The loss of biodiversity is affecting all ecosystems on Earth, one of the greatest threats to biodiversity being climate change. Forests have been highlighted for the potential to mitigate climate change by storing carbon above- and belowground in soils. In this thesis, I studied the effects of tree diversity on carbon cycling in subtropical Chinese forests. I aimed to explore the mechanisms behind tree diversity effects on carbon cycling by focusing on microbial-based processes and the consequences of tree diversity-induced spatial heterogeneity. First, my colleagues and I tested the effects of tree diversity on litterfall spatial patterns and the consequences for litter decomposition and quantified the importance of microbial community in decomposition processes. Second, we explored the effects of tree diversity on relationships between soil microbial facets and soil microbial functions. Third, we tested the effects of tree diversity on soil microbial biomass and carbon concentrations, and their mediation by biotic and abiotic conditions. Finally, we explored the consequences of diversifying forests for re-/afforestation initiatives and plantations to reduce atmospheric carbon levels, and the benefits of diversity for mitigating the effects of climate change on ecosystems and human well-being. We highlighted the positive effects of tree diversity on tree productivity. By increasing the amount and diversity of litterfall, tree diversity increased litter decomposition and subsequently the assimilation of tree products into the forest soils. Our investigation has shown the key role of microbial communities for forests carbon dynamics by carrying out litter decomposition, soil heterotrophic respiration, and soil carbon stabilization. Most notably, tree diversity effects on soil microbial respiration were mainly mediated by soil microbial biomass rather than soil microbial community taxonomic or functional diversity. The effects of tree diversity on microbial biomass were mediated by biotic and abiotic conditions. Taken together, we revealed the importance of considering space to understand biodiversity-ecosystem functioning relationships. Finally, we argued that tree diversity is a promising avenue to maximize the potential of re-/afforestation projects to mitigate increasing atmospheric carbon. Moreover, we highlighted that diversifying forests in re-/afforestation initiatives can help to reduce climate change effects on ecosystems: first, by increasing resistance and resilience to extreme climatic events, and second, by buffering microclimatic conditions in natural and urban areas. My investigation highlighted that tree diversity effects on ecosystem functioning could be explained by both mass and diversity effects on higher trophic levels and their functions. In addition, I showed the key role of tree diversity-induced spatial heterogeneity and the need to consider space and time in further research. Moreover, these results need to be combined with practitioner constraints to enable feasible restoration projects.:Summary table Bibliographic information .................................................................................... I ~ XV Main body ......................................................................................................... 1 ~ 212 Supplementary materials ..................................................................................... i ~ xv Scientific supplementary materials ............................................................. -1- ~ - 154- Table of Contents Table of figures .......................................................................................................... XI Table of scientific supplementary materials ............................................................. XIII Glossary ................................................................................................................... XV Introduction ................................................................................................................. 3 Chapter I - Tree diversity effects on litter decomposition are mediated by litterfall and microbial processes .................................................................................................. 35 Transition I - II ........................................................................................................... 67 Chapter II - Tree diversity and soil chemical properties drive the linkages between soil microbial community and ecosystem functioning................................................ 71 Transition II - III ....................................................................................................... 107 Chapter III - Abiotic and biotic drivers of scale-dependent tree trait effects on soil microbial biomass and soil carbon concentration ................................................... 