Soil and plant responses to pyrogenic organic matter: carbon stability and symbiotic patterns Edvaldo Sagrilo Summary Pyrogenic organic matter (PyOM), also known as biochar, is the product of biomass combustion under low oxygen concentration. There is currently a growing interest in research on the use of PyOM as a soil amendment, inspired by the existence of highly fertile, PyOM-rich anthropogenic soils in the Amazon basin. The presence of PyOM in these so-called Amazonian Dark Earths (ADE) in quantities larger than in the non-anthropogenic surrounding soils is considered one of the main reasons for their high fertility. Soil additions of PyOM have been suggested to increase soil fertility and crop yields, simultaneously providing additional important environmental services. The offset of CO2 emissions through sequestration of a larger pool of recalcitrant soil organic carbon (SOC) is one of these services. This would at the same time sustain soil microbial activity, which is directly associated to soil quality, for instance, nutrient cycles and plant growth. This multiple win scenario suggests that the addition of PyOM to the soil would be the solution for the “carbon dilemma”. The dilemma states that the main biological benefits from soil organic matter are a consequence of its decay. Therefore, it is unlikely that increased C sequestration and the benefits from its decay can be simultaneously maximized. Rather than win-win, PyOM would then also be subjected to inevitable trade-offs. Additions of PyOM can modify the turnover rate of native SOC by either accelerating or decelerating its decomposition through a mechanism known as priming. Although positive priming by PyOM has been reported, negative priming has also been found. The higher amount of non-pyrogenic C in ADE, compared to non-anthropogenic surrounding soils has been considered evidence that PyOM can stabilize SOC in the long-term. A complicating issue in studies is that short-term increases in CO2 emission can be due to decomposition of labile PyOM fractions, erroneously suggesting positive priming of SOC. Addition of PyOM can also lead to modifications in the microbial activity and assemblages. Changes in microbial populations can have impacts on their functionality, favouring mutualistic root symbioses such as the arbuscular mycorrhizal fungal (AMF) symbiosis and the rhizobial symbiosis with legumes that is responsible for biological nitrogen fixation (BNF). Although soil amendments with PyOM can stimulate AMF and BNF, results are contrasting and mechanisms are not clear. Most studies of PyOM effects on SOC and on mutualistic root symbioses are from short-term experiments, often conducted in greenhouse or laboratory. Although such studies provide insights in potential factors driving changes in SOC and symbiotic relationships in PyOM-amended soils, they do not assess changes under realistic conditions over periods of time longer that one or a few cropping cycles. Therefore, there is still a gap in our understanding regarding the duration and magnitude of effects over time under field conditions and possible mechanisms involved. This thesis addresses these gaps. The aim of this research was to provide a better understanding of interactions between PyOM and SOC and the factors controlling symbiotic patterns in a tropical soil amended with PyOM. To reach this aim, I combined greenhouse and field studies. I also used meta-analytic methods in order to quantitatively synthesize data in literature. In Chapter 2, I combined the results of 46 studies in a meta-analysis. I investigated changes in CO2 emission patterns from an array of PyOM-amended soils and identified the causes of these changes and the possible factors involved. I showed an overall increase of 29% in CO2 emission from PyOM-amended soils. Such increases were only evident in soils amended with a PyOM-C (PyC):SOC ratio >2. These data are consistent with the hypothesis that increased CO2 emission after PyOM addition is additive and mainly derived from PyOM’s labile C fractions rather than from SOC. Therefore, positive priming is not a main driver of increases in CO2 emission in PyOM-amended soils. This PyC:SOC ratio provided the best predictor of increases in CO2 production after PyOM addition to soil. This meta-analysis indicates (i) the importance of taking into account the amount of applied PyC in relation to SOC for designing future decomposition experiments and that (ii) the recalcitrance of PyOM in soil-PyOM mixtures may be less than usually assumed. A technical problem of separating PyOM-induced priming on SOC from other non-additive interactions is the uncertainty regarding the origin of the respired CO2 (whether from SOC or PyOM). This issue can only be solved with the use of isotopes. In a field study (Chapter 3), I quantified changes in the PyOM and SOC stocks over four soybean cropping cycles (CC) in a sandy Ferralsol, previously supporting a vegetation with C4 