intensity and average tree size. Overestimation is clearly a more serious error than underestimation (Satoo & Madgwick, 1982) because it does not err on the side of safety i.e. for carbon accounting procedures. CHAPTER III 84 Wood moisture content and basic density change along the stem (Satoo & Madgwick, 1982) (Figs. 3.5 and 3.6), and affect the estimation of dry biomass by the CW and PW methods respectively. Minimum moisture content and basic density occur in the basal part of the stem, which is obviously where most of the accumulated weight and volume occur. This effect must therefore be taken into account with a sufficient and well distributed number of subsamples along the stem. One way of addressing this problem, when taper functions are available, is the density integral approach (Parresol & Thomas, 1989). The weighed average is an alternative method that gives more importance to those observations in the lower part of the stem, and therefore more closely related to volume. Chave et al. (2001) reported that the biomass values of the smallest trees strongly affect the values of the model parameters in the allometric relation. This effect is even stronger when a weighted adjustment methodology is used, because the smallest trees, which are less variable, are more important than the largest trees because of heteroscedasticity correction. It is therefore advisable to obtain the complete dry weight of the stem of small trees. The degree of accuracy required depends on the objective of the estimation, although equilibrium between sampling intensity and the level of precision must be ensured (Brown et al., 1995). If ratio-type estimators are chosen for stem dry biomass estimation, a relatively intensive subsampling scheme should be implemented, as others authors indicated for both ratio-type and density-integral methods (Parresol, 1999). Comparing the methods considered here, the CW method produced better results for the largest dimensional class than the PW method (Figs. 3.3 and 3.4). This is because, for a given length of cubed and weighed part, the proportion over total stem (as an indicator of sampling intensity) differs depending on tree size, and therefore becomes less important as tree size increases. This must be taken into account because the PW method is usually used for large trees in which complete weighing is time-consuming. The results clearly show the trends in relative errors derived from a default consideration of the bottom disk or the bottom log as the first section to measure. It is advisable, if systematic sampling is to be used, to establish the subsampling intensity before randomizing the position along the stem of the first disk or log to be measured. In the case of the PW method it is not recommended to take only one sample log per tree, although this was the approach used in this study. The subsampling intensity should be split along the stem, and a good representation of the bole area where average basic density is likely to be found is advisable. Most published papers do not provide information about the proportion of weighed and cubed logs or their distribution along the stem, although the most reasonable distribution would be systematic or random, with the subsampling intensity chosen on the basis of statistical criteria. 3.4.2. Biomass equations from stem and crown variables Although it is known that d, h and W are closely related (Satoo & Madgwick, 1982), h is not always included in biomass equations together with d because both are correlated and inclusion of IMPROVEMENT IN ACCURACY OF ABOVEGROUND BIOMASS ESTIMATION IN EUCALYPTUS NITENS PLANTATIONS: EFFECT OF BOLE SAMPLING INTENSITY AND EXPLANATORY VARIABLES 85 h adds only negligible accuracy (Jokela et al., 1986; Ter-Mikaelian & Korzukhin, 1997; Johansson, 1999; Verwijst & Telenius, 1999; Snowdon et al., 2000; Brown, 2002; Porté et al., 2002; Jenkins et al., 2003). In this study, inclusion of h together with d only resulted in improved accuracy in the case of wood, although other authors have reported significant improvement for several fractions (Loomis et al., 1966; Pearson et al., 1984; Bartelink, 1996; Reed & Tomé, 1998; Monserud & Marshall, 1999). In their study on Eucalyptus globulus, António et al. (2007) observed improvements in the sum of residual squares of 72%, 8%, 12% and 10% for wood, bark, leaves and branches respectively, after inclusion of h together with d. It is possible that in the present study the tree sample was not representative of the entire variability in height for a given diameter. Some studies included h and d in biomass models, together with density, age and site index (Ter-Mikaelian & Parker, 2000; António et al., 2007), and these models are therefore suitable for comparing different sites (Ketterings et al., 2001). Other studies included age as an independent variable in biomass equations (Porté et al., 2002; Saint-André et al., 2005), thus producing dynamic models with which biomass increments can be estimated by derivative analysis. Although individually dst worked well as a predictor, it is seldom measured in forest inventories. On the other hand, it is sometimes useful to estimate dry biomass when trees are already cut down and only stump dimensions are available. It has been observed that some crown variables work well as predictors of crown fractions (Clark, 1982; Satoo & Madgwick, 1982; Carvalho & Parresol, 2003). In the present study, inclusion of crown variables improved the RMSE by 1.8%, 10.8%, 19.1% and 17.3% for respectively dead branches, thick branches, twigs and leaves, in the individual fit. These improvements are smaller than those obtained by António et al. (2007) for Eucalyptus globulus in Portugal, probably because of the lower genetic variability in the plantations considered in our study. The best improvement was for leaves, which implies better estimations of a fraction that is very difficult to predict and is very important as regards nutrition and ecology. Overall, the results indicate a low accuracy of estimation of the bark fraction in the present study, in comparison with reports for other species of Eucalyptus. Wood, bark and thin branches depend on the same variables in both systems of equations, and the results obtained by simultaneous fitting were generally only slightly less accurate. For leaves, the reduction in R2Adj. derived from simultaneous fitting was 6.3%. The ability of the fitted biomass equations to evaluate the proportion of each aboveground biomass component for a range of diameters has seldom been studied. The proportions are often considered as parameters for ecophysiological models, particularly for small diameters (Sands & Landsberg, 2002). The present results show that, if a threshold diameter is considered for defining a wood component, minimum breast height diameter must be considered to define the range of use of the biomass components equations, if sound estimation of these percentages is sought. CHAPTER III 86 3.5. Conclusions Two systems of equations were fitted for aboveground biomass components of Eucalyptus nitens. The inclusion of crown variables as predictive variables provided poorer results for total biomass, wood and thin branches, but improved the accuracy of estimation for twigs, leaves, thick branches and dead branches. Stem subsampling affects estimation of the wood fraction. If a systematic subsample of disks or logs is taken, the variation in moisture content or basic density along the stem should be considered. Less intensive sampling usually leads to overestimation of biomass with both methods (complete fresh weighing or partial fresh weighing). The minimum subsampling intensity for an assumed ±5% RD in wood dry biomass estimation depends on the tree diameter class, with a range of 0.75 to 0.95 disks m-1 in the CW method. Use of the PW method would require very intense subsampling to reduce the relative error, independently of tree size. The average basic density usually occurs at a relative height of 30-35% along the stem. It is not always advisable to choose the first section of study at the bottom of the stem. 3.6. Bibliography Álvarez-González, J.G.; Rodríguez-Soalleiro, R. & Rojo-Alboreca, A. 2007. Resolución de problemas del ajuste simultáneo de sistemas de ecuaciones: heterocedasticidad y variables dependientes con distinto número de observaciones. En: (Jarandilla de la Vera (Cáceres)). Mayo, . António, N.; Tomé, M.; Tomé, J.; Soares, P. & Fontes, L. 2007. Effect of tree, stand, and site variables on the allometry of Eucalyptus globulus tree biomass. Canadian Journal of Forest Research. 37 (5): 895-906. Bartelink, H.H. 1996. Allometric relationships on biomass and needle area of Douglas-fir. Forest Ecology and Management. 86 (1-3): 193-203. Breusch, T.S. & Pagan, A.R. 1979. A simple test for heteroscedasticity and random coefficient variation. Econometrica. 47 (5): 1287-1294. Briggs, R.D.; Cunia, T.; White, E.H. & Yawney, H.W. 1987. Estimating sample tree biomass by subsampling: some empirical results. Estimating Tree Biomass Regressions and Their Error.Proceedings of the Workshop on Tree Biomass Regression Functions and their Contribution to the Error of Forest Inventory Estimates.Compiled by EH Wharton and T.Cunia.USDA For.Serv.Gen.Tech.Rep.NE-117. : 119-127. Broad, L.R. 1998. Allometry and Growth. Forest Science. 