111 Transition III - IV ..................................................................................................... 155 Chapter IV – Diverse forests are cool: promoting diverse forests to mitigate carbon emissions and climate change ............................................................................... 159 General discussion ................................................................................................. 173 Abstract .................................................................................................................. 195 General acknowledgments ..................................................................................... 209 Supplementary materials ..............................................................................................i

Submitted by MARIÂNGELA BRITO FREIBERGER null (maia_bf@yahoo.com.br) on 2016-04-01T13:08:09Z No. of bitstreams: 1 TESE-MARIÂNGELA BRITO FREIBERGER.pdf: 1663017 bytes, checksum: 79b257695de46e620dc8065c2ad772ce (MD5) / Approved for entry into archive by Felipe Augusto Arakaki (arakaki@reitoria.unesp.br) on 2016-04-05T14:11:18Z (GMT) No. of bitstreams: 1 freiberger_mb_dr_bot.pdf: 1663017 bytes, checksum: 79b257695de46e620dc8065c2ad772ce (MD5) / Made available in DSpace on 2016-04-05T14:11:18Z (GMT). No. of bitstreams: 1 freiberger_mb_dr_bot.pdf: 1663017 bytes, checksum: 79b257695de46e620dc8065c2ad772ce (MD5) Previous issue date: 2016-02-22 / Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / O estudo teve por principal objetivo avaliar o estoque de carbono (C), as alterações na quantidade de C microbiano, a qualidade da matéria orgânica (MO) e a emissão de CO2 em decorrência de aplicações de lodos de esgoto em área sob semeadura direta. A área experimental, que apresenta Latossolo Vermelho de textura argilosa e está localizada em Botucatu, SP, tem sido conduzida com os mesmos tratamentos desde 2002. Os resultados constantes no presente estudo, entretanto, foram obtidos no período de outubro de 2012 a outubro de 2014. O delineamento experimental utilizado foi o de blocos ao acaso em esquema fatorial 2 x 4, com quatro repetições. Os tratamentos correspondem à aplicação bienal de dois resíduos (lodo biodigerido – LB e lodo centrifugado – LC) em quatro doses: 0, 2, 4 e 8 Mg ha-1 (base seca) e o sistema de produção utilizado nesse período foi a sucessão soja / aveia-preta. Em cada um dos cultivos avaliou-se características nutricionais e de produtividade das culturas e a emissão de CO2 a partir do solo. Ao final do estudo foram coletadas amostras de solo para análise química básica, fracionamento da MO, C microbiano e estoque de C. A aplicação continuada de LC promoveu aumento do pH do solo e do teor de macronutrientes, principalmente Ca, bem como maior produção de matéria seca e acúmulo de nutrientes na parte aérea da aveia-preta. O LC também promoveu maior acúmulo de C nas plantas e maior atividade dos microrganismos do solo, o que acarretou em maior teor de C da biomassa microbiana (até 390 mg kg-1 na camada superficial), maior decomposição de MO leve e, consequente maior fluxo de CO2 para atmosfera (de 4,8 a 6,2 µmol m-2 s-1). Aplicações de longa data de lodo de esgoto (LB ou LC) resultam no aumento do teor de micronutrientes no solo, a ponto de Cu, Fe, Mn e Zn se apresentarem em níveis que podem ser prejudiciais às plantas. A produtividade da soja foi maior (até 3.232 kg ha-1) quando do uso de doses de lodos equivalentes a 4,5 a 5,3 Mg ha-1. A aplicação de lodos de esgoto resulta em aumento dos teores de C orgânico total (até 19,8 g kg-1), C da fração particulada (até 0,88 g kg-1) e C associado a minerais (até 19,0 g kg-1) somente na camada superficial do solo. Dentre as substâncias húmicas, a fração humina foi a que mais contribuiu com o estoque de C no solo (até 13,8 g kg-1). Após seis aplicações de lodo de esgoto, independentemente da dose e tipo de lodo, o estoque de C no solo aumentou apenas na camada superficial, e correspondeu a 106,2 Mg ha-1. / The study had as main objective to evaluate carbon (C) stock, changes in the amount of microbial C, quality of the soil organic matter (SOM) and CO2 emission as affected by sewage sludge applications in area under no-till. The experimental area, which shows a clayey Rhodic Ferralsol and is located in Botucatu, SP, has been conducted with the same treatments since 2002. The results of the present study, however, were obtained in the period from October 2012 to October 2014. A complete randomized blocks design arranged in a 2x4 factorial scheme and with four replicates was used. The treatments are represented by biennial application of two sewage sludge types (biodigested sludge - BS and centrifuged sludge - CS) in four rates: 0, 2, 4 and 8 Mg ha-1 (dry basis). The cropping system used in the study was a soybean/black oat succession. Yield and nutritional aspects of crops and CO2 emissions from soil were evaluated in each one of