plants, amended with different rates of PyOM (0, 5, 10, 20 and 40 Mg ha-1). The PyOM was produced from C3 woody species using traditional pyrolysis methods employed in Northeast Brazil. I used 13C isotopic analysis to discriminate the origin of the C in the soil and quantify the decomposition rates of native SOC and PyOM. I showed that decomposition of traditionally produced PyOM is faster (25-60% within first year) than normally assumed (10-20% within 5-10 years), which was higher than that of native SOC (5-14%). The data indicate preferential decomposition of PyOM compared to native SOC. The intensity of that effect depends on the rate of PyOM applied to the soil. Only on the longer term (>1 yr) addition of PyOM seems to stabilize SOC. In Chapter 4 I explored mechanisms controlling AMF activity and crop yield in PyOM-amended soils through the use of path analysis. I tested the effects of PyOM rates and P fertilization on soybean root colonization by AMF, soil P and plant performance over four cropping cycles (CCs). Data showed a major effect of CC and P, as well an interaction effect of PyOM x CC on mycorrhizal colonization. There was a linear decrease in root colonization by AMF in CC1 with increasing PyOM rates in contrast to a consistent linear increase in CC4. Plant performance was mainly affected by CC, but a significant interactive effect of PyOM x P was also observed on grain yield. Grain yield was highest at high PyOM rates (20 and 40 Mg ha-1) in the P-fertilized treatments in CC4. Soil pH increased in CC1 with increasing PyOM rates, but no effects were observed in CC4. Path analysis indicated that PyOM effects on root colonization by AMF were not mediated by changes in soil pH or P content. My data are consistent with the hypothesis that interference of PyOM in signalling processes is an important driver of change in AMF activity and that positive effects of PyOM on AMF and crop yield develop with time. In Chapter 5, I assessed the effects of PyOM application rates and P fertilization on BNF in soybean inoculated with Bradyrhizobium japonicum over four cropping cycles. Again I observed that CC had a significant main effect on most dependent variables, while PyOM was not a significant source of variation. There was a significant PyOM × CC interaction effect on shoot N concentration. In CC1 shoot N concentration after application of 5 Mg PyOM was significantly lower than that of plants grown on plots to which 10 or 20 Mg PyOM was applied. In CC4 shoot N concentration was not affected by PyOM. The major effect of CC was explained through changes in nutrient management, more specifically the addition of micronutrients in CC3 and CC4. Alleviation of micronutrient deficiency increased BNF and also resulted in a positive effect of P on BNF. I conclude that under conditions of adequate management, PyOM application does not improve BNF in soybean. In Chapter 6 (General Discussion) I synthesize the findings of the previous chapters and use data from additional greenhouse and litterbag field experiments to integrate the results. Data from Chapters 2 and 3 show that if any positive priming occurs due to PyOM addition, it is a small short-term event and does not lead to significant losses of native SOC in the long-term. This was confirmed by data from a 2 yr litterbag experiment, which showed no interaction between decomposition of PyOM and fresh organic matter. Stability of SOC has been considered an ecosystem property rather than a consequence of recalcitrance, but this definition has not yet been extended to PyOM. In this thesis I demonstrated that stability of PyOM can also be influenced by the soil environment. In order to link PyOM effects to SOC and on root symbioses, I performed path analysis integrating root colonization by AMF, SOC content and Ndfa in one model. We found no significant path coefficients linking AMF and BNF. The model indicated a significant positive path coefficient linking AMF root colonization and SOC in CC4, but not in CC1. The data suggest that PyOM may increase SOC stability through increased AMF activity. Soil aggregation and C sequestration are tightly correlated with abundance of AMF in the soil. I propose that the same mechanism through which AMF stabilizes native SOC may also positively influence PyOM stabilization in the long-term. In conclusion, I have shown that main beneficial effects of PyOM on AMF and crop yield develop with time, but in well-managed soils increased crop yield is not a direct consequence of increased AMF due to PyOM addition. Finally, although PyOM additions represent an effective form of sequestering C, positive effects of PyOM on crop yield are likely to occur after partial decomposition of PyOM. Therefore, although some benefits of adding PyOM can be simultaneously obtained (C sequestration and increased crop yield), they cannot be simultaneously maximized. This means that the carbon dilemma can only be partially solved by adding PyOM to the soil.