44: 458-464. Brown, I.F.; Martinelli, L.A.; Thomas, W.W.; Moreira, M.Z.; Ferreira, C.A.C. & Victoria, R.A. 1995. Uncertainty in the biomass of Amazonian forests: an example from Rondônia, Brazil. Forest Ecology and Management. 75 (1-3): 175-189. Brown, S. 2002. Measuring carbon in forests: current status and future challenges. Environmental Pollution. 116 (3): 363-372. Carvalho, J.P. & Parresol, B.R. 2003. Additivity in tree biomass components of Pyrenean oak (Quercus pyrenaica Willd.). Forest Ecology and Management. 179 (1-3): 269-276. IMPROVEMENT IN ACCURACY OF ABOVEGROUND BIOMASS ESTIMATION IN EUCALYPTUS NITENS PLANTATIONS: EFFECT OF BOLE SAMPLING INTENSITY AND EXPLANATORY VARIABLES 87 Chave, J.; Riéra, B. & Dubois, M.A. 2001. Estimation of biomass in a neotropical forest of French Guiana: spatial and temporal variability. Journal of Tropical Ecology. 17 (01): 79-96. Clark, A. 1982. Predicting biomass production in the South. En: Predicting Growth and Yield in the Mid-South 119-139 pp. Clutter, J.L.; Fortson, J.C.; Pienaar, L.V.; Brister, H.G. & Bailey, R.L. 1983. Timber management: a quantitative approach John Wiley and Sons, New York. 333 pp. Cunia, T. 1987. Construction of tree biomass tables by linear regression techniques. En: Estimating Tree Biomass Regressions and their Error: Proceedings of the Workshop on Tree Biomass Regression Funktions and their Contribution to the Error of Forest Inventory Estimates 27-36 pp. Cunia, T. 1979. On sampling trees for biomass tables construction: some statistical comments. En: Forest resource inventories, Vol. 2. Ed. W.E. Frayer, Colorado State University, Fort Collins, Colorado. 643-664 pp. Gregoire, T.G.; Valentine, H.T. & Furnival, G.M. 1995. Sampling methods to estimate foliage and other characteristics of individual trees. Ecology. 76 (4): 1181-1194. Harvey, A.C. 1976. Estimating Regression Models with Multiplicative Heteroscedasticity. Econometrica. 44 (3): 461-465. Jenkins, J.C.; Chojnacky, D.C.; Heath, L.S. & Birdsey, R.A. 2003. National-Scale Biomass Estimators for United States Tree Species. Forest Science. 49 (1): 13-35. Johansson, T. 1999. Biomass equations for determining fractions of pendula and pubescent birches growing on abandoned farmland and some practical implications. Biomass and Bioenergy. 16 (3): 223-238. Jokela, E.J.; Van Grup, K.P.; Briggs, R.D. & White, E.H. 1986. Biomass estimation equations for norway spruce in New York. Canadian Journal of Forest Research. 16 (2): 413-415. Ketterings, Q.M.; Coe, R.; van Noordwijk, M.; Ambagau, Y. & Palm, C.A. 2001. Reducing uncertainty in the use of allometric biomass equations for predicting above-ground tree biomass in mixed secondary forests. Forest Ecology and Management. 146 (1-3): 199-209. Kleinn, C. & Pelz, D.R. 1987. Subsampling trees for biomass. En: Estimating Tree Biomass Regressions and their Error: Proceedings of the Workshop on tree Biomass Regression Functions and their Contribution to the Error of Forest Inventory Estimates. Eds. E.H. Wharton & T. Cunia, Gen.Tech.Rep. NE-117 Edn., USDA Forest Service, Broomall, PA: U.S. 225-227 pp. Loomis, R.M.; Phares, R.E. & Crosby, J.S. 1966. Estimating foliage and branchwood quantities in shortleaf pine. Forest Science. 12 (1): 30-39. Martínez Cortizas, A. & Pérez Alberti, A. 1999. Atlas Climático de Galicia Xunta de Galicia, Santiago de Compostela. 207 pp. Monserud, R.A. & Marshall, J.D. 1999. Allometric crown relations in three northern Idaho conifer species. Canadian Journal of Forest Research. 29 (5): 521-535. Neter, J.; Wasserman, W.; Kutner, M.H. & William Wasserman, M.H.K. 1989. Applied linear regression models Irwin Homewood, Ill, New York. Parresol, B.R. 2001. Additivity of nonlinear biomass equations. Canadian Journal of Forest Research. 31 (5): 865-878. Parresol, B.R. 1999. Assessing Tree and Stand Biomass: A Review with Examples and Critical Comparisons. Forest Science. 45: 573-593. Parresol, B.R. 1993. Modeling multiplicative error variance: an example predicting tree diameter from stump dimensions in baldcypress. Forest Science. 39 (4): 670-679. Parresol, B.R. & Thomas, C.E. 1989. A density-integral approach to estimating stem biomass. Forest Ecology and Management. 26 (4): 285-297. CHAPTER III 88 Pearson, J.A.; Fahey, T.J. & Knight, D.H. 1984. Biomass and leaf area in contrasting lodgepole pine forests. Canadian Journal of Forest Research. 14 (6): 259-265. Pérez-Cruzado, C. 2009. Herramientas de gestión para plantaciones de Eucalyptus nitens (Deane & Maiden) Maiden con el objetivo de fijación de carbono. Trabajo de Investigación Tutelado. Escuela Politécnica Superior de Lugo, Universidad de Santiago de Compostela. 77 pp. Porté, A.; Trichet, P.; Bert, D. & Loustau, D. 2002. Allometric relationships for branch and tree woody biomass of Maritime pine (Pinus pinaster Ait.). Forest Ecology and Management. 158 (1-3): 71-83. Reed, D. & Tomé, M. 1998. Total aboveground biomass and net dry matter accumulation by plant component in young Eucalyptus globulus in response to irrigation. Forest Ecology and Management. 103 (1): 21-32. Saint-André, L.; M'Bou, A.T.; Mabiala, A.; Mouvondy, W.; Jourdan, C.; Roupsard, O.; Deleporte, P.; Hamel, O. & Nouvellon, Y. 2005. Age-related equations for above-and below-ground biomass of a Eucalyptus hybrid in Congo. Forest Ecology and Management. 205 (1-3): 199214. Sands, P.J. & Landsberg, J.J. 2002. Parameterisation of 3-PG for plantation grown Eucalyptus globulus. Forest Ecology and Management. 163 (1-3): 273-292. SAS Institute Inc 2004. SAS/STAT 9.1 User's Guide Cary, N.C. Satoo, T. & Madgwick, H.A.I. (eds) 1982. Forest Biomass, JSTOR, The Netherlands. Schaegel, B.E. 1982. Boxelder (Acer Negundo L.) Biomass Component Regression Analysis for the Mississippi Delta. Forest Science. 20 (4): 617-628. Snowdon, P.; Eamus, D.; Gibbons, P.; Khanna, P.K.; Keith, H.; Raison, R.J. & Kirschbaum, M.U.F. 2000. Synthesis of allometrics, review of root biomass and design of future woody biomass sampling strategies. NCAS Technical Report Edn., Australian Greenhouse Office, Canberra. 114 pp. Ter-Mikaelian, M.T. & Korzukhin, M.D. 1997. Biomass equations for sixty-five North American tree species. Forest Ecology and Management. 97 (1): 1-24. Ter-Mikaelian, M.T. & Parker, W.C. 2000. Estimating biomass of white spruce seedlings with vertical photo imagery. New Forests. 20 (2): 145-162. Valentine, H.T.; Tritton, L.M. & Furnival, G.M. 1984. Subsampling trees for biomass, volume, or mineral content. Forest Science. 30 (3): 673-681. Verwijst, T. & Telenius, B. 1999. Biomass estimation procedures in short rotation forestry. Forest Ecology and Management. 121 (1-2): 137-146. White, H. 1980. A heterocedasticity-consistent covariance matrix estimator and a direct test for heterocedasticity. Econometrica. 48 (4): 817-838. Zianis, D. & Mencuccini, M. 2004. On simplifying allometric analyses of forest biomass. Forest Ecology and Management. 187 (2-3): 311-332. 89 Chapter IV A management tool for estimating bioenergy production and carbon sequestration in Eucalyptus globulus and Eucalyptus nitens grown as short rotation woody crops in north-west Spain 90 A MANAGEMENT TOOL FOR ESTIMATING BIOENERGY PRODUCTION AND CARBON SEQUESTRATION IN EUCALYPTUS GLOBULUS AND EUCALYPTUS NITENS GROWN AS SHORT ROTATION WOODY CROPS IN NORTH-WEST SPAIN 91 4. A management tool for estimating bioenergy production and carbon sequestration in Eucalyptus globulus and Eucalyptus nitens grown as short rotation woody crops in north-west Spain Abstract This study proposes stand level models for estimating biomass yield, total energy and carbon sequestration in Eucalyptus globulus and Eucalyptus nitens plantations, on the basis of measurements made in 131 plots established at the usual range of initial forest densities for southwestern Europe. The timber volume, total aboveground biomass, logging residue biomass, crown biomass, carbon in aboveground biomass and soil organic layer, energy in aboveground biomass, energy in logging residue biomass and usable cellulose yield were represented in the form of isolines (taking mortality into account) and plotted against dominant height. These variables were calculated and compared with previously published data on two silvicultural options for short rotation forestry, one destined for bioenergy production and the other consisting of the standard silviculture regime applied to both species in southern Europe, considering the average site index for each species. Yield levels were higher in Eucalyptus nitens than in Eucalyptus globulus for all variables because of faster diameter increment at similar densities. The total yield in terms of biomass was 13.9-14.6 Mg ha-1 yr-1 for Eucalyptus globulus and 20.4-21.5 Mg ha-1 yr-1 for Eucalyptus nitens. Energy in aboveground biomass ranged between 233 and 245 GJ ha-1 yr-1 for Eucalyptus globulus and 345 and 364 GJ ha-1 yr-1 for Eucalyptus nitens, carbon accumulation rate in aboveground biomass and soil organic layer was 6.9-7.2 Mg ha-1 yr-1 for Eucalyptus globulus and 12.7-13.5 Mg ha-1 yr-1 for Eucalyptus nitens, and usable cellulose was 5.7-5.9 Mg ha-1 yr-1 for Eucalyptus globulus and 9.0-10.1 Mg ha-1 yr-1 for Eucalyptus nitens. It was found that 50% increments in the initial density result in only marginal increments in biomass and usable cellulose yields. Keywords: short rotation forestry, eucalypts, woody crops, density management diagrams, bioenergy production, carbon sequestration. 