the cultivations. At the end of the study, soil samples were collected for analysis of soil fertility, OM fractionation, microbial C and C stock. The continued application of CS increased the pH and macronutrient levels in the soil, mainly Ca, as well as increased dry matter production and nutrient accumulation in aerial part of black oat. CS application also promoted greater accumulation of C in plants and greater activity of soil microorganisms, which led to a greater level of microbial biomass C (up to 390 mg kg-1 in the superficial layer), greater decomposition of light OM and consequently greater CO2 fluxes to the atmosphere (from 4.8 to 6.2 µmol m-2 s-1). Long time applications of sewage sludge (either BS or CS) resulted in increase of micronutrients levels in the soil, up to the point of Cu, Fe, Mn and Zn reach levels that may be harmful to plants. The soybean yield was higher (up to 3,232 kg ha-1) when sludge rates equivalent to 4.5 to 5.3 Mg ha-1 were used. The long-term application of sewage sludge increases the levels of total organic C (up 19.8 g kg-1), particulate fraction of C (up to 0.88 g kg-1) and C associated with minerals (up 19.0 g kg-1) only in the superficial layers of soil. Among the humic substances, the fraction that most contributed to the soil organic C (up 13.8 g kg-1) was humin. After six sewage sludge applications, regardless of the rate and type of sludge, the soil C stock increased only in the surface layer, and in total corresponded to 106.2 Mg ha-1. / FAPESP: 2011/21276-9 / CNPq: 152725/2012-1

L'objectif principal de ce travail est d'évaluer si l'amendement calco-magnésien de deux bassins versants acidifiés du massif vosgien (un sur grès et un sur granite) engendre des changements de la morphologie des humus, de l'abondance et de la diversité de la mésofaune, de la biomasse fongique, du stockage de carbone dans les horizons organiques 5 et 7 ans après l'amendement. Par ailleurs, nous avons effectué une étude dans des conditions contrôlées de laboratoire en utilisant des microcosmes contenant les horizons organiques du sol provenant de bassins versants acidifiés sur grès et sur granite afin d'évaluer les effets à court terme de l'amendement. Les résultats indiquent que les effets de l'amendement sont beaucoup plus marqués sur le substrat le plus acide. À court terme, sur substrat gréseux, nous observons une diminution de l'abondance de la mésofaune puis une augmentation de celle-ci 5 ans après l'amendement avant de ne montrer plus aucun effet 7 ans après l'épandage. L'effet inverse ou aucun effet n'est par contre observé sur substrat granitique. Le contenu en Ca et Mg de l'horizon OL a augmenté avec l'amendement, mais cet effet n'est également significatif que sur grès. La biomasse fongique du sol n'a pas été affectée par le chaulage, mais est significativement plus élevée sur granite que sur grès. Le chaulage a augmenté l'épaisseur de l'horizon OH sur grès alors qu'il a diminué l'épaisseur de l'horizon OL sur granite, probablement en relation avec les modifications de l'activité biologique (mésofaune ou microorganismes). Enfin, sur substrat gréseux, le stock de carbone de l'humus a augmenté du fait d'un épaississement notable de l'horizon OH avec le chaulage. / The main objective of this work is to assess the consequences of calco-magnesium liming on two acidified forested catchments in Vosges Mountain (sandstone and granite) on humus morphology, abundance and diversity of mesofauna, fungal biomass and carbon storage in organic horizons 5 and 7 years after liming. Moreover, we performed a study to verify under laboratories conditions by using microcosms filled with soil from acidified catchments lying on sandstone and granite in order to study the short-term effect of calco-magnesium liming. During the two field campaigns, we determined calcium and magnesium contents of OL horizon, pH, humus morphology, mesofauna abundance and diversity, fungal biomass, soil carbon storage of organic layers OL, OF and OH.Results indicated that calco-magnesium liming affects studied parameters in different ways according to elapsed time after liming and the geological substrate (sandstone and granite), the more acidic catchment (sandstone) showing the most remarkable reaction. Results of immediate effects of liming showed a decrease of mesofauna abundance and then gradually increasing tendencies in field campaigns after 5 years while no effect after 7 years after liming. The opposite or no effects were observed on granite substrate. Ca and Mg contents in OL horizon increased by liming but this effect were only significant on sandstone substrate. Soil fungal biomass was not affected by liming but was significantly higher on granite substrate than sandstone. Liming increased the OH horizon thickness on sandstone while decreased the OL horizon on granite, under the influence of biological activity (mesofauna vs microorganisms).

