CHANGE OF SOIL MICRO-ENVIRONMENTS DURING PLANT DECOMPOSITION AND ITS EFFECT ON CARBON AND NITROGEN DYNAMICS By Kyungmin Kim A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Crop and Soil Sciences – Doctor of Philosophy 2021 ABSTRACT CHANGE OF SOIL MICRO-ENVIRONMENTS DURING PLANT DECOMPOSITION AND ITS EFFECT ON CARBON AND NITROGEN DYNAMICS By Kyungmin Kim Detritusphere is one of the most important hotspots; it consists of soil surrounding dead organic materials and is affected by their decomposition and recycling. When assessing C and N dynamics associated with decomposition, micro-environments created near decomposing residues should be considered because the microbial processes are affected by the condition of detritusphere micro-environments, not that of the bulk soil. This could possibly overcome current inaccuracy of global greenhouse gas emission and biogeochemical cycle models. The goal of my Ph.D. research was to investigate the change of soil micro-environmental conditions within detritusphere during plant residue decomposition and to understand their role in and interactions with C and N transformation and dynamics. In Chapter 1, I evaluated the water absorption by decomposing plant roots, based on the finding of water absorption by leaves (a.k.a. sponge effect). In addition to this ‘sponge effect’ in root residues, I assessed the soil moisture gradient created by it by using micro-computed 1the initial soil moisture content and the pore characteristics nearby the roots. It also suggested that the anaerobic micro-environment formed near the roots might influence the N2O emission in the early stage of the decomposition process. In chapter 2, I hypothesized that the influence of moisture redistribution on the N2O emission found in previous chapter is mediated by reduced O2 availability near plant residues. I measured O2 and N2O concentrations in the pores adjacent to leaf and root residues by using electrochemical microsensors. The leaf residues had lower O2 availability near them due to greater water absorption and microbial O2 consumption. Both N2O production and emission were negatively correlated to O2 availability, supporting the initial hypothesis. In Chapter 3, I investigated the fate of C and N during the decomposition of switchgrass roots grown in contrasting soil pores, to test if the micro-environmental characteristics described in former chapters have significant influence on decomposition dynamics. Comprehensive assessments of CO2 and N2O emissions, priming effects, and C and N remaining in soil were performed using dual-isotope labeling (13C and 15N) techniques. There were enhanced influences of soil pore sizes on plant-driven N2O emission, N2O priming, and enzyme activity in in-situ grown root systems. The study also confirmed that detritusphere micro-environments formed in large- pore soils are more favorable for microbial activity and denitrification processes. My dissertation contributed to the characterization of micro-environmental conditions in detritusphere, and their relevance to C and N cycling. It stresses the importance of hotspot micro- environments in predicting greenhouse gas emission and related microbial processes and urges further research to understand the full mechanism and incorporate those in greenhouse gas prediction models. ACKNOWLEDGEMENTS First and utmost, it was my honor to conduct doctoral research with the help of Dr. Sasha Kravchenko. Without her patience and support, I would not have been able to grow this far. She always gave me best research opportunities and chance to collaborate with people from various disciplines and organizations. She encouraged me to participate in academic conferences, seminars, and teaching practices at which I learned many different aspects of scientific research. She is the best advisor and teacher, and scientist I want to emulate. I also thank Dr. Andrey Guber, who was always supportive in designing the experimental setup. All new experimental techniques used in my dissertation were devised by him and I would like to thank him again for his help. Before I even started my Ph.D., I admired Dr. Phil Robertson and his research. It was my honor to have him as my committee member, and I appreciate his outstanding expertise and experience in soil science. I also thank Dr. Nathaniel Ostrom, for introducing me to the world of isotope biogeochemistry. Significant findings in my dissertation research were possible due to delicate isotope analysis, and it is all thanks to his knowledge and help. Dr. Dirk Colbry, the best programmer I know, has always provided practical solutions to data processing and image analysis. Thanks to his advice and guidance, I received a ‘Certificate in Computational Modeling’. I thank again for the support and encouragement from my guidance committee members. I would like to thank Kravchenko Lab members. Maxwell Oerther, who is indispensable to our lab, helped me with field works and all laboratory experiments. He was such a great friend and without his support, I would not have been able to even finish the experiments required for the thesis. Dr. Jenie Gil-Lugo was the best collaborator I met during the Ph.D. program. I will not forget our memories in and outside the campus, and hope we work together again in the future. My senior Dr. Michelle Quigley helped me to adjust to the new environment and gave me practical iv advice to survive in Ph.D. life. I would also like to deliver my gratitude to Dr. Archana Juyal and Dr. Maik Lucas, the best postdoctoral researchers in our lab; and Jinho Lee and Sukhdeep Singh, my fellow doctoral students who always have been encouraging me. I would like to mention special thanks to Dr. Hasand Gandhi and Sam Decamp for their help with isotope analysis. I also thank Dr. Jessica Fry, Dr. Turgut Kutlu, Dr. Linh Nguyen, Dr. Hongbing Zheng, Dr. Weiqing Zhang, and many other researchers who passed through Kravchenko Lab. There are many other colleagues from Kellogg Biological Station and Great Lakes Biological Research Center, and I appreciate their support and collaboration as well. My journey in doctoral degrees was not only about research, but also about adapting to a new culture and language. I would like to thank my family in Korea for their emotional and mental support. Also, I thank the friends I met in Lansing and Ann Arbor, for making good memories in Michigan. I cannot forget the help of Pastor Jeong, Pastor Park, Dr. Pyeon and Dr. Dong and many other ministers in New Hope Baptist Church. I feel grateful for their prayer and care which helped me through the hard times. I would like to give special thanks to Jason Byun for his love and trust in making my Ph.D. journey a success. Finally, I give all the glory to God who guide my life. v TABLE OF CONTENTS LIST OF TABLES ...................................................................................................................... viii LIST OF FIGURES ....................................................................................................................... ix CHAPTER 1: Introduction ............................................................................................................. 1 REFERENCES ...............................................................................................................................5 CHAPTER 2: Water absorption by decomposing plant residues detected via X-ray micro computed tomography approach .....................................................................................................8 Abstract .......................................................................................................................................8 2.1 Introduction ..........................................................................................................................9 2.2 Materials and Methods ......................................................................................................11 2.2.1 Soil microcosm preparation ...........................................................................................11 2.2.2 KI experiment .................................................................................................................14 2.2.3 Incubation experiment ...................................................................................................15 2.2.4 X-ray µCT scanning and image analysis .......................................................................16 2.2.5 Statistical analysis .........................................................................................................20 2.3 Results .................................................................................................................................21 2.3.1 Sponge effect in decomposing roots assessed through iodine distribution patterns .....21 2.3.2 Root decomposition ........................................................................................................24 2.3.3 CO2 and N2O emissions during the incubation .............................................................25 2.4 Discussion ...........................................................................................................................30 2.4.1 Water absorption by decomposing plant roots – the sponge effect ...............................30 2.4.2 Root decomposition and CO2 emission ..........................................................................32 2.4.3 N2O emission ..................................................................................................................33 2.5 Conclusion ..........................................................................................................................35 ACKNOWLEDGEMENTS ........................................................................................................40 REFERENCES ..........................................................................................................................41 CHAPTER 3: O2 depletion and N2O production near decomposing plants directly measured using microsensors ........................................................................................................................48 Abstract .....................................................................................................................................48 3.1 Introduction ........................................................................................................................49 3.2 Materials and Methods ......................................................................................................52 3.2.1 Soil and plant residues ...................................................................................................52 3.2.2 Microsensor experiment .................................................................................................53 3.2.3 Microsensor data processing .........................................................................................56 3.2.4 Plant analysis .................................................................................................................58 3.2.5 Statistical analysis .........................................................................................................61 3.3 Results .................................................................................................................................62 3.3.1 Microsensor experiments ...............................................................................................62 3.3.2 Plant analysis .................................................................................................................70 3.4 Discussion and Conclusion .................................................................................................71 vi 3.4.1 Effect of incorporated switchgrass leaves and roots on N2O production and emission 71 3.4.2 Effect of incorporated switchgrass leaves and roots on N2O temporal dynamic ..........74 3.4.3 O2 at plant residue surfaces and its relationship with N2O ...........................................75 3.4.4 Evaluation of microsensors as a tool for hotspot detection ...........................................77 ACKNOWLEDGEMENTS .......................................................................................................80 REFERENCES ..........................................................................................................................81 CHAPTER 4: Development of soil micro-environment in plant detritusphere and its contribution to C and N cycling ........................................................................................................................87 Abstract .....................................................................................................................................87 4.1 Introduction .........................................................................................................................88 4.2 Materials and Methods ......................................................................................................93 4.2.1 Rhizobox preparation .....................................................................................................93 4.2.2 13C and 15N plant labeling .............................................................................................96 4.2.3 Incubation experiment and N2O, N2, and CO2 analyses .................................................97 4.2.4 Zymography analysis .....................................................................................................99 4.2.5 Plant and soil analysis .................................................................................................102 4.2.6 Calculation methods ....................................................................................................104 4.2.7 Statistical analysis .......................................................................................................105 4.3 Results ...............................................................................................................................107 4.3.1 Switchgrass growth and its effects on soil characteristics ...........................................107 4.3.2 N2O, CO2, and N2 emissions ........................................................................................108 4.3.3 Chitinase activity dynamics during root decomposition ..............................................115 4.3.4 Soil inorganic N, plant-derived C and N, and microbial biomass C ...........................118 4.4 Discussion .........................................................................................................................123 4.4.1 In-situ grown roots in root decomposition studies ......................................................123 4.4.2 Soil pore size as a key driver of N2O emissions ...........................................................123 4.4.3 Relationships between chitinase activity, root decomposition, and N2O emissions ....127 4.4.4 Longevity of high N2O emissions .................................................................................129 4.4.5 Decoupled N2O and CO2 production in large-pore dominated soils ..........................130 4.4.6 N2O priming .................................................................................................................131 4.5 Conclusion ........................................................................................................................132 ACKNOWLEDGEMENTS .....................................................................................................134 REFERENCES ........................................................................................................................135 CHAPTER 5: Conclusion ...........................................................................................................149 vii LIST OF TABLES Table 2.1: Total C, N, and inorganic N in the soil materials used in the study. Values in parenthesis are standard errors (n=3). For all soil characteristics the differences between the small- and large- pore soil materials were not statistically significant (p< 0.05). ....................................................12 Table 2.2: Summary of F-tests for the statistical model fitted to the iodine content data in soil and root media. Shown are F-values and p -values for the main and interaction effects. Please see the model description in the Materials and Methods.........................................................................................22 Table 2.3: Effect of pore size on iodine content at given medium types and moisture levels. Statistical differences marked on Fig. 2.4 are based on the p-value presented here......................22 Table 2.4: Summary of F-tests for the statistical models fitted to iodine content data in 7 soil layers. Shown are F-values and p -values for the main and interaction effects. . ..........................................36 Table 2.5: Effect of pore size on iodine content at given WFPSs and soil layers. Statistical differences marked in Fig. 2.5 are based on the p-value presented here. ......................................37 Table 2.6: Summary of F-tests for the statistical models fitted to volume loss data. Shown are F-values and p -values for the main and interaction effects. ...........................................................................37 Table 2.7: Effect of WFPS at given pore sizes (above) and effect of pore size at given WFPSs (below) on volume loss data. Shown are F-values and p-values for WFPS and pore size effect, when the level of another factor is fixed. Statistical differences marked in Fig. 2.6b and c are based on the p-values presented here. ...............................................................................................................................38 Table 2.8: Summary of F-tests for the statistical models fitted to CO2 and N2O flux data. Shown are F-values and p -values for the main and interaction effects. ............................................................38 Table 2.9: N2O and CO2 emissions with original units. Means and 95% confidence interval was back- transformed from the results of statistical analysis. ..........................................................................39 Table 3.1: Total carbon and nitrogen contents in the plant residues. ...........................................71 Table 4.1: C and N characteristics of soil and roots after plant growth and labeling. Standard deviations in parenthesis. Letters indicate significant differences between large- and small-pore soils (p< 0.05). ............................................................................................................................107 Table 4.2: Summary of statistical analysis for gas emissions. .....................................................108 Table 4.3: Summary of statistical analysis for chitinase activity.................................................117 viii LIST OF FIGURES Figure 2.1: Procedure for the microcosm preparation. .................................................................14 Figure 2.2: Example of a histogram of scanned images (33.069 KeV). The two peaks