4.1. Introduction Short rotation woody crops (SRWCs) (grown in short rotation forestry: SRF) are an important potential source of cellulosic biomass, which can be used as solid fuel in the form of wood chips, CHAPTER IV 92 pellets or charcoal, transformed into ethanol via a cellulosic platform and/or used in pyrolysis to generate syngas and other products (Johnson et al., 2007). The establishment of single stem stands and subsequent replanting or coppicing is the most common management regime in SRF, and decisions about replanting and coppicing should consider: (i) the yields from both options (ii) establishment costs, and (iii) desired dimensions of the final product (Rockwood et al., 2006). The genus Eucalyptus has been used in forestation in Europe since the early 19th century because of its high productivity and plasticity. The total area currently occupied by Eucalyptus plantations in southern Europe is approximately 14000 km2, with Eucalyptus globulus being the most common species but with an increasing proportion of Eucalyptus nitens, which is grown successfully as a frost-tolerant species. Both species belong to the subgenus Symphyomyrtus, known to produce larger average tree sizes and to be more productive than species of the subgenus Monocalyptus (Davidson & Reid, 1980; Turnbull et al., 1993; Sims et al., 1999a). The management objective of these plantations in southern Europe is currently the production of wood pulp or fibreboard, although logging residues and the bark derived from the harvesting operations are increasingly used as biofuel to produce thermal energy and electricity. The ideal characteristics of an energy crop are: (i) high yield, (ii) low energy input for production, (iii) low cost, (iv) minimal contents of contaminants and (v) low nutrient requirements (McKendry, 2002). In this sense, Eucalyptus species adapt well to energy production, because of the high yield and low water and nutrient requirements than for poplars and willows (Johnson et al., 2007); Eucalyptus species accounted for 38% of total SRF plantations throughout the world in 2003 (FAO, 2003). Before establishment of a plantation it is important to consider the desired combination of stem density and rotation. For species that must be established from seedlings, the application of narrow interand intrarow spacing would lead to very high costs and low ratios of wood/other biomass components, wood/bark or percentage of cellulose. Moreover, the average size of tree decreases as N increases (Bullard et al., 2002) and larger logs are more dense. Nutrient depletion is also less likely for longer rotations, which also provide product flexibility (Ericsson, 1994; Guo et al., 2002). Eucalypts in southern Europe established at initial densities of 1000 to 2400 stems per ha can therefore supply the bioenergy industry as the main plantation objective or through the use of logging residues for energetic purposes. Forest plantations develop from a collection of individual, freely growing trees, through the onset of competition, to full site occupancy and self-thinning. Stand development is commonly displayed as a trajectory of increasing mean tree size with decreasing stand density (Long et al., 2004). Dynamic stand density management diagrams (SDMDs) illustrate the relationships among yield, density and density-dependent mortality at all stages of stand development. Their use has been proven to be an effective method for the design, display and evaluation of alternative density management regimes in the field of even-aged forestry (Newton, 1997). The adaptation of such management tools to the field of short rotation forestry for bioenergy production may assist in the assessment of energy yield potential, optimum stand management in terms of density and rotation in comparison with other potentially useful woody species. A MANAGEMENT TOOL FOR ESTIMATING BIOENERGY PRODUCTION AND CARBON SEQUESTRATION IN EUCALYPTUS GLOBULUS AND EUCALYPTUS NITENS GROWN AS SHORT ROTATION WOODY CROPS IN NORTH-WEST SPAIN 99 against stand density to determine the maximum average crown biomass at 1000 stems ha-1 (Wc1000, kg tree-1). The FDC equation was then derived from equation [4.14] by substituting K for 2 3 1000 1000 c W. Although the -3/2 power law of self-thinning (Yoda et al. 1963) is assumed to be valid for all species and locations, some factors such as severe nutrient deficiencies (Lonsdale & Watkinson, 1982; Morris & Myerscough, 1991), climatic conditions and artificial shading (Aikman & Watkinson, 1980) may affect the trajectory of FDC. This implies that it is more appropriate to define a full density zone (FDZ), in which the probability of mortality due to the self-thinning effect is very high (Jack & Long, 1996). The lowest boundary limit of FDZ was determined as 70% of Wc1000 (DeBell & Whitesell, 1988). Both limits of FDZ were represented in the diagrams by solving equation [4.14] for N. Mechanized harvesting implies a size limitation due to the machinery used. Isolines of 10 and 20 cm of average stump diameter ( st d, cm) were therefore represented in the diagrams. For this, a linear relationship between dg and d was fitted for both species. st d can be estimated accurately from the average diameter at breast height by use of a linear model (Diéguez-Aranda et al., 2003). The model developed by Sánchez et al. (2004) was used to estimate E. globulus st d, and a similar model, developed with data collected in the plots established in the present study, was used to estimate E. nitens st d. 4.3. Results 4.3.1. Model parameters and additional relations Results of the non-linear fit of equations [4.2-4.10], the coefficient estimates and the regression statistic values are shown in Table 4.3. All coefficients were significant at p<0.05, and the models accounted for more than 84% of the total variability in the quadratic mean diameter, and more than 94% of the total variability in productivity, equations [4.3-4.10]. As expected, the least accurate models were those predicting dg, since the approach used in developing this type of diagram is to predict yield rather than increment, which makes them of little use, relative to dynamic growth models, for simulating a broad range of silvicultural regimes (García, 1994). With regard to the productivity equations, the C model was the least accurate, because the carbon in the soil organic layer is barely related to the independent variables considered, and is more closely related to time since last perturbation. The changes in total carbon in the aboveground biomass and in the soil organic layer over time are shown in Fig. 4.2. CHAPTER IV 100 Table 4.3. Non-linear regression coefficients and statistics obtained from simultaneous fitting of the system of 10 equations predicting quadratic mean diameter (dg, cm), stand volume (V, m3 ha-1), total aerial biomass (W, Mg ha-1), crown biomass (Wc, Mg ha-1), total carbon in aboveground biomass and soil organic layer (C, Mg ha-1), logging residue biomass (Ww, Mg ha-1), total aerial energy without leaves (E, TJ ha-1), logging residue energy (Ew, TJ ha-1), usable cellulose production (UC, Mg ha-1). Equation Eucalyptus globulus Parameter estimates Adjusted R2 RMSE [4.2] b0=16.37744 (9.7711) b1=-0.38706 (0.0699) b2=0,.868226 (0.0858) 0.8480 3.1225 [4.3] b3=0.000064 (0.000018) b4=2.047556 (0.0504) b5=0.756605 (0.0482) b6=0.981972 (0.0235) 0.9950 17.8455 [4.4] b7= 0.000034 (5.775E-6) b8=2.132935 (0.0164) b9=0.732106 (0.0175) b10=0.97052 (0.0133) 0.9984 6.0913 [4.5] b11=0.000012 (1.79E-6) b12=2.457291 (0.0145) b13=0.077533 (0.0148) b14=1.036985 (0.0134) 0.9987 0.8934 [4.6] b15=0,.000021 (2.023E-6) b16=2.259985 (0.00986) b17=0.03626 (0.0109) b18=1.014372 (0.00884) 0.9992 0.4474 [4.7] b19=0.000073 (0.000029) b20=2.058496 (0.0528) b21=0.512052 (0.0602) b22=0.899273 (0.0426) 0.9911 6.9936 [4.8] b23=5.544E-7 (9.644E-8) b24=2.130056 (0.0168) b25=0.745888 (0.0179) b26=0.969512 (0.0136) 0.9984 0.1043 [4.9] b27=1.888E-7 (2.813E-8) b28=2.459492 (0.0146) b29=0.07792 (0.0149) b30=1.037102 (0.0135) 0.9987 0.0142 [4.10] b31=0.00001 (2.413E-6) b32=2.073132 (0.0228) b33=0.918631 (0.0247) b34=0.953745 (0.0180) 0.9974 3.3231 Equation Eucalyptus nitens Parameter estimates Adjusted R2 RMSE [4.2] b0=23.23792 (11.4785) b1=-0.34626 (0.0586) b2= 0.70549 (0.0728) 0.8437 3.0194 [4.3] b3=0.000068 (0.000017) b4=1.936645 (0.0752) b5=0.800026 (0.0628) b6=1,005736 (0.0252) 0.9956 14.9311 [4.4] b7=0.000025 (2.157E-6) b8=2.24867 (0.0140) b9=0.665185 (0.0128) b10=0.980934 (0.00781) 0.9987 4.2858 [4.5] b11=0.000026 (4.245E-6) b12=2.353757 (0.0326) b13=0.08703 (0.0265) b14=0.991036 (0.0168) 0.9960 1.5209 [4.6] b15= 2.714E-6 (5.996E-7) b16=2.756195 (0.0409) b17=0.180144 (0.0308) b18=1.002181 (0.0218) 0.9941 1.1395 [4.7] b19=0.000231 (1.97E-5) b20=1.902716 (0.1489) b21=0.485519 (0.1429) b22=0.834961 (0.1109) 0.9475 14.9173 [4.8] b23= 4.196E-7 (3.662E-8) b24=2.24873 (0.0140) b25=0.667739 (0.0128) b26=0.980905 (0.00783) 0.9987 0.0734 [4.9] b27=4.219E-7 (7.089E-8) b28=2.355724 (0.0341) b29=0.089347 (0.0277) b30=0.99033 (0.0175) 0.9957 0.0256 [4.10] b31=0.000016 (2.892E-6) b32=1.844636 (0.0275) b33=0.93374 (0.0290) b34=0.973451 (0.0137) 0.9956 3.4338 A MANAGEMENT TOOL FOR ESTIMATING BIOENERGY PRODUCTION AND CARBON SEQUESTRATION IN EUCALYPTUS GLOBULUS AND EUCALYPTUS NITENS GROWN AS SHORT ROTATION WOODY CROPS IN NORTH-WEST SPAIN 101 Eucalyptus globulus 0 50 100 150 200 250 0 5 10 15 20 25 t (years) Carbon (Mg ha -1 ) Eucalyptus nitens 0 50 100 150 200 250 0 5 10 15 20 25 t (years) Carbon (Mg ha -1 ) ABOVEGROUND BIOMASS SOIL ORGANIC LAYER Figure 4.2. Carbon accumulation in aboveground biomass and soil organic layer over time (Mg ha-1) for Eucalyptus globulus and Eucalyptus nitens plantations. The dynamics of accumulation were quite different in both compartments (Fig. 4.2), while there was a high rate of accumulation of aboveground biomass carbon in both species (although dependent of site quality, Carbon at 8.3 Mg ha-1 yr-1 on average), the carbon in the litter tends to stabilize at 5-7 years in both species, and reaches values of Carbon at 10 and 25 Mg ha-1 at this age in E. globulus and E. nitens plantations. The relation between the average crown biomass (kg tree-1) and the number of trees per hectare for each plot and species, including several isolines of Wc1000 is shown in Fig. 4.3. If we consider only the plots with low IH values and that probably undergo self-thinning (Lonsdale, 1990), the most reasonable values for the Wc1000 are 60 kg tree-1 for E. globulus and 50 kg tree-1 for E. nitens. This indicates that E. nitens tolerates lower densities than E. globulus, which is consistent with previous findings for this species (Sims et al., 1999a; Sims et al., 2001). These values are equivalent to aboveground biomass of 360 kg tree-1 for E. globulus and 285 kg tree-1 for E. nitens, values similar to the self-thinning threshold defined to parameterize the 3-PG model for E. globulus in Australia (300 kg tree-1, Sands & Landsberg, 2002). Eucalyptus nitens 45 50 55 60 65 70 0 20 40 60 80 100 120 400 600 800 1000 1200 1400 1600 1800 2000 2200 N Average crown biomass Wc1000 Eucalyptus globulus 45 50 55 60 65 70 0 20 40 60 80 100 120 400 600 800 1000 1200 1400 1600 1800 2000 N Average crown biomass Wc1000 Figure 4.3. Relation between average crown biomass and number of stems per hectare (N) for Eucalyptus globulus and Eucalyptus nitens plantations: determination of self-thinning threshold. CHAPTER IV 102 The parameter values for the linear models of destimation from dg for each species are shown in Table 4.4. The model accuracy was very high for both species, allowing the incorporation of the harvesting limits for st dof 10 and 20 cm. Table 4.4. Equations for dg estimation from d for Eucalyptus globulus and Eucalyptus nitens plantations. Equation Adjusted R2 RMSE Eucalyptus globulus 4067004051 .d.dg    0.9954 0.5474 Eucalyptus nitens 344003221 .d.dg    0.9937 0.4365 4.3.2. Stand Density Management Diagrams The SDMDs obtained for E. globulus and E. nitens are shown in Figs. 4.4-4.7. The isolines for quadratic mean diameter, carbon in biomass and organic soil, Hart-Becking index and mortality, as well as the positions of the sample plots are shown in Figs. 4.4 and 4.6. The isolines for usable cellulose, aboveground biomass energy, harvest limits and mortality are shown in Figs. 4.5 and 4.7. Although SDMDs provide useful graphic information about stand development stages, for more accurate estimation, equations [4.2-4.10] and the parameters shown in Table 4.3 must be used. As can be seen in the SDMD, dg decreases for a given H0 as N increases because of increased competition for resources, which results in a smaller average tree size (Hamilton, 1969; Assmann, 1970; Curtis, 1970; Dean & Long, 1992). The pattern of dg isolines is parallel to that observed for the harvest limits, and thus a threshold established for basal diameter would mean that a combination of initial planting density and dominant height at harvest would have to be considered. For comparable stages of stand development (in terms of H0 and N), E. nitens accumulates more carbon in biomass and soil organic layer, and this effect is partly attributable to the higher rate of accumulation of C in soil organic layer (Figs. 4.4 and 4.6). The shape of biomass-related isolines (W, Ww, Wc, E and Ew) are more vertical than the volume-related isolines (V and UC) (Figs. 4.5 and 4.7). This indicates that biomass-related isolines are less sensitive to changes in N than volume related isolines. Moreover, for comparable stages of stand development, E. nitens has more energy in aboveground biomass and usable cellulose than E. globulus, because of faster diameter growth at comparable levels of N and H0. These results are consistent with those of other studies for single stem crops of E. nitens at 2200 trees ha-1 (Sims et al., 1999a). A MANAGEMENT TOOL FOR ESTIMATING BIOENERGY PRODUCTION AND CARBON SEQUESTRATION IN EUCALYPTUS GLOBULUS AND EUCALYPTUS NITENS GROWN AS SHORT ROTATION WOODY CROPS IN NORTH-WEST SPAIN 103 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 DOMINANT HEIGHT (m) DENSITY (stems ha -1 ) 600 700 12 18 20 26 30 32 14 800 900 16 22 24 28 10 500 1400 1600 2400 1800 2000 1200 1000 400 2200 8 34 36 10 6 38 40 30 QUADRATIC MEAN DIAMETER (cm) HART-BECKING INDEX (%) MORTALIT Y CARBON IN BIOMASS AND ORGANIC SOIL (Mg·ha-1) PLOTS 20 20 40 60 80 100 120 140 160 180 200 220 240 40 Figure 4.4. Stand Density Management Diagram for Eucalyptus globulus with isolines for: quadratic mean diameter, carbon in biomass and organic soil, Hart-Becking index, mortality and sample plots. 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 DOMINANT HEIGHT (m) DENSITY (stems ha -1 ) 600 700 800 900 500 1400 1600 2400 1800 2000 1200 1000 400 2200 15 45 60 75 90 120 150 180 210 240 1,8 2,4 4,2 1,2 0,6 3 3,6 4,8 5,4 6 6,6 7,2 7,8 8,4 9 10 20 30 ABOVEGROUND BIOMASS ENERGY (TJ·ha -1 ) USABLE CELLULOSE (Mg·ha -1 ) MORTALIT Y HARVEST LIMITS (cm) Figure 4.5. Stand Density Management Diagram for Eucalyptus globulus with isolines for: usable cellulose, aboveground biomass energy, harvest limits and mortality. CHAPTER IV 104 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 DOMINANT HEIGHT (m) DENSITY (stems ha -1 ) 600 700 12 18 20 26 30 32 14 800 900 16 22 24 28 10 500 1400 1600 2400 1800 2000 1200 1000 400 2200 8 34 36 10 20 3040 38 40 20 40 60 80 100 120 140 160 180 200 220 240 260 QUADRATIC MEAN DIAMETER (cm) HART-BECKING INDEX (%) MORTALIT Y CARBON IN BIOMASS AND ORGANIC SOIL (Mg·ha -1 ) PLOTS Figure 4.6. Stand Density Management Diagram for Eucalyptus nitens with isolines for: quadratic mean diameter, carbon in biomass and organic soil, Hart-Becking index, mortality and sample plots. 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 DOMINANT HEIGHT (m) DENSITY (stems ha -1 ) 600 700 1,8 5,4 6 7,2 7,8 2,4 800 900 4,2 6,6 1,2 500 1400 1600 2400 1800 2000 1200 1000 400 2200 0,6 15 30 45 60 75 90 105 120 150 180 33,6 4,8 10 20 A BOVEGROUND BIOMASS ENERGY (TJ·ha -1 ) USABLE CELLULOSE (Mg·ha -1 ) MORTALIT Y HARVEST LIMITS (cm) Figure 4.7. Stand Density Management Diagram for Eucalyptus nitens with isolines for: usable cellulose, aboveground biomass energy, harvest limits and mortality. A MANAGEMENT TOOL FOR ESTIMATING BIOENERGY PRODUCTION AND CARBON SEQUESTRATION IN EUCALYPTUS GLOBULUS AND EUCALYPTUS NITENS GROWN AS SHORT ROTATION WOODY CROPS IN NORTH-WEST SPAIN 105 The changes in live-tree densities are shown in all the diagrams and allow estimation of expected natural mortality after planting in a scenario of no thinning, which is generally applied to woody crops destined for energy or fibre production. It must be considered that these trends were calculated by considering the average site index for both species and thus a site specific calculation using equations [4.11] and [4.12] is recommended if greater accuracy is sought. A high plantation density, together with longer rotations, results in high self-thinning mortality and unstable stems, as already corroborated (Sims et al., 2001). As the initial density increases, the boundary of the FDZ is closer in terms of H0 and therefore the management window is smaller. Although the relationship between the number of trees per hectare and the average tree size is a good indicator of density and therefore of competence, there are other factors that affect the self-thinning process, such as tree distribution and accumulation of gaps (Li et al., 2000), which is not important when a regular square distribution is considered, but could be in the case of paired row layouts. In the SDMD proposed here, several harvesting limits were represented. The st d10 cm limit defines the area in which a chip harvester may be more profitable. The st d 10 and 20 cm limits define the area where a multi-cutter harvester may be more cost effective, whereas average basal diameters above st d20 cm would limit the choice to traditional single-stem harvesters. Since all models depend on N, the production cost can be easily determined, and together with a correct estimate of site quality, will determine the time needed to obtain the desired products and therefore the economic profitability of the plantation. 