L’utilisation des Produits Résiduaires Organiques (PRO) par l’agriculture peut permettre de substituer une partie des engrais minéraux et augmenter les niveaux de matière organique des sols. Elle peut également générer des pollutions azotées via la lixiviation de nitrate ou des émissions gazeuses. Mieux connaitre le devenir du carbone (C) et de l’azote (N) après applications de PRO sur les sols agricoles contribue à améliorer les bénéfices de cette pratique et à en limiter les impacts environnementaux. Cette thèse a pour objectif: (i) de prédire les dynamiques du C et du N en cas d’applications de PRO sur les sols agricoles, (ii) d’étudier les effets de scenarios d’apports de PRO sur les stocks de carbone dans les sols, la substitution des engrais azotés et les pollutions azotées dans le contexte du territoire de la Plaine de Versailles, en prenant en compte la diversité des sols, de leurs teneurs en C et des successions culturales, (iii) d’améliorer ces bénéfices à l’échelon du territoire via une distribution optimale des PRO. Le modèle CERES-EGC a été utilisé pour simuler les effets de 13 ans d’apports de PRO sur les dynamiques de C et de N dans le système sol-plante-atmosphère de l’expérimentation QualiAgro, située au cœur du territoire d’étude. Le sous-modèle NCSOIL a été paramétré à partir de cinétiques de minéralisation de C et N mesurées en conditions contrôlées de laboratoire. Le transfert de ces paramètres dans CERES-EGC a permis de simuler correctement les évolution des stocks de carbone dans les sols, les rendements et les prélèvements de N par les cultures ainsi que l’évolution des stocks de N minéral dans les sols. Les dynamiques de minéralisation de C et N des 18 PRO disponibles sur le territoire ont été déterminées. NCSOIL a été paramétré à partir des caractéristiques analytiques et des résultats de fractions biochimiques des PRO. Les types de sol du territoire n’ont pas impacté significativement le paramétrage. Les PRO ont été classés selon 4 types : (1) composts stables, (2) composts moins matures restant plus réactifs et les fumiers stables, (3) des fumiers très réactifs correspondant à des fumiers de chevaux, (4) des PRO très réactifs tels que des boues et des fientes pouvant plutôt être utilisés comme fertilisants. De nombreux scenarios d’apport de PRO, contraints par les flux de phosphore et de N apportés, ont été simulés sur 20 ans dans tous les contextes de sols, successions culturales et teneurs en C organique des sols du territoire. Le type de sol a impacté le plus le stockage de C et les lixiviations de nitrate tandis que les successions culturales étaient prépondérantes sur la substitution des engrais. Des composts ont permis de stocker 1.1 t C ha- 1 an-1 allant jusqu’63% du C apporté. Des substitutions de 74 kg N ha-1 an-1 ont été atteintes avec une boue mais aussi un compost en prenant en compte la substitution des engrais liée aux augmentations de matière organique dans les sols. La substitution des engrais a pu atteindre plus de 90% du N apporté par les PRO, ceci étant lié aux arrières-effets sur le stockage de matière organique dans les sols et dépendant des hypothèses de calcul de la substitution des engrais azotés. Un modèle d’optimisation a été développé pour sélectionner les scenarios d’apports de PRO pour chaque ilot du territoire (sol x succession de culture x teneur en C organique initiale x surface) tenant compte de la disponibilité des PRO et avec pour objectif de maximiser le stockage de C ou les substitutions en engrais azotés ou de minimiser la lixiviation de nitrate au niveau du territoire. En apportant préférentiellement les PRO les plus stables sur les sols à