represent are soil pore and soil mineral portions of the image. The same method was used for the images scanned under 28 KeV. . ...............................................................................................................................18 Figure 2.3: Procedure for assessing the distribution of the liquid as a function of the distance from roots. (a) Original image. (b) Segmented roots. (c) Root dilation results. (d) Final layer used for measuring KI contents. Note that only the soil mineral voxels from the layers identified on (c) were used, while all pore voxels were excluded. (e) Difference image from dual energy scanning (33.269 keV - 33.069 keV) (f) Thresholded KI image reflecting the distribution of the liquid. (g) Examples of the layers overlaying the original image. .................................................................19 Figure 2.4: Water absorption by dry decomposing roots assessed though iodine gradients. Shown are an example of a 3D visualization of a root, soil, and iodine (a), percent of iodine occupied voxels within the root voxels at 50% WFPS (b) and 75% WFPS (c), and percent of iodine occupied voxels within the soil matrix voxels at 50% WFPS (d) and 75% WFPS (e). Symbol ** marks statistically significant differences in iodine levels between large- and small-pore microcosms (p< 0.05). .............................................................................................................................................23 Figure 2.5: Percent of iodine occupied voxels within the soil matrix voxels as a function of the distance from the roots. Gray dashed line is the average iodine content in the bulk soil matrix within the same WFPS. Symbol ** marks statistically significant difference between iodine levels in large- and small-pore microcosms at 0-48 μm layer (p< 0.05). ...............................................24 Figure 2.6: Root decomposition during the 21-day incubation. Shown are an example of a 3D visualization of a root before (left) and after (right) incubation (a), and the root volume losses (%) in the large- and small-pore microcosms at 50% WFPS (b) and 75% WFPS (c). Shown are the treatment means, the error bars represent standard errors (n=4). Volumes were calculated from the number of voxels in μCT image stacks. Symbol ** indicates statistically significant differences between pore size treatments at the same WFPS (p< 0.05), and different letters indicate statistically significant differences between WFPSs at the same pore size group (p< 0.10). ...............................25 Figure 2.7: CO2 and N2O fluxes during the 21-day incubation in large- and small-pore size microcosms at the two studied WFPS, grouped by moisture content. (a) CO2 emission at 50% WFPS, (b) CO2 emission at 75% WFPS, (c) N2O emission at 50% WFPS, and (d) N2O emission at 75% WFPS. Shown are the treatment means, the error bars represent standard errors (n=5). Symbols * and ** mark significant differences between pore sizes within the same day (p< 0.10 and p< 0.05, respectively). ........................................................................................................................................................27 Figure 2.8: Cumulative CO2 and N2O emission during the incubation. Symbols * and ** indicate the differences between pore sizes with 0.10 and 0.05 significance level, respectively. The different lowercase letters represent the differences between moisture contents in the same pore size group. ........................................................................................................................................................28 ix Figure 2.9: N2O fluxes during 21-day incubation in large- and small-pore size microcosms at the two studied WFPSs, grouped by pore-size. Shown are the treatment means, the error bars represent standard errors (n=5). Symbol ** marks the differences between WFPSs within the same day (p< 0.05). ..............................................................................................................................................29 Figure 2.10: N2O emission from control soil during the incubation. Symbol ** marks the difference between the pore sizes within the same day (p< 0.05). ................................................................30 Figure 3.1: Experimental setup for microsensor measurements. (a) Flattened residue fragments in the holder. Holder was used to fix the location of residue fragments in the soil box. (b) The experimental setup. It shows boxes containing soil with incorporated residues, microsensors inserted into the boxes, and tubing connecting the air chambers above the soil boxes with the PAS device for measuring air concentrations of N2O and CO2. (c) Schematic representation of the microsensor locations in vicinity to the residue within the soil box. ......................................54 Figure 3.2: Dynamics of N2O concentrations in the soil microcosms with (a) leaf and (b) root residues. Black dotted lines indicate the timepoint at which the lag time and the peak (maximum) concentration were determined. Colored area under each curve presents cumulative N2O production for 2 days. Red circles mark an example of artificial fluctuations in one of the experimental runs. .........................................................................................................................57 Figure 3.3: Examples of (a) boxes with soil and plant residues used for soil zymography and (b) resultant zymography images. Yellow dotted rectangles on (a) mark the areas that were subjected to zymography, i.e., membrane placement. The white dotted rectangle on (b) encompasses the area used to calculate the enzyme activity for the incubated plant residue. .........................................60 Figure 3.4: (a) Microsensor recorded O2 concentrations near switchgrass residues and control soil during the experiment. Water addition started at time 0. (b) Cumulative O2 depletion plotted vs. cumulative N2O production during the first 2 days of the experiment. Dotted line is the linear regression model fitted to the data (p< 0.10, one-tailed test). ......................................................63 Figure 3.5: Boxplot of quantitative measurements from microsensor and Photoacoustic Spectroscopy. (a) Peak N2O production - the maximum N2O concentration observed from 5 days of microsensor recordings, (b) Cumulative N2O production - the area under the microsensor curves and above zero for 2 days, (c) N2O emission –N2O from the surface of the soil measured from headspace, (d) lag N2O production – the time elapsed from the start of the soil wetting until the peak production, (e) minimum O2 concentration - the lowest O2 concentration observed from 5 days of microsensor recordings, and (f) O2 depletion for 2 days - the area under the base O2 concentration. ** and * indicate significant differences between leaves and roots (p< 0.05 and 0.10). Black dots are individual observations from each run. The coefficients of variation were presented as percentage. ................................................................................................................64 Figure 3.6: (a) Peak N2O production plotted vs. cumulative N2O production for day 1 of the experiment. (b) The mass of plant residue incorporated into soil vs. peak N2O production from the residue. Dotted lines represent the regression models fitted to the entire data set (black) and to leaf data only (green). The regression model for roots (blue) was not statistically significant. Symbols * and ** mark the models statistically significant p< 0.10 and 0.05, respectively. The red circle marks the outlier data point that was not included in the regression analysis. .............................66 x Figure 3.7: (a) N2O and (b) CO2 emission rates from the soil boxes with incorporated leaf and root residues, measured using Photoacoustic Spectroscopy. Vertical lines represent standard errors. Asterisk * indicates significant differences between leaves and roots (p< 0.10). ........................68 Figure 3.8: Relationship between CO2 and N2O emission (day 1-5). Dotted lines are regression models for leaf (green) and root (blue). Symbol *** marks the significance of the slopes (p< 0.01). ........................................................................................................................................................69 Figure 3.9: Relationship between cumulative N2O productions (measured from soil pore using microsensor) in and emissions (measured from headspace using photoacoustic spectroscopy) from the microcosms (a) for 1 day, (b) 2 days, and (c) 3 days of the experiment. Dotted lines represent linear regression models. All regressions were statistically significant at p< 0.01 and 0.05 (marked with *** and **). There were no significant differences between regression slopes of leaves and roots. ..............................................................................................................................................69 Figure 3.10: (a) Average water absorption levels by leaf and root residues. The difference between leaves and roots is significant at p< 0.10. (b) Average enzyme (β-glucosidase) activity at the surface of the plant residues at day 1 and 3 of the experiment. Different letters mark significant differences between the days within each residue type (p< 0.10). Symbol ** indicates the significant differences between residue types at a given day (p< 0.05). Vertical lines represent standard errors. ..............................................................................................................................70 