4.3.3. Practical example of SDMD for determining energy production SDMDs can be used to estimate production from a given stand development stage (N and H0), or to estimate the minimum density that can provide a certain output for an average tree size. As a practical example, two silvicultural alternatives were simulated for each species, with initial densities of 2400 and 1600 stems ha-1, representative of an energy option and a standard southern European pulp production target, respectively. The rotation time (T) was defined as the time that allows efficient harvesting of the crop by a multi-cutter harvester ( st d 20 cm) in the case of the energy target for an average site index. For the standard pulp silviculture, stand growth was simulated until one of the two species reached the FDZ area, for which E. nitens occurred at a dominant height of 29 m. The stand development stages at the time of harvest and rotation time for the average site index are shown for both alternatives and species (Table 4.5). The predicted dg at the end of rotation and yield values for V, W, Ww, Wc, C, E, Ew and UC for the two silvicultural alternatives are also shown. The total aboveground biomass yield for all fractions ranged between 13.9 and 14.6 Mg ha-1 yr-1 for E. globulus and between 20.4 and 21.5 Mg ha-1 yr-1 for E. nitens (Table 4.5). The mean annual increment in terms of biomass was only marginally higher for the high density silviculture, although the rotation age was also shorter in this case. The values for the expected yield in terms CHAPTER IV 106 of energy or usable cellulose were higher for E. nitens, although direct comparison is difficult, as the species grow in different areas. Table 4.5. Predicted yield for the two silvicultural alternatives considered. N0=2400 N 0=1600 Eucalyptus globulus Eucalyptus nitens Eucalyptus globulus Eucalyptus nitens N1 (stems ha-1) 1700 2100 925 1300 H01 (m) 27.6 27.2 28.5 28.7 T (years) 14 11 15 12 dg (cm) 16.7 16.8 20.3 20.2 Yield units ha-1 ha-1 yr-1 ha-1 ha-1 yr-1 ha-1 ha-1 yr-1 ha-1 ha-1 yr-1 V (m3) 356.6 25.5 503.0 45.7 356.3 23.8 489.7 40.8 W (Mg) 204.1 14.6 236.3 21.5 208.2 13.9 244.6 20.4 Ww (Mg) 33.5 2.4 52.9 4.8 34.6 2.3 53.9 4.5 Wc (Mg) 24.8 1.8 25.5 2.3 24.9 1.7 28.0 2.3 C (Mg) 101.1 7.2 148.4 13.5 104.2 6.9 151.9 12.7 E (TJ) 3.430 0.245 4.000 0.364 3.501 0.233 4.141 0.345 Ew (TJ) 0.531 0.038 0.865 0.079 0.549 0.037 0.882 0.074 UC (Mg) 83.2 5.9 110.9 10.1 85.0 5.7 108.4 9.0 4.4. Discussion 4.4.1. Model performance and limitations The model presented here represents a system of related equations that enable accurate estimation of crop yield in terms of oven dry biomass, total energy, usable cellulose and other variables, and thus provides a powerful tool for decision making as regards woody crops of the species studied. The diagrams show the crop development as dominant height increases, and are therefore independent of age and valid for applying to different breeding materials and sites for each species, provided the change in dominant height with age is known in each case. These empirical statistically based tools can also be combined with process-based models if the detailed information required in this case is available. Woody biomass can be converted via combustion, gasification, pyrolysis and fermentation, and the energy recovered depends on the conversion technology (McKendry, 2002). Although the energy obtained from LHV is a theoretical value that can only be achieved at 0% moisture, the information obtained from the diagrams, particularly the share of logging residues, and the proportion of cellulose or stem volume, is useful for determining the suitability of the biomass produced for subsequent processing. A MANAGEMENT TOOL FOR ESTIMATING BIOENERGY PRODUCTION AND CARBON SEQUESTRATION IN EUCALYPTUS GLOBULUS AND EUCALYPTUS NITENS GROWN AS SHORT ROTATION WOODY CROPS IN NORTH-WEST SPAIN 107 The models provided here are only valid for single stem rotations, which is an important constraint to be considered. However, this is not a key limitation in the case of E. nitens, because some studies have shown the poor coppicing ability of E. nitens (Sims et al., 2001; Little & Gardner, 2003), at least for the breeding materials currently in use. Changes in growth patterns and relationships among stand variables after coppicing have been shown, particularly the change in basal area derived from the sprout number per stool (Sims et al., 1999b). It is therefore necessary to obtain a similar model for E. globulus coppice stands in order to assess the long term productivity. One of the main advantages of the models provided in this paper is the possibility of assessing the economic profitability of SRC with eucalypts, because many variables are needed to calculate the cash flow throughout the rotation. Most production costs, from establishment to delivery, can be calculated as they are density-dependent, as is the case of plantation, localized fertilization and weed control. Plant material may account for up to 65% of establishment costs and any advantage gained by high planting rates may be outweighed by increasing costs (Mitchell et al., 1999). Stand harvesting costs constitute a major portion of total production costs, and may have effects as important as those of stand establishment costs (Whitesell et al., 1992). Harvest operations represent up to 70% of the cost in the overall supply chain, and therefore less frequent harvesting reduces the cost of biomass production per unit (Mitchell et al., 1999). Moreover, average tree size is the most important factor in harvesting costs (Whitesell et al., 1992). The information provided in the diagrams, particularly the average basal diameter, is also essential for deciding what type of harvesting to carry out, i.e. whole stem harvesting or chip harvesting. 4.4.2. Carbon sequestration The diagrams obtained in this paper and the fitted equations constitute a powerful tool for assessing carbon sequestration in eucalypts grown in SRF. The isolines plotted in the diagrams range between 20 and 200 Mg ha-1 for carbon in aboveground biomass and litter. Since no information was obtained for root biomass it was not possible to assess sequestration in this compartment, which is an important consideration because root biomass is usually left in situ after the crop is harvested. The values for carbon in litter show greater accumulation of C in the litter resulting from E. nitens than in E. globulus crops. Net production in a forest should be assessed as the sum of biomass accumulation and annual litter production, in which E. globulus plantations of density 4167 stems ha-1 may reach 13.4 Mg ha-1 yr-1, and as much as 10-20% of the total biomass production at age 3 years (Toky & Ramakrishnan, 1983). Although this litter can hardly be used for energy production, estimation of the quantity is important for estimating carbon accumulation and nutrient cycle return. Although the soil mineral layer was not considered here, carbon accumulation in this pool is very important because the lifespan of the carbon is longer than in biomass and in the soil organic layer (Romanya et al., 2000). CHAPTER IV 108 4.4.3. Comparison of data from other studies The biomass yields (Table 4.5) are higher than the 1-9 Mg ha-1 yr-1 reported for 4-year-old E. globulus planted at a density of 2196 stems ha-1 (Cromer et al., 1975), but lower than the 24 Mg ha-1 yr-1 observed in 3-year-old plantations of the same species in New Zealand, established at a density of 4167 stems ha-1 and irrigated with effluent. No irrigated stands yielded as much as 19.3 Mg ha-1 yr-1 (Guo et al., 2002). E. globulus planted at densities of 20000, 30000 and 40000 stems ha-1 in Portugal (Pereira et al., 1994) yielded 16, 21 and 19 Mg ha-1 yr-1 after 2 years. The general effect is therefore a similar average yield (obtained early on because of the initial high stocking density) but with a decrease in the average tree size (Dickmann, 2006). Referenced yield data for E. nitens in SRF are scarce and values are very low compared with those observed in the present study. Sims et al. (2001) reported 3-7.3 Mg ha-1 yr-1 for single stems established at a density of 5000 stems ha-1 and first coppice rotation, respectively, with very low survival. As regards other Eucalyptus species, Wise & Pitman (1981) collected yield data for several Eucalyptus species in Australia, and observed values of between 11 and 16 Mg ha-1 yr-1. Sachs et al. (1980) reported Eucalyptus yields as high as 40 Mg ha-1 yr-1 in a wide range of sites. The average yield for Eucalyptus species in a single stem rotation at a density of 2200 trees ha-1 reported by Sims et al. (1999a) ranged between 9.6 and 15.5 Mg ha-1 yr-1, which are the highest reported values for the species, with high survival rates and larger average tree size. The values observed here were similar or higher and were determined from commercial sized plots with conservative assumptions for site quality. With regard to energy production at the end of rotation, the predicted values ranged between 3.4-3.5 and 4.0-4.1 TJ ha-1 for E. globulus and E. nitens respectively, well above the values reported for poplar (173-259 GJ ha-1; 10-15 Mg ha-1 yr-1) and willow (187-280 GJ ha-1; 