fort potentiel de stockage de C (fortes teneurs en argile et calcaire), jusqu’à 0.47 t C ha-1 an- 1 pourrait être stocké. En appliquant préférentiellement les PRO fertilisants sur les successions à maïs, et les PRO amendants sur les successions à colza jusqu’à 53 kg N ha-1 an-1 de N de synthèse pourrait être économisés. / The use of Exogenous Organic Matter (EOM) in agriculture could be an efficient way to substitute mineral fertilisation and increase soil organic matter (SOM) enhancing soil fertility and storing carbon (C). It could also cause nitrogen (N) pollutions such as nitrate leaching and gas emissions. Better understanding of C and N fate after EOM applications on cropped soils would allow improving these benefits while limiting environmental impacts. This thesis aims at: (i) predicting EOM impacts when applied on cropped soils, (ii) studying the effects of various scenarios of EOM applications in terms of C storage, synthetic N saving and N pollutions in the context of the Plain of Versailles region (221 km²) and taking into account soil diversity, crop successions and soil organic C contents, (ii) studying the potential for improving these benefits at the regional scale with an optimal distribution of EOM. The CERES-EGC crop model was used to simulate the effects of repeated applications of EOM over 13 years on both soil C and N dynamics in the soil-crop-water-air system of the long-term field experiment QualiAgro located within the region. The sub-model NCSOIL was parameterised from C and N mineralisation kinetics of EOM measured in laboratory conditions. When transposing the parameters into the CERES-EGC model, C storage at the field scale was well simulated, together with crop N uptake and yields, as well as soil mineral N contents. The kinetics of C and N mineralisation of the 18 EOM available in the region were used along with EOM biochemical fractionations for parameterising the NCSOIL model. The soil type did not significantly change EOM parameters. Four groups of EOM were distinguished based on their C and N dynamics: (i) stable composts, (2) more reactive and less mature composts and stable manures, (3) manures with reactive OM corresponding to horse manures and (4) very reactive EOM as sludges, litters that should be used as fertilisers. Numerous scenarios of EOM applications, constrained on the phosphorus and N quantities they bring (and limiting the input in trace elements), were simulated for 20 years in all regional contexts of soil, crop successions and soil organic C contents. The soil type was the main factor controlling C storage and N leaching while it was crop successions for N saving. Some composts allowed C storage up to 1.1 t C ha-1 yr -1 reaching 63% of C applied. N saving of 74 kg N ha-1 yr -1 were possible with a dried sewage sludge and a compost. N substitution could reach more than 90% of N applied with EOM, these high percentages being related with the indirect effect of EOM on soil OM and the hypothesis made for N substitution An optimisation model was developed to select EOM application scenario for each crop plot (soil x crop succession x initial soil OC content x area) accounting for EOM availability in the region with the objectives of maximising C storage or synthetic N saving or minimising N leaching at the regional scale. Applying preferentially the most stable EOM on soils with the highest potential for C storage i.e. with the highest calcareous and clay contents, up to 0.47 t C ha-1 yr-1 could be stored. Applying preferentially fertilising EOM on crop succession with maize and amending EOM on succession with rapeseed, up to 53 kg N ha-1 yr -1 could be saved.