Figure 4.1: Switchgrass 15N and 13C labeling procedure. (a) Rhizoboxes placed on a rack tilted at 50⁰ to induce root growth on the surface of the box for subsequent zymography. (b) Rhizobox placed on the beaker containing 15N Hoagland solution to label the plant without directly adding 15 N to the soil. (c) Rhizoboxes and beakers sealed with aluminum foil to avoid evaporation. (d) Rhizoboxes and beakers placed in the chambers for pulse labeling the plants with 13CO2. .........96 Figure 4.2: Processes to obtain the reference image of roots. (a) Raw image taken prior to zymography measurement. (b) Raw image cut into region of interest. (c) Image after background removal, default adjustment of brightness and contrast, and Gaussian blur. (d) Image after default threshold applied. (e) Result of particle analyzer that separates all particles detected in the image. (f) Final reference image obtained by selecting root particles. ...................................................101 Figure 4.3: Illustration of zymography procedures. (a) Example of a raw image that captures the florescence developed from enzyme-substrate reaction. A white dashed box in (a) represents the region of interest used for further zymography analysis. (b) A map of enzyme activity calculated from the set of raw images taken every 5 minutes. (c) A final reference image in which black and white indicate the root and soil, respectively. (d) Chitinase activity on the surface of the roots – average chitinase activity on the root area. (e) Chitinase activity in soil – average chitinase activity on the soil area. Brown area represents root (d) or soil (e) surfaces that has no enzyme activity. ......................................................................................................................................................102 Figure 4.4: Dynamics of N2O emission during the incubation: (a) total N2O emission rate at 40% WFPS and (b) 70% WFPS; and (c) the fraction of root-derived N2O across both WFPS treatments. Error bars are standard errors of the mean. ‘Large’ and ‘Small’ in the legends indicate the soils dominated by large (> 30 µm Ø) and small (< 10 µm Ø) pores, respectively. Controls are the rhizoboxes that were not planted and did not contain switchgrass roots. The gray box presents significant results of the factors and their interactions. Red asterisks * and *** mark statistical xi significance (p< 0.10 and 0.01). P-values of the factors and their interactions were presented in Table 4.2 (supplementary material). ...........................................................................................110 Figure 4.5: Temporal dynamics of CO2 emission and its fraction derived from the decomposing roots. Total CO2 emission rate (a) at 40% WFPS and (b) at 70% WFPS, and (c) fraction of root- derived CO2. Error bars are standard errors of the mean. Gray boxes present significant results of the factors and their interactions. Red asterisks *, ** and *** mark statistically significant differences between large- and small-pore soils (p< 0.10, 0.05, and 0.01). ...............................111 Figure 4.6: Cumulative N2O emissions from the (a) large- and (b) small-pore soil; and cumulative CO2 emission from the (c) large- and (d) small-pore soil after 21 days of incubating the rhizoboxes with in-grown switchgrass roots. Control refers to unplanted soil boxes incubated under each WFPS. Differences between soil-derived gas emission and the control gas emission (priming effect) are presented as red arrows. Red asterisks ** and *** mark the cases where the priming effect was significantly different from 0 (p< 0.05 and 0.01). NS stands for ‘Not Significant’. Letters indicate the significant differences between root-derived cumulative gas emissions (white, p< 0.05) and priming effect (red, p< 0.10) at a given moisture level. Error bars are standard errors of the mean. Note the different y-axis scales between (a) and (b). ..................................................................112 Figure 4.7: Dynamics of N2O priming effect at (a) 40% WFPS and (b) 70% WFPS in the large- and small-pore dominated soils. Red asterisks ** and *** indicate significant differences between large- and small-pores (p< 0.05 and 0.01). Letters indicate the significant differences between moisture levels at given soil materials and time (p< 0.05). ........................................................113 Figure 4.8: Relationship between total CO2 and total N2O emission rates in the two studied soils and moisture levels. Regression line of small-pore dominated soils was presented in green. Blue plots present the average emission rates of the unplanted controls. R2 was reported separately for large-pore and small-pore soils (p< 0.01). Regression slope from the large pores was not presented due to low R2. ..............................................................................................................................114 Figure 4.9: Dynamics of N2 emission during the root decomposition in (a) large-pore soils and (b) small-pore soils. Shown are the differences between the planted soil and soil without plants (control). Asterisks ** mark the cases when the differences were greater than zero (p< 0.05). Error bars are standard errors of the mean. Gray lines show the actual N2 emission (average of the two moisture levels) from planted soils. ............................................................................................115 Figure 4.10: Dynamics of chitinase activity on root and soil surfaces in the large- and small-pore dominated soils at the two studied soil moisture levels. Example zymograms are shown on (a) and (d); root chitinase activity on (b) and (e), and soil chitinase activity on (c) and (f) for 40% and 70% WFPS, respectively. Asterisks *, **, and *** indicate the significant differences between the two soils at a given day of incubation and moisture content, at significance levels of 0.1, 0.05, and 0.01, respectively. The letters mark differences between time (incubation day) within given soil and moisture. Error bars are standard errors of the mean. .................................................................116 Figure 4.11: Relationships between chitinase activity and emissions of root-derived N2O or CO2. (a) Correlation between root chitinase activity and root-derived N2O emission rate, and (b) root- derived CO2 emission rate. (c) Correlation between soil chitinase activity and soil-derived N2O xii emission rate, and (d) soil-derived CO2 emission rate. Black dash lines represent linear regression models. N.S indicates that the regression slope was not significantly different from zero. .......118 Figure 4.12: Dynamics of soil (a) NH4+ and (b) NO3- during rhizobox incubation. Asterisks ** and *** mark significant differences between large- and small-pore dominated soils at the given day of incubation at p of 0.05 and 0.01, respectively. Letters mark the differences between days of incubation in the given soil (p< 0.05). Error bars are standard errors of the mean. ...................119 Figure 4.13: Dynamics of (a) dissolved organic C (DOC) and (b) dissolved organic N (DON) during the rhizobox incubation. Asterisks *, **, and *** indicate the significant differences between large-pore and small-pore soil (on the figure) and the significant effect of the factor at p levels of 0.10, 0.05, and 0.01. .....................................................................................................120 Figure 4.14: Microbial biomass C (MBC) derived from soil organic matter (gray) and labeled decomposing root residues (blue) on day 39 of the incubation. Asterisk ** indicates significant effect of the root-derived MBC at 0.05 error probability level. Error bars are standard errors of the mean. ...........................................................................................................................................121 Figure 4.15: (a) Total soil C and (b) total soil N, and fractions of (c) root-derived total C and (d) root-derived total N in the root-associated soil and in the bulk soil on day 39 of the incubation. Symbol ** and *** indicates significant difference of the factor at 0.05 and 0.01 error probability level. Error bars are standard errors of the mean. .......................................................................122 Figure 4.16: Graphical representation of the micro-environments in the rhizoboxes dominated with large-pore soil (left) and small-pore soil (right). Large-pore dominated soil has higher fungal growth, enzyme activity, and microbial activity near roots due to rhizosphere legacy. It leads to faster decay of roots and SOM in large-pore dominated soil, where larger amount of both root- driven and soil-driven organic substances can be used as substrates for N2O emission. Sponge effect (water absorption by decaying roots) is more prominent in large-pore dominated soil because of its lower water retention. Anaerobic detritusphere (near roots) and aerobic surrounding large pores together make optimum condition for denitrification and increase both root-driven and soil- driven N2O emission. ..................................................................................................................124 xiii CHAPTER 1: Introduction Global climate change is now one of the most important issues to be solved for human life, and the United Nations designated climate action as a major goal for sustainable development (SDG, 2019). Soil is considered as a possible and feasible contributor to mitigate climate change - for example, ‘4 per mille Soils for Food Security and