10-15 Mg ha-1 yr-1) grown as short rotation woody crops. Mean values ranged from 245 to 345 GJ ha-1 yr-1, still far from the average range for eucalypt plantations in Aracruz (450 to 650 GJ ha-1 yr-1, (Moreira, 2006)). The annual logging residue energy yield was 33-35 and 53-54 GJ ha-1 yr-1 for E. globulus and E. nitens respectively, similar to the 65 GJ ha-1 yr-1 estimated for both species together, also in northern Spain (Pérez et al., 2008). Rotations simulated for the bioenergy alternative in this paper are longer than the average considered in SRC, but there are positive effects of this practice. If wood is the main biomass compartment desired, longer rotations provide higher wood:bark ratios, as shown in Fig. 4.8 for the data used in the present study. The wood:bark ratio can reach up to 80% in E. globulus at 10 years (Guo et al., 2002), but data reported for E. nitens at 3 years indicate a rather low value of 45% (Sims et al., 1999a). Moreover, longer rotations result in larger average tree size, and therefore harvest machinery is better able to discriminate between leaves and other tree fractions, which reduces the nutrient exports and ash production. Bark and leaves contain the highest amounts of ash in all aboveground biomass compartments, with reported values of 1.5-2 times (Pérez et al., 2008), and even 10 times (Ragland & Aerts, 1991) the wood ash content; this is an important APPLICATION OF CALORIMETRY AND THERMAL ANALYSIS TO STUDY THE STABILIZATION OF SOIL ORGANIC MATTER IN AFFORESTED SOILS 115 5. Application of calorimetry and thermal analysis to study the stabilization of soil organic matter in afforested soils Abstract The study of soil organic matter dynamics requires highly reproducible and accurate techniques for application to large numbers of samples. This is especially important for modelling changes in soil organic matter in relation to land use changes, as required by the Kyoto protocol. The main objective of this study was to apply calorimetry and thermal analysis, as novel techniques to elucidate how afforestation affects the nature of SOM and soil microbial metabolism. The techniques were applied to study the SOM dynamics in two afforested stands differing in the tree species used, Pinus radiata and Eucalyptus globulus, established on pastures in a humid temperate region. The application of differential scanning calorimetry and solid state nuclear magnetic resonance revealed that the soil organic matter was constituted by carbohydrates, carbonyl/carboxyl groups, aliphatic components and aromatic carbon in the first years of the rotation. All these fractions became degraded after afforestation. The degradation was monitored by calorespirometry, which provided the calorespirometric ratio of the soil basal metabolism, together with the active biomass and the metabolic quotient. These indexes proved to be sensitive parameters that provided information about changes in the pattern of microbial metabolism in response to changes in the nature and redox state of the carbon substrates, demonstrating degradation of the aromatic and aliphatic organic matter fraction. The techniques were able to distinguish differences in soil organic matter dynamics in the two types of stands, attributable to the different development of understory vegetation and litter composition. Keywords: calorimetry, thermal analysis, afforestation, SOM, biodegradation. 5.1. Introduction Changes in land cover and certain types of land management (such as intensive forestry and agriculture), wildfires, and drainage alter the pools and turnover rates of soil organic carbon (SOC). Since soils represent one of the largest reservoirs of organic C on a global scale, huge amounts of CO2 and other greenhouse gases are being transferred from the soil to the atmosphere as a consequence of such activities. Although important advances have been made in research on this CHAPTER V 116 topic in the last decade, current knowledge of the mechanisms involved in the dynamics of soil organic matter (SOM) still does not enable prediction of the ecosystem response to such perturbations. A substantial part of SOM can be restored by means of conversion of arable land to forest land. For this reason, afforestation is a major strategy included in the Kyoto Protocol on climate change, because of the capacity of forests to restore C in biomass and soils. Although these measures contribute to C sequestration in biomass, the impact on soil C is less clear. Since C sequestration in soils is influenced by numerous factors, such as previous land use, climate and tree species (Paul et al., 2002), it is difficult to establish a common protocol predicting the changes in C in relation to afforestation. Thus, different studies have detected important decreases in SOM during the first years after afforestation as a consequence of the increased C mineralization (Turner & Lambert, 2000; Davis & Condron, 2002). The loss of SOM is attributed to the reduced inputs of labile organic compounds to the mineral soil, which creates an imbalance between inputs of C and respired C (Chen et al., 2000; Saggar et al., 2001), indicating the crucial role of microorganism activity in SOM dynamics. One of the most important difficulties in modelling the soil C sequestration that occurs after afforestation is the complexity of the SOM dynamics, resulting from simultaneous mechanisms, such as molecular recalcitrance, spatial inaccessibility of the decomposer and stabilization by clays and ions (e.g. Sollins et al., 1996; von Lützow et al., 2006). Moreover, it is known that the degradation of recalcitrant OM can easily take place in upper horizons (80% of the SOM in most soils), in which other stabilization mechanisms, such as the spatial inaccessibility of microorganism to SOM and enzymes or interaction with mineral surfaces and metal ions, are less active than in deeper soil horizons (von Lützow et al., 2006). As a consequence, SOM turnover is usually rapid in the upper horizons, but takes longer at greater depths (Fierer et al., 2003; Goberna et al., 2006). For all of these reasons, evaluation of the effects of afforestation on C balance and SOM turnover must be made by characterization of the composition and the stability of SOM substances and by the use of indicators of microbial activity. Analytical techniques such as solid-state 13 C CP-MAS NMR, Fourier transform infrared spectroscopy and pyrolysis/GC-MS are widely used to characterize the structure and composition of SOM. In addition, different fractionation methods have been developed in the last decade to evaluate the relative abundance of labile and recalcitrant forms of C (e.g von Lützow et al., 2007). Same examples of such fractionation approaches include chemical and physical procedures (Sohi et al., 2001; Rovira & Vallejo, 2002; Marriott & Wander, 2006). Moreover, microbial biomarkers and compound-specific stable isotope and radiocarbon analysis enable differentiation of functional fractions (Rethemeyer et al., 2004; Brovkin et al., 2008). All of these strategies constitute valuable tools that have provided important insights into SOM dynamics. However, some studies on dynamics of SOM, especially those focussed on modelling, require simple techniques that could be used for large number of samples, while at the same time offering a high degree of reproducibility and accuracy. In this sense, calorimetry and thermal techniques avoid the need for some of these long procedures, and provide quantitative assessment of SOM turnover and stabilization. APPLICATION OF CALORIMETRY AND THERMAL ANALYSIS TO STUDY THE STABILIZATION OF SOIL ORGANIC MATTER IN AFFORESTED SOILS 117 In recent years, calorimetry and thermal analysis have been use to meet the increasing demand for rapid and more reproducible assessment of SOM properties (Barros et al., 2007). Differential Scanning Calorimetry (DSC) provides information about thermal properties of SOM in relation to its composition and stabilization. These techniques are based on the different temperatures of exothermic decomposition of carbohydrates and carboxyl groups, aliphatic substances and more refractory aromatic C. Thus, DSC has been used to offer a relatively rapid and simple analytical method to study SOM fractions of soils subjected to different types of management or perturbations, such as wildfires (De la Rosa et al., 2008; Duguy & Rovira, 2010), land uses changes (Lopez-Capel et al., 2005; Salgado et al., 2010) and intensive tillage (Plante et al., 2005). The combined use of DSC and other analytical techniques enables better characterization of SOM. Thus, recent studies have combined DSC with e.g. microbial activity indicators (Marinari et al., 2010), SOM chemical and physical fractionation (Dell'Abate et al., 2002; Lopez-Capel et al., 2005; Plante et al., 2005), changes in isotopic C (Kuzyakov et al., 2006; Dorodnikov et al., 2007), pyrolysis/GC-MS (De la Rosa et al., 2008), Fourier transform infrared spectroscopy (Marinari et al., 2010) and NMR (Lopez-Capel et al., 2005; Barros et al., 2011). In the study of the relationships between microbial activity and SOM stabilization, certain indicators such as the microbial biomass pool, the microbial quotient (microbial biomass/SOC) and metabolic quotient (soil respiration/microbial biomass, qCO2) have been used to assess substrate degradability and the degree of substrate limitation for soil microbes (Dilly & Munch, 1998; Bastida et al., 2008). Thus, the application of these parameters provides information about how rapidly organic substrates are metabolized. Soil microbial metabolism can also be measured by isothermal calorimetry, which has been used to study metabolism in microorganisms, animals and plants (Hansen et al., 2002). Despite the different advantages of this technique, its application to soils is still quite limited (Barros et al., 2007; Barros et al., 2011). Since this technique is not invasive, it can improve the accuracy and reproducibility of the determinations. It is also a simple, sensitive and reliable technique, which reduces the problems derived from the lengthy handling procedures, and enables continuous monitoring of soil microbial activity. Thus, calorimetry provides data on microbial biomass (Sparling, 1983) and simultaneous measurement of metabolic heat rate (Φ) and soil respiration, data that are well correlated with those measured by other more traditional techniques (Sparling, 1981; Critter et al., 2004b). Moreover this technique can also provide quantitative indices of microbial metabolism (such as the heat released per unit of microbial biomass), which provide information about the efficiency of carbon utilization by soil microorganisms (Barros & Feijóo, 2003; Zheng et al., 2009). Another useful and novel index obtained by this technique is the calorespirometric ratio (Φ/RCO2), which provides information about the redox state of substrates being metabolized by the microorganisms, and the efficiency of conversion of substrate carbon into living cells in processes associated with microbial biomass gain (Hansen et al., 2004; Wadsö et al., 2004). This parameter has been successfully used in plants and insect studies (Acar et al., 2004; Summers et al., 2009), but CHAPTER V 118 determination of the Φ/RCO2, was not reported for basal metabolism in soils until recently (Barros et al., 2011). The latter study shows that measurements of this ratio can provide information on the nature of the organic substrates being degraded in the soil and may therefore contribute to improving our knowledge about SOM turnover. The main objective of this study was to use calorimetry and thermal analysis to elucidate how afforestation affects the nature and dynamics of SOM in a humid temperate region, where the SOM dynamics are particularly rapid. DSC was used in combination with 13C CP-MAS NMR to study the changes in SOM throughout the rotation, after afforestation. The effect of changes in SOM composition on microbial metabolism was studied by isothermal calorimetry. This should contribute to our understanding of the SOM stabilization processes by providing new insight into SOM properties. 5.2. Material and Methods 5.2.1. Stands selected and sampling The study was carried out in NW Spain (Lugo). All stands were located on former pastures, in which low intensive management was applied over many years, and which have recently been afforested with E. globulus and P. radiata. The stands were located no more than 30 km apart and they had comparable land use history prior to afforestation. Similar forest management was carried out in each type of forest plantation. To ascertain that all sites were similar as regards soil type and land use, selection of the study sites was based on direct observation of the terrain of adjacent pastures, representing the average tendency for each species reported in Chapter II, as well as consultations with local landowners. The ages of the Eucalyptus globulus stands were 1, 5 and 18 yr, and of the Pinus radiata stands 3, 13, 28, 35 and 40 yr, which corresponded to different stages of the rotation of these species in the region (15-18 yr in the former and 35-40 yr in the latter). These stages were: (i) establishment (stands 1E and 3P), which in the eucalypts and pine stands are 1 and 3 yr of age, respectively; (ii) young stages, immediately after canopy closure (5E and 13P); (iii) mature plantations (28 and 35P), and (iv) end of rotation (18E and 40P). The 20 year annual average rainfall of the area is 1158 mm, and the temperature, 15.1 ºC. The wettest month is November, with an average rainfall of 139 mm, and the driest August, with 45 mm. The lowest mean monthly temperature 9.7 ºC occurs in February, and the highest 19.1 ºC, in August. The soils were developed from schist and quartzite, and were classified as Alumi-humic Umbrisol (IUSS Working Group WRB, 2006). The soil has a loam or sandy loam texture and is well drained. The A horizon is rich in organic matter and strongly acidic (Table 5.1). The soil humidity APPLICATION OF CALORIMETRY AND THERMAL ANALYSIS TO STUDY THE STABILIZATION OF SOIL ORGANIC MATTER IN AFFORESTED SOILS 119 and temperature regimes are Udic (mean period with partial drought, 1 month) and Mesic (mean frost-free period, 10 months), respectively. Table 5.1. Characteristics of stands and selected properties of the soil samples (0-5 cm), and organic layer. Age Stocking Tree height Organic Layer Clay Tree Species Sample (yr) (tree ha-1) (m) (Mg ha-1) pH (%) Soil Texture Eucalyptus globulus 1E 1 1146 1.3 0.00 5.88 13 Loam 5E 5 1432 7.7 2.70 5.57 15 Sandy Loam 18E 18 1146 23.0 10.14 4.95 11 Loam Pinus radiata 3P 3 668 1.91 0.31 5.82 16 Silty Loam 13P 13 891 15.15 30.96 5.62 17 Loam 28P 28 859 20.46 56.22 5.32 5 Sandy Loam 35P 30 923 19.54 49.07 5.36 6 Sandy Loam 40P 40 923 25.05 69.32 4.69 17 Silty Loam For soil sampling, a 50 m x 50 m plot was selected within the stand, at a distance of more than 30 m from the edge of the stand. In each plot, forest floor and mineral soil samples were taken at 8 randomly distributed points. The organic horizons were collected with the aid of a frame (25 cm x 25 cm). Sub-samples of the mineral soil layer (0-5 cm) were collected with a steel corer, and were combined to form one bulk sample per plot. Site preparation for forest establishment consisted of ripping, and no fertilization, tillage or weed control was carried out in the plantations. 5.2.2. General soil properties The pH of the soil was measured in 0.1 M KCl with a glass electrode. Total C and N were analyzed with a LECO Elemental analyzer (LECO TruSpec CHNS). Soil particle analysis was performed by laser diffractometry, with a Mastersizer 2000 diffractometer. 5.2.3. Thermal analysis The soil was dried and gently ground in an agate mortar for DSC measurements (DSC Q100 TA instruments). These experiments were conducted at a heating rate of 10°C min–1 under a flux of dry air from 20 to 600 ºC, as previously described (Dell'Abate et al., 2000). The direct integral of the exothermic DSC curves under the flux of dried air with respect to zero gives the heat of combustion of the soil in kJ g-1. All measurements were made on dry weight basis. CHAPTER V 120 5.2.4. Calorimetric measurements Soil samples were sieved (2 mm) and stored at 4 ºC in polyethylene bags. Prior to calorimetric experiments, samples were brought to 25% moisture content, and incubated for about 24 hours at the temperature of the calorimetric measurements, 25ºC. The soil basal metabolism was monitored in a TAM 2277 calorimeter (TA Instruments), which is a very sensitive heat conduction calorimeter with 3 calorimeter channels. Each channel has two calorimeter ampoules: one for the sample and the other for reference samples. The calorimeter was statically calibrated for the soil measurements at an amplifier setting of 300 microwatts (μW). Each soil sample was prepared for calorimetric measurements after equilibrating at 25 ºC by weighing 1.5 g into three 4 ml stainless steal ampoules; the open ampoules were then left, together with a vial containing water, in a sealed polyethylene bag, for 48 hours. This treatment allows the soil metabolism to equilibrate from the storage temperature, 4ºC, to the measurement temperature, 25ºC, and allows the samples to reach vapour equilibrium. A small vial containing 0.2 mL of 0.4M NaOH was then placed in one of the sample ampoules. The three sample ampoules were then closed and placed in the calorimeter at the same time, together with the reference ampoules filled with silica sand. The calorimetric channels with soil measure basal metabolic heat rate, Φ continuously in microwatts, μW or μJ s-1, while the channel with NaOH measures the soil basal metabolic heat rate plus the heat rate from reaction between metabolic CO2 and NaOH, both of which are exothermic. The enthalpy of the CO2 reaction is -108.5 kJ mol-1 at this concentration of NaOH (Criddle et al., 1991; Russell et al., 2006). Φ was measured at 3 minute intervals for between 20 and 48 hours. At the end of the experiment, the NaOH was removed from the calorimeter and the sample resealed and replaced in the calorimeter to check the basal metabolic rate of the soil sample containing the NaOH. This procedure enables the reproducibility of the registered basal metabolism in the three soil samples to be checked. The metabolic heat rate, Φ, the rate of CO2 production, RCO2, and the ratio of metabolic heat rate to CO2 rate, Φ/RCO2, were calculated from the tabulated Φ data by averaging the Φ values from the two channels with only soil and subtracting the values from the Φ measured from the soil sample with NaOH. The results obtained are the tabulated Φ data for the reaction between the CO2 and the NaOH, which were then divided by the enthalpy change (-108.5 kJ mol-1) for the reaction between the NaOH and the CO2, to give RCO2 at each data point in mol CO2 per second. The quotient between the basal metabolic Φ and the tabulated RCO2 values yields Φ/RCO2. The Φ, RCO2, and Φ/RCO2 can be reported quantitatively as the average values of the tabulated data. The standard deviation of the tabulated data provides information about the variability in these values. The quantitative RCO2 was related to the active biomass to give the metabolic quotient, qCO2. Soil active biomass was calculated by the Sparling method (Sparling, 1983) and related to the soil C content to give the ratio of microorganisms to soil carbon, Cmic-C. APPLICATION OF CALORIMETRY AND THERMAL ANALYSIS TO STUDY THE STABILIZATION OF SOIL ORGANIC MATTER IN AFFORESTED SOILS 121 5.2.5. Solid state 13C CP-MAS NMR Solid NMR experiments were performed at 298 K in a 17.6 T Varian Inova-750 spectrometer (operating at 750 MHz proton frequency) equipped with a T3 Varian solid probe (Agilent, Inc, USA). Solid NMR samples were prepared in 3.2 mm rotors with an effective sample capacity of 22 µL, which corresponds to approximately 30 mg of the powdered sample. Carbon chemical shifts were referred to the carbon methylene signal of solid adamantine, at 28.92 ppm. This sample was also used to calibrate the 1D CP-MAS experiments. Cross Polarization Magic Angle Spinning (1D CP-MAS) experiments were carried out with the samples, under the following conditions: the inter-scan delay was set at 0.5 s, the number of scans was 100000 and the MAS rate was 15 kHz. Heteronuclear decoupling during acquisition of the FID was performed with Spinal-64, at a proton field strength of 70 kHz. The cross polarization time was set at 1 ms. During cross polarization, the field strength of the proton pulse was held constant at 75 kHz, and that of the 13C pulse was linearly ramped with a 20 kHz ramp near the matching sideband. The NMR spectra were processed and the area of the signals was quantified with MestreNova software (Mestrelab Research inc). 5.2.6. Statistical analysis The quantitative indices are given as the average of three replicates and the standard deviation. The data were compared by graphical analysis. All the correlations were significant at p<0.05. 5.3. Results 5.3.1. Changes in the SOM in the afforested soils The changes in the soil organic layer and SOC in the uppermost, superficial mineral soil layer (0-5 cm) throughout the rotation period for each species (15 yr and 35-40 yr are the usual rotation periods for E. globulus and P. radiata in the region) are shown in Table 5.1, and as expected, litter accumulation increased with age in both types of stands, although the process was much faster in the pine stands than in the eucalyptus stands. In the P. radiata stands litter accumulation started earlier, and the amount accumulated at the end of the rotation was 7 times higher than in the latter. The mineral soils of both plantations were subjected to large decreases in SOC throughout the rotation. The soils under pines lost more SOC (maximum recorded loss, 70% after 30 yr) and over a longer time that the soil under eucalyptus (maximum recorded, 50% after 5 yr). At the end of the rotation both soils partially recovered their SOC contents. In the pine chronosequence, the C gain CHAPTER V 122 in mineral soil occurred when the litter accumulation was close to steady state. The C/N ratio decreased in the same way as SOC (Table 5.2), and reached values of around 10 in the soils with the highest SOC losses. Table 5.2. SOC, SON and SOM contents of the samples, together with the carbon to nitrogen ratio (C/N) and the microbial active biomass. Tree Species Samples C (%) N (%) C/N SOM(1) (%) Microbial Biomass (μg CX g-1) E. globulus 1E 11.0 0.83 13.2 24.07±0.45 553±16 5E 5.2 0.56 9.3 12.53±0.80 387±28 18E 9.1 0.75 12.2 17.37±0.46 270±14 P. radiata 3P 12.4 0.88 14.1 27.22±1.77 767±98 13P 6.5 0.47 13.9 16.97±0.19 475±97 28P 3.7 0.35 10.5 5.26 ± 0.10 273±19 35P 4.4 0.44 10.0 9.41±0.18 69±12 40P 6.7 0.46 14.7 13.00±0.33 183±45 (1) Soil organic matter. 5.3.2. SOM composition in the mineral soil horizon: DSC and NMR The heat of combustion (Q) of the soil samples, in Joules per gram of soil, determined by direct integration of DSC curves (Fig. 5.1), and the maximum heat of combustion temperatures for SOM are shown in Table 5.3. Q was positively correlated with SOC and SOM (R2 = 0.83 p< 0.001; R2 = 0.88 p<0.001 respectively), reflecting the direct connection between Q and the SOM content of the samples. Such correlations have been reported in previous studies (Barros et al., 2008) and indicate that the heat of combustion of the soil, determined by DSC, is only a function of the SOC and SOM contents when given in Joules per soil mass. Thus, the largest areas limited by the DSC curves for samples 1E and 3P was related to the higher SOM contents, whereas the small areas limited by the DSC curves for samples 5E, 28P and 35P corresponded to the lower SOM contents. Typical DSC curves of the soil samples for the chronosequences of eucalypts and pine respectively, are shown in Figs. 5.1a and 5.1b. Different studies have demonstrated that these curves reflect the properties of the different organic compounds constituting the SOM, on the basis of their different thermal stabilities (Barros et al., 2007). In all soils, the DSC curves exhibit a prominent exothermic combustion peak in a narrow range of 325-339 ºC, which is attributed to the combustion of carbohydrates (Dell'Abate et al., 2002; Lopez-Capel et al., 2005), and named Exo 1 (Grisi et al., 1998). The heights of these Exo 1 peaks varied greatly in the samples, following the changes in C content throughout afforestation. APPLICATION OF CALORIMETRY AND THERMAL ANALYSIS TO STUDY THE STABILIZATION OF SOIL ORGANIC MATTER IN AFFORESTED SOILS 123 a b Figure 5.1. DSC curves for soil mineral samples (0-5 cm) from the E. globulus stands (Fig. 5.1a) and P. radiata stands (Fig. 5.1b). Different heights reflect different SOM contents. The name of the sample denotes the age of the stand and the tree species (e.g. 1E: one-year-old eucalyptus stand). Each curve represents the average of three replicates. In accordance with the higher SOM contents, the samples of the youngest stands of each chronosequence (1E and 3P) exhibited the highest Exo 1 peaks (Figs. 5.1 and 5.2). In these samples, a secondary peak, named Exo 2, with a maximum at about 380 ºC and a tail until about 500 ºC, was also distinguished. In this peak the heat flow generated can be attributed to aliphatic C and fulvic acids (Cuypers et al., 2002; Barros et al., 2011), whereas the heat flow at temperatures higher than 400 ºC would reflect the presence of aromatic compounds in the SOM (Leinweber & Schulten, 1992; Leinweber & Schulten, 1999). Thus, according to this interpretation, the SOM in these samples, corresponding to the first years after afforestation, was characterized by large amounts of carbohydrates and carboxylic C, and lower amounts of aliphatic and aromatic compounds. CHAPTER V 124 Table 5.3. Thermal properties of the soil samples. Heat of combustion (Q) of the soil, and combustion temperatures for Exo 1 and Exo 2 peaks in the DSC curves. Tª end is the temperature at which the combustion ended. The names of the samples denote the age of the stand and the tree species (e.g. 1E: one-year-old E. globulus stand). Data represent the average of 3 replicates. N=3 ± SD. Tree Species Samples Q (kJ g-1) Exo 1 T1 (ºC) Exo 2 T2 (ºC) Tª end (ºC) E. globulus 1E 3.92±0.06 332 ±1 395±1 539±3 5E 1.91±0.01 325 ±2 395±2 505±5 18E 2.13±0.03 329 ±1 490±1 P. radiata 3P 4.45±0.09 330±1 398±2 539±2 13P 2.54±0.03 330±1 417±5 520±1 28P 1.08± 331±1 414±3 502±1 35P 1.13±0.01 326±1 427±2 502±1 40P 1.65±0.01 339±1 427±1 516±4 Figure 5.2. DSC curves for samples taken in the most recent afforested soils (1E and 3P are the samples taken from E. globulus and P. radiata plantations, 1 and 3 yr after afforestation, respectively). The combustion temperatures indicate the same SOM composition in both samples. Each curve represents the average of three replicates. Characterization of SOM in the soils from the youngest stands was complemented by 13C CPMAS NMR analysis for the P. radiata samples (Fig. 5.3). The 13C CP-MAS NMR spectrum of the sample 3P, revealed the presence of signal corresponding to saturated aliphatic chains, branched alkyl-C and CH3O-C groups (0-60 ppm), carbohydrate C (55-110 ppm), aromatic groups (110-165 ppm) and carboxylic C (165-200 ppm). The most prevalent signals in the 13C CP-MAS spectrum correspond to a broad band in the region 45–110 ppm. This region is typical of O-alkyl C groups and is generally attributed to polysaccharide material such as cellulose (Kögel-Knabner, 1997). In the aliphatic region, the spectrum shows two peaks, one at 32 ppm assigned to acetyl C and methyl C in lipid, cutin, suberin or amino acids (Golchin et al., 1996), and the other at 55 ppm,