Climate’ launched at the COP21 suggests that sequestering atmospheric carbon (C) as soil C by 4 ‰ per year can compensate the global greenhouse gas (GHG) emissions from human activity (Minasny et al., 2017). Despite evidence of soil’s current substantial contribution to GHGs and its potential to mitigate GHGs, quantifying emissions and reductions of GHGs yet remains as a substantial challenge (Paustian et al., 2016). One of the reasons is that the current models that project feedbacks of climate and soil nutrient cycles omit key biogeochemical mechanisms of soil C (Wieder et al., 2013). For example, direct microbial control over soil C dynamics is not included, and the ability to project soil’s response in a changing environment remains unresolved (Ise and Moorcroft, 2006; Manzoni and Porporato, 2009). Better understandings of underlying microbial processes and environmental conditions in soil can improve such models for projection and mitigation of GHGs. Soil is a complex and heterogeneous material. Different shapes and arrangements of soil minerals and pores in addition to locations of newly added organic materials result in heterogeneous microscale distributions of water, air, organic substrates, and microorganisms. This spatial heterogeneity plays an important role in persistence of soil organic C (Keiluweit et al., 2017; Lehmann et al., 2020) and nitrous oxide (N2O) production and emission (Kravchenko et al., 2017). Varying micro-scale environments suggest that there can be different pathways and process rates of microbial activity even in the same body of soil. High micro-scale soil heterogeneity might explain the high variability and uncertainty of current earth system models. Still, measuring 1 environmental conditions at microscale and scaling them up to regional and global climate models requires tremendous time and financial resources. The most challenging, yet most effective, approach could be gaining understanding of micro-environmental conditions in “soil microsites” where most microbes reside and interact with the physicochemical environment. Such soil microsites where microbial activity is most vigorous and consequently contributes to a significant proportion of GHG emissions from the soil are referred to as “hotspots”. Microbial hotspots are characterized by faster process rates and more intensive interactions compared to the average soil conditions (Kuzyakov and Blagodatskaya, 2015). Major hotspots in soil are found in the rhizosphere (Hinsinger et al., 2009), deteritusphere (Kögel-Knabner, 2002), biopores (Schrader et al., 2007), and on aggregate surfaces (Kaiser and Kalbitz, 2012). Detritusphere is one of the most important hotspots; it consists of soil surrounding dead organic materials and is affected by their decomposition and recycling (Kuzyakov and Blagodatskaya, 2015). The micro-scale environment formed near decomposing residues can be distinctly different from the bulk soil characteristics. For example, decomposing plant residues induce a gradient of soil organic C and N to an extent of 1.5 – 3.0 mm (Gaillard et al., 2003; Gaillard et al., 1999; Poll et al., 2006). The detritusphere micro-environment is also characterized by unique microbial community structure (Nicolardot et al., 2007) and high enzyme activity (Spohn and Kuzyakov, 2014). These findings imply that the basic assumption that the bulk soil properties regulate the microbial processes can be misleading, which can possibly contribute to inaccurate predictions of GHGs from soils. Rather, the micro-environments created near decomposing residues should be considered while assessing C and N dynamics associated with decomposition, because the microbial processes are affected by the condition of these micro- environments, not that of the bulk soil. In other words, studying micro-environments in hotspots 2 of high environmental significance can be more relevant to the microbial processes and nutrient dynamics at greater spatial scales, such as those of soil horizon or even the entire soil profile. The presence of plant residues within the soil matrix can change moisture distributions and gas concentrations in pores near the residues (Kravchenko et al., 2017; Li et al., 2016), creating unique micro-environments distinguished from the bulk soil. Plant leaf’s water retention capacity (Iqbal et al., 2013) leads to absorption of nearby pore water (Kravchenko et al., 2018; Kravchenko et al., 2017), creating anaerobic conditions within the detritusphere. Also, the interactions between residues and pore characteristics can regulate the fate of C and N from dead plant residues by formulating distinct environmental conditions (Toosi et al., 2017). Although several studies have discussed these micro-environments of detritusphere hotspots, limited studies have provided experimental data on how exactly they are created, and their potential influence in a larger-scale biogeochemical cycle. Most importantly, the detritusphere micro-environments’ actual contribution to important microbial processes related to GHG emissions has not yet been accurately evaluated. The goal of my Ph.D. research was to investigate the change of soil micro-environmental conditions within detritusphere during plant residue decomposition and to understand their role in and interactions with C and N transformation and dynamics. In Chapter 1, I evaluated the water absorption by decomposing plant roots, based on the finding of water absorption by leaves (a.k.a. sponge effect). In addition to this ‘sponge effect’ in root residues, I assessed the soil moisture gradient created by it by using micro-computed tomography. I hypothesized that the moisture redistribution near decomposing roots depends on the initial soil moisture content and the pore characteristics nearby the roots. The influence of micro-environments on CO2 and N2O emission was explored as well. In chapter 2, I hypothesized that the influence of moisture redistribution on 3 the N2O emission explored in previous chapter is mediated by reduced O2 availability near plant residues. By measuring O2 and N2O concentrations in the pores adjacent to leaf and root residues using electrochemical microsensors, I investigated the relationships among O2 availability near residues, N2O production from residues, and total N2O emission. Temporal dynamics of those gas concentrations were measured in different type of residues (leaf vs. root). In Chapter 3, I investigated the fate of C and N during the decomposition of switchgrass roots grown in contrasting soil pores, to test if the micro-environmental characteristics explored in former chapters have significant influence on decomposition dynamics. Comprehensive assessments of CO2 and N2O emissions, priming effects, and C and N remaining in soil were performed using dual-isotope labeling (13C and 15 N) techniques. Biological influences of soil pore sizes and moisture contents on micro-environments was quantified using 2-dimensional mapping of extracellular enzymes (zymography). Combination of isotopic and spatial measurements allowed to explore complex mechanisms contributing to GHG emission from the detritusphere influenced by former rhizosphere. 4 REFERENCES 5 REFERENCES Gaillard, V., Chenu, C. and Recous, S., 2003. Carbon mineralisation in soil adjacent to plant residues of contrasting biochemical quality. Soil biology and biochemistry, 35(1): 93-99. Gaillard, V., Chenu, C., Recous, S. and Richard, G., 1999. Carbon, nitrogen and microbial gradients induced by plant residues decomposing in soil. European Journal of Soil Science, 50(4): 567-578. Hinsinger, P., Bengough, A.G., Vetterlein, D. and Young, I.M., 2009. Rhizosphere: biophysics, biogeochemistry and ecological relevance. Plant and soil, 321(1): 117-152. Iqbal, A., Beaugrand, J., Garnier, P. and Recous, S., 2013. Tissue density determines the water storage characteristics of crop residues. Plant and Soil, 367(1): 285-299. Ise, T. and Moorcroft, P.R., 2006. The global-scale temperature and moisture dependencies of soil organic carbon decomposition: an analysis using a mechanistic decomposition model. Biogeochemistry, 80(3): 217-231. Kaiser, K. and Kalbitz, K., 2012. Cycling downwards–dissolved organic matter in soils. Soil Biology and Biochemistry, 52: 29-32. Keiluweit, M., Wanzek, T., Kleber, M., Nico, P. and Fendorf, S., 2017. Anaerobic microsites have an unaccounted role in soil carbon stabilization. Nature communications, 8(1): 1-10. Kögel-Knabner, I., 2002. The macromolecular organic composition of plant and microbial residues as inputs to soil organic matter. Soil biology and biochemistry, 34(2): 139-162. Kravchenko, A., Fry, J. and Guber, A., 2018. Water absorption capacity of soil-incorporated plant leaves can affect N2O emissions and soil inorganic N concentrations. Soil Biology and Biochemistry, 121: 113-119. Kravchenko, A. et al., 2017. Hotspots of soil N2O emission enhanced through water absorption by plant residue. Nature Geoscience, 10(7): 496-500. Kuzyakov, Y. and Blagodatskaya, E., 2015. Microbial hotspots and hot moments in soil: concept & review. Soil Biology and Biochemistry, 83: 184-199. Lehmann, J. et al., 2020. Persistence of soil organic carbon caused by functional complexity. Nature Geoscience, 13(8): 529-534. Li, X., Sørensen, P., Olesen, J.E. and Petersen, S.O., 2016. Evidence for denitrification as main source of N2O emission from residue-amended soil. Soil Biology and Biochemistry, 92: 153-160. Manzoni, S. and Porporato, A., 2009. Soil carbon and nitrogen mineralization: theory and models across scales. Soil Biology and Biochemistry, 41(7): 1355-1379. 6 Minasny, B. et al., 2017. Soil carbon 4 per mille. Geoderma, 292: 59-86. Nicolardot, B., Bouziri, L., Bastian, F. and Ranjard, L., 2007. A microcosm experiment to evaluate the influence of location and quality of plant residues on residue decomposition and genetic structure of soil microbial communities. Soil Biology and Biochemistry, 39(7): 1631-1644. Paustian, K. et al., 2016. Climate-smart soils. Nature, 532(7597): 49-57. Poll, C., Ingwersen, J., Stemmer, M., Gerzabek, M. and Kandeler, E., 2006. Mechanisms of solute transport affect small‐scale abundance and function of soil microorganisms in the detritusphere. European Journal of Soil Science, 57(4): 583-595. Schrader, S., Rogasik, H., Onasch, I. and Jégou, D., 2007. Assessment of soil structural differentiation around earthworm burrows by means of X-ray computed tomography and scanning electron microscopy. Geoderma, 137(3-4): 378-387. SDG, U., 2019. Sustainable development goals. The energy progress report. Tracking SDG, 7. Spohn, M. and Kuzyakov, Y., 2014. Spatial and temporal dynamics of hotspots of enzyme activity in soil as affected by living and dead roots—a soil zymography analysis. Plant and Soil, 379(1-2): 67-77. Toosi, E., Kravchenko, A., Guber, A. and Rivers, M., 2017. Pore characteristics regulate priming and fate of carbon from plant residue. Soil Biology and Biochemistry, 113: 219-230. Wieder, W.R., Bonan, G.B. and Allison, S.D., 2013. Global soil carbon projections are improved by modelling microbial processes. Nature Climate Change, 3(10): 909-912. 7 CHAPTER 2: Water absorption by decomposing plant residues detected via X-ray micro computed tomography approach* Abstract The “sponge effect”, or water absorption by incorporated plant leaf residues, was recently identified as one of the mechanisms that drives activity in microbial hotspots. I explored the presence of the sponge effect in plant root residues, and its role in root decomposition and associated N2O and CO2 emissions. Young soybean (Glycine max) plants were grown in microcosms with two soil materials dominated by (i) large (> 30 µm Ø) and (ii) small (< 10 µm Ø) pores. After termination, the microcosms with the decomposing roots were incubated at 50% and 75% water-filled pore space (WFPS) soil moisture levels. Root decomposition, water absorption by the decomposing roots, and water redistribution were quantified using X-ray computed micro-tomography, including dual-energy scanning. The results demonstrated occurrence of the sponge effect in decomposing, young, in-situ grown soybean roots and sharp gradients in the distribution of the added water within ~150 µm distance from the decomposing roots. At 50% WFPS the large pore soil emitted 185% more N2O than the small pore soil; and, during the first 5 days of incubation, more N2O than the large pore soil at 75% WFPS. This finding indicates that the decomposing roots acted as hotspots of N2O production, potentially due to sponge effect and associated anoxic conditions. This study suggests that the interactions between pore characteristics and soil moisture can play a significant role in defining the contribution of detritusphere, specifically, decomposing young roots, to soil biogeochemical processes, including microbial activity and denitrification dynamics. * Originally published as: Kim, K., Guber, A., Rivers, M., & Kravchenko, A. (2020). Contribution of decomposing plant roots to N2O emissions by water absorption. Geoderma, 375, 114506. doi: 10.1016/j.geoderma.2020.114506 8 2.1 Introduction While residue incorporation brings multiple benefits to soil fertility and sustainability (Lehtinen et al., 2014), crop residues can also stimulate the emission of greenhouse gases (GHGs) from the soil (Baggs et al., 2000; Jin et al., 2014 ; Köbke et al., 2018). Roots account for 13 – 67 % of the whole plant biomass (Roy et al., 2001; Bolinder et al., 2002), thus mass of root residues remaining in the soil after the harvest can be substantial, reaching 0.5–2 Mg·ha-1 (Tufekcioglu et al., 1998). Water absorption by dry plant residue incorporated into the soil (hereafter referred to as “sponge effect”) affects decomposition (Iqbal et al., 2013) and has been recently identified as one of the mechanisms that enhances local anoxic conditions and promotes hotspots of N2O emission (Kravchenko et al., 2017). Decomposition of incorporated corn and soybean leaves was faster when leaves were surrounded by soil with prevalence of > 30 µm Ø pores as opposed to soil dominated by ~10 µm Ø pores (Negassa et al., 2015; Kravchenko et al., 2017). Greater sponge effect, i.e., higher (up to 120%) water absorption by the leaf residues, in the soil dominated by the large pores was suggested as one of the drivers of faster decomposition (Kravchenko et al., 2017; Kravchenko et al., 2018). Lower water retention capacity in large-pore dominated soil allowed greater sponge effect compared to small-pore dominate soils. However, the past work has been conducted using only aboveground plant residues. Examining the sponge effect in other types of decomposing plant tissues, especially in roots, is necessary to understand the role of N2O hotspots induced by plant residue decomposition in total N2O emissions from the soil. Roots distinctly differ from aboveground biomass in terms of tissue traits and chemical composition (Kumar and Goh, 1999; Moretto et al., 2001; Kuzyakov et al., 2007; Begum et al., 2014). Such differences can cause dissimilarities in decomposition rates and C and N dynamics 9 between the aboveground residues and roots. Lower residue quality, higher C:N ratio, and higher lignin content of root residues lead to lower C mineralization, lower denitrification, and slower decomposition compared to the leaf residues (Vanlauwe et al., 1996; Velthof et al., 2002; Bird et al., 2008; Hansson et al., 2010; Steffens et al., 2015). Roots also differ from the incorporated aboveground residues in terms of the impact they make on the physical and chemical properties of adjacent soil. Aboveground residues are typically incorporated by mixing with tillage-disturbed soil, while roots grow, die, and decompose in situ, and even tillage does not fully separate decomposing roots from their associated soil. During growth, roots alter soil properties in their immediate vicinity by changing soil density, hydraulic properties, and C and N levels, as well as composition of microbial communities (Angers and Caron, 1998; Carminati and Vetterlein, 2012; Meier et al., 2017). The micro-environmental conditions created when the root was alive (i.e., rhizosphere) directly affect the conditions at which it decomposes (i.e., detritusphere). The differences in compositions and in the properties of the surrounding soil can potentially lead to differences in contributions of incorporated above- and belowground plant residues to CO2 and N2O emissions. However, the question of how much the decomposing roots contribute to GHG emissions has not received as much attention as the contributions of the incorporated aboveground plant residues (Hobbie et al., 2010; Chirinda et al., 2012). Moreover, in most studies the roots for incubation experiments were taken from and washed of their native soil before being placed into incubated microcosms (Jung et al., 2011; Bai et al., 2016; Wang et al., 2017; Shahbaz et al., 2018). Thus, the potential impact on emissions from the micro-environmental soil conditions created by the in-situ grown roots was neglected. 10 The objectives of this study were: (i) to quantify the magnitude of the sponge effect in in- situ grown roots decomposing in soils with contrasting pore size distributions and moisture content levels, (ii) to examine whether sponge effect changes water distribution patterns in the vicinity of decomposing root residues, and (iii) to quantify N2O and CO2 emissions from soil with decomposing root residue at contrasting soil pore size distributions and moisture content levels. 2.2 Materials and Methods 2.2.1 Soil microcosm preparation Soil used in the study was collected from experimental plots of the biologically based agronomic treatment (corn-soybean-wheat rotation) of Long Term Ecological Research site at the W.K. Kellogg Biological Station (KBS), Michigan, U.S.A. (85⁰24' W, 42⁰24' N). The soil was Kalamazoo loam (fine-loamy, mixed, mesic, Typic Hapludalf), developed on glacial outwash. Composite soil samples for the study were collected from 0 – 15 cm depth and air-dried. The treatment design of the study consisted of two experimental factors: (i) prevalent pore size, with two levels, namely, soils with prevalence of > 35 µm Ø and < 10 µm Ø pores; and (ii) soil moisture content during the incubation, with two levels, namely, 50% and 75% WFPS. To generate the two soil materials with contrasting pore sizes, I followed a method described by Toosi et al. (2017). The soil material with prevalence of > 35 µm Ø pores was created from 1 – 2 mm Ø fraction by sieving the air-dried soil. The soil matrial with prevalence of < 10 µm Ø pores was created from the 1 – 2 mm Ø fraction by a series of gentle grindings using mortar and pestle, followed by sieving through a 0.053 mm sieve. The remaining particles, primarily small stones, were re-collected and completely ground using a shatter box to make it pass through the 0.053 mm sieve. The use of sequential grinding to procure most of the small pore material minimized the negative effects of soil grinding on microorganisms. The two soil fractions were 11 hereafter referred to as large-pore and small-pore soils. Soil organic C and total N of the two materials were measured using Costech elemental combustion system (Costech Analytical Technologies, U.S.A) with 3 replicates. For inorganic N, soils were extracted using 1M KCl with soil-solution ratio of 1:5. Soil extracts were then mixed with premade reagent packets (Hach GSA, U.S.A). Salicylate method was used for ammonium and cadmium reduction method was used for nitrate (Sinsabaugh et al., 2000; Doane and Horwáth, 2003). The level of inorganic N was then determined using SYNERGY H1 (BioTek, U.S.A). Using the large soil fractions to produce the small fractions enabled us to minimize differences in soil mineralogy and microbial properties. The two soil materials were not significantly different in terms of their soil organic C, total N, NO3, and NH4 contents (Table 2.1). The levels of the soluble organic carbon where 292 (std error 25) µg C · g-1 soil in the large-pore and 344 (std error 19) µg C · g-1 soil in the small-pore soil material and not significantly different from each other (p< 0.05) (reported as supplemental information by Kravchenko et al., 2017). Table 2.1: Total C, N, and inorganic N in the soil materials used in the study. Values in parenthesis are standard errors (n=3). For all soil characteristics the differences between the small- and large-pore soil materials were not statistically significant (p< 0.05). Soil characteristic Small-pore soil Large-pore soil Total C (%) 1.46 (0.08) 1.69 (0.26) Total N (%) 0.14 (0.003) 0.16 (0.02) + NH4 (mg N · kg-1 soil) 7.11 (1.78) 7.00 (0.51) NO3- (mg N · kg-1 soil) 7.81 (0.09) 7.69 (0.06) A total of 40 microcosms were built by packing plastic cylinders (8 mm Ø, 40 mm height) with soil of the two studied materials to a bulk density of 1.1 g cm-3. The relatively small size of the microcosms was chosen in order to accommodate quantification of the root decomposition via X-ray µCT. A disadvantage of working with in-situ grown roots is the unknown initial mass of 12 roots in the system, thus the unknown loss of root mass during decomposition. To quantify root decomposition during the incubation I scanned all microcosms before and after the incubation (as described in section 2.2.4). Then the loss of the root volume was obtained as the difference in root volumes before and after the incubation µCT images. While this approach enabled acceptable quantification of the root volume loss, the size of the microcosms had to be kept relatively small to provide sufficiently high scanning resolution. On top of each microcosm cylinder, I placed a larger cylinder (16 mm Ø, 5 mm height) and filled it with loose soil. The purpose of the large cylinder was to accommodate the initial growth of the soybean (Glycine max) seeds, which required more space than what was available within the small cylinders. The seeds were germinated for 4-5 d on wet cloth. After germination, one soybean seed was carefully inserted in the middle of each large cylinder (Fig. 2.1). During the plant growth the soil moisture within the microcosms was initially adjusted to ~50% of WFPS and 0.2 mL water was added on a daily basis to maintain optimal condition for the plant growth. No fertilizers were used. The plants were allowed to grow for 4 days, the period of time during which the soybeans roots grew through the entire length of the studied microcosms. Then the plants were cut, and the microcosms were air-dried for 5 days. 13 Figure 2.1: Procedure for the microcosm preparation. 2.2.2 KI experiment Eight microcosms (2 replicates of each pore size and WFPS treatment combination) were used to quantify the sponge effect and the spatial patterns in water distribution within the decomposing roots. Upon root termination and air-drying, 10% potassium iodide (KI) solution was added to each microcosm. Iodine is a chemical dopant that enhances the contrast of liquid phase against other phases during X-ray µCT scanning (Wildenschild et al., 2002; Wildenschild and Sheppard, 2013). The volume of the KI solution added to the 50% and 75% WFPS treatment microcosms was equal to the respective amounts of water that were added to the counterpart microcosms of these WFPS treatments in the incubation experiment (described in section 2.2.3). 14 The microcosms were allowed to equilibrate with added KI solution for ~24 hr and then subjected to dual-energy X-ray µCT scanning (described in section 2.2.4). 2.2.3 Incubation experiment The incubation experiment was a full factorial design with two factors: pore size with two levels (prevalent large and small pores), and water content with 2 levels (50% WFPS and 75% WFPS). Due to loss of 5 microcosms during handling and transporting, a total of 27 microcosms (5-8 replicates of each pore size and WFPS treatment combination) were used to assess the CO2 and N2O emissions and to quantify the root volume loss during the decomposition. To determine the root volume loss the microcosms were µCT scanned twice: first, before and then, after the incubation. Upon root termination and air-drying, the microcosms were subjected to X-ray µCT scanning (as described in section 2.2.4). Then, for the incubation, distilled water was added to the tops of the microcosms to achieve the desired WFPS levels: 50% WFPS and 75% WFPS. Each microcosm was placed in a 130 mL Mason jar, and a small water-filled plate was placed within the jar along with the microcosm for maximizing air humidity and reducing evaporation from the soil during the incubation. Completely sealed jars were incubated in the dark at 22 ⁰C. Concentrations of CO2 and N2O were measured on days 1, 3, 7, 14, and 21 of the incubation using Infrared Photoacoustic Spectroscopy (INNOVA Air Tech Instruments, Denmark). After each gas measurement the jars were flushed with fresh air. After the incubation, the microcosms were air-dried again (for 5 days) and scanned at the same µCT settings as before-incubation. The µCT images obtained before and after incubation were used to calculate the volume loss during the decomposition (as described in section 2.2.4). Please note that even though there was a 5-day time lapse between the last gas measurement and the scanning, root decomposition during that period was negligible. First, the peak of active decomposition greatly subsided by the end of the 15 21-day incubation period; second, the microcosms were placed for drying into a ventilated hood and due to their small size dried very quickly (in < 3 hours). 2.2.4 X-ray µCT scanning and image analysis All microcosms were scanned at a resolution of 4.03 – 5.32 µm at sector 13-BM-D, GeoSoilEnvironCARS, Advanced Photon Source, Argonne National Laboratory, IL. During scanning two-dimensional projections were taken with 2 second exposure time and 0.25⁰ rotation angle (Quigley et al., 2018). Original projection images were reconstructed as 1200 slice images with 1,920 by 1,920 pixels. Image analyses were conducted with ImageJ software (Schneider et al., 2012). Before the main analysis all of the scanned images were preprocessed with Gaussian blur 3D (3x3x3 window) to reduce random noise. The sponge effect was assessed with the dual-energy approach (Kutlu et al., 2018), where microcosms from KI experiment were scanned at two energies, 33.269 keV and 33.069 keV, which are above and below the K-shell edge for iodine (33.169 keV). The mass attenuation coefficients of soil particles, air and water do not change considerably when switching from one energy level to another, however they do change for iodine. Therefore, subtracting the below K-edge images from those obtained at the energy above the K-edge provides a map of iodine distribution within the microcosm (Kutlu et al., 2018; Deboodt et al., 2019), which informs of the distribution patterns of water in the soil of the studied pore size and WFPS treatments. Difference images (33.269 keV - 33.069 keV) were converted into binary images reflecting presence and absence of iodine. Threshold value for iodine was determined according to the volume ratio of KI and total soil sample. The iodine content in each medium (roots and soils) was calculated as the number of medium's voxels occupied by the iodine divided by the total number of the medium's voxels, and was expressed as percent. 16 The thresholds for root and soil segmentation were computed using the minimum error thresholding approach (Kittler and Illingworth, 1986). The peaks corresponding to pore space and soil mineral material were clearly visible on the histograms of images (Fig. 2.2). Two Gaussian distributions, for pore and soil mineral, were fitted to histograms of grayscale images (Nakagawa and Rosenfeld, 1979; Kravchenko et al., 2019). The grayscale value corresponding to the pore peak, i.e., pore mean, plus two standard deviations was used as the lower boundary for root identification. The grayscale value corresponding to the mineral peak, i.e., mineral mean, minus two standard deviations was used as the upper boundary for root identification. After the initial root thresholding using the lower and upper boundaries, surfaces of identified roots were manually cleaned to increase the accuracy of root separation, followed by a series of filling holes, erosion, and dilation (1 iteration) operations using 3D erode and dilate tools of BoneJ. That helped with removing the partial volume effects and cleaning the surface of the root residue. Final removal of the remaining artifacts was achieved using particle identification tool of BoneJ, ‘Particle Analyzer’, with the options of minimum value as 6; maximum value as infinite; surface resampling and volume resampling as 2; and gradient split as 0. 17 Figure 2.2: Example of a histogram of scanned images (33.069 KeV). The two peaks represent are soil pore and soil mineral portions of the image. The same method was used for the images scanned under 28 KeV. From the dual-energy scanned images, I assessed the iodine contents as a function of the distance from the roots. For that I used 3D dilation tools from BoneJ plugin of ImageJ (Doube et al., 2010) to create 7 layers around each root. The layers followed the shape of the root and covered distances 0-48, 48-96, 96-144, 144-192, 192-240, 240-480, and 480-720 µm from the surface of the root (Fig. 2.3). Only the soil mineral voxels from the layers were used in iodine calculations, while all pore voxels were excluded. The iodine contents in each layer was calculated as the percent of the voxels occupied by iodine divided by the total number of soil mineral voxels within the layer. 18 Figure 2.3: Procedure for assessing distribution of iodine as a function of the distance from roots. (a) Original image. (b) Segmented roots. (c) Root dilation results. (d) Final layer used for measuring KI contents. Note that only the soil mineral voxels from the layers identified on (c) were used, while all pore voxels were excluded. (e) Difference image from dual energy scanning (33.269 keV - 33.069 keV) (f) Thresholded KI image reflecting the distribution of the liquid. (g) Examples of the layers overlaying the original image. 19 The microcosms from the incubation experiment were scanned at 28 keV energy both before- and after-incubation. Root binary images of before- and after-incubation were obtained, and the decomposition of the root was expressed as the root volume loss (%). ! 𝑉𝑜𝑙𝑢𝑚𝑒 𝑙𝑜𝑠𝑠 = (1 − !" ) × 100 # where, Va and Vb are the numbers of root voxels in the image sequences scanned after- and before- incubation, respectively. 2.2.5 Statistical analysis The statistical models used in the data analyses varied for different response variables depending on the specific experimental design settings. The root volume loss data originated from a completely randomized design and were analyzed using the statistical model with fixed factor i.e., pore size and WFPS, and their interaction. The iodine content in the two media, i.e., soil vs. root, within each microcosm was analyzed using the statistical model with the fixed effects of pore size, WFPS, medium type, their interactions, and a random effect of the microcosm nested within pore size and WFPS, which was used as an error term to test the effect of the pore size and WFPS. The data on iodine content as a function of the distance from the root were analyzed using the statistical model with the fixed effects of pore size, WFPS, layer, and their interactions. Microcosm was included as a random effect and used as an error term to test the effect of the pore size and WFPS. The CO2 and N2O fluxes were analyzed using a repeated measures approach as described in Milliken and Johnson (2009). For that, the statistical model consisted of fixed effects of pore size, WFPS, incubation time, and their interactions. The model also included a random effect of the microcosms nested within pore size and WFPS, which was used as an error term to test their effects and as a subject for repeated measurements. Model selection was conducted using Akaike 20 Information Criterion and Bayesian Information Criterion. All analyses were conducted in PROC MIXED of SAS 9.4 (SAS Inc, 2017). Summary of the F-tests for the studied statistical models are shown in Supplementary Tables 2.2 – 2.8. In all analyses the normality assumption was checked using normal probability plots of the residuals. The equal variance assumption was evaluated by examining the plots of the predicted versus residual values and the side-by-side box plots of the residuals (Fernandez, 1992; Kuehl, 2000; Ott and Longnecker, 2015). When the assumptions were found to be violated, the data were subjected to natural log-transformation. Reported are log-transformed values, but back- transformed means and 95% confidence intervals are provided in Table 2.9. Slicing, a.k.a. simple effect test of the interactions, was performed for all pre-planned interaction comparisons. The differences between the treatment means were reported as statistically significant based on the slicing results. The results are reported as statistically significant at p< 0.05 and as trends and tendencies at p< 0.10 levels. The figures were produced using Python version 3.6 (Python Software Foundation, https://www.python.org/). Error bars in all figures indicate standard errors. 2.3 Results 2.3.1 Sponge effect in decomposing roots assessed through iodine distribution patterns Roots held significantly higher amounts of added iodine than surrounding soil in all WFPS and pore size groups (p< 0.05, Fig. 2.4 and Table 2.2). At 50% WFPS, the root volumes with iodine in the large-pore microcosms were 9.4 % greater than that in the roots in the small-pore microcosms (p< 0.05, Table 2.3). However, the root volume with iodine at 75% WFPS was not significantly different between the pore size groups. 21 Table 2.2: Summary of F-tests for the statistical model fitted to the iodine content data in soil and root media. Shown are F-values and p -values for the main and interaction effects. Please see the model description in the Materials and Methods. Iodine content Effect F Value p-value Pore size (large vs. small) 0.27 0.6232 WFPS (50% vs. 75%) 56.28 0.0004 Medium type (soil vs. root) 891.25 <.0001 Pore size*WFPS 2.01 0.2101 Medium type*Pore size 0.68 0.4438 Medium type*WFPS 2.83 0.1478 Medium type*Pore size*WFPS 0.32 0.5911 Table 2.3: Effect of pore size on iodine content at given medium types and moisture levels. Statistical differences marked on Fig. 2.4 are based on the p-value presented here. Effect of pore size on Iodine content Medium type Moisture F Value p-value Root1 50% WFPS 8.49 0.0435 Root 75% WFPS 0.32 0.6037 Soil 50% WFPS 0.04 0.8437 Soil 75% WFPS 0.20 0.6789 1 Each line presents the F-test for the pore size effect at the specified medium type and WFPS. 22 Figure 2.4: Water absorption by dry decomposing roots assessed though iodine gradients. Shown are an example of a 3D visualization of a root, soil, and iodine (a), percent of iodine occupied voxels within the root voxels at 50% WFPS (b) and 75% WFPS (c), and percent of iodine occupied voxels within the soil matrix voxels at 50% WFPS (d) and 75% WFPS (e). Symbol ** marks statistically significant differences in iodine levels between large- and small-pore microcosms (p< 0.05). Iodine content in the soil immediately adjacent to the roots (~48 µm) was noticeably higher than in the bulk soil matrix (> 720 µm from roots; p< 0.05). In all treatments, iodine content decreased markedly at 0-96 µm distance from the roots and reached its background level (i.e., 23 iodine content in the bulk soil matrix) ~150 µm from the roots (Fig. 2.5). While there was no significant difference between pore sizes at 75% WFPS, large-pore microcosms had greater iodine content in the soil at 0-48 µm distance from the roots, compared to small-pore soils at 50% WFPS (p< 0.05, Fig. 2.5 and Table 2.5). That is, the gradient created at 0-96 µm distance from the roots in large-pore soil was higher than the one created in small-pore soil. Figure 2.5: Percent of iodine occupied voxels within the soil matrix voxels as a function of the distance from the roots. Gray dashed line is the average iodine content in the bulk soil matrix within the same WFPS. Symbol ** marks statistically significant difference between iodine levels in large- and small-pore microcosms at 0-48 μm layer (p< 0.05). 2.3.2 Root decomposition The loss of root volume was higher in the microcosms of the large- than the small-pore size group at 50% WFPS (p< 0.05, Fig. 2.6 and Table 2.7). In the small-pore microcosms, 75% WFPS tended to lead to a greater root volume loss compared to that in 50% WFPS (p< 0.10, Table 24 2.7). Figure 2.6: Root decomposition during the 21-day incubation. Shown are an example of a 3D visualization of a root before (left) and after (right) incubation (a), and the root volume losses (%) in the large- and small- pore microcosms at 50% WFPS (b) and 75% WFPS (c). Shown are the treatment means, the error bars represent standard errors (n=4). Volumes were calculated from the number of voxels in μCT image stacks. Symbol ** indicates statistically significant differences between pore size treatments at the same WFPS (p< 0.05), and different letters indicate statistically significant differences between WFPSs at the same pore size group (p< 0.10). 2.3.3 CO2 and N2O emissions during the incubation The large-pore microcosms had higher CO2 emission rates compared to the small-pore microcosms on days 3, 14, and 21 of the incubation at 50% WFPS (Fig. 2.7a). However, there was no significant difference in CO2 emissions between the two pore sizes at 75% WFPS (Fig. 2.7b). 25 WFPS had no effect on the cumulative amounts of emitted CO2. At 50% WFPS, N2O emission in the large-pore microcosms was significantly higher than that in the small-pore microcosms throughout the incubation period (Fig. 2.7c). In contrast, at 75% WFPS, N2O emission tended to be higher in the small- than in the large-pore microcosms (Fig. 2.7d). The difference was especially pronounced during the first 3 days of the incubation and disappeared afterwards. Cumulative N2O emission exhibited a similar pattern; at 50% WFPS emission from the large-pore microcosms exceeded that from the small-pore microcosms, while at 75% WFPS small-pore emissions exceeded the large-pore ones (Fig. 2.8). The effect of WFPS on N2O emissions from the large- and small-pore microcosms depended on the incubation time (Fig. 2.9). In the small-pore microcosms greater N2O emissions at 75% than at 50% WFPS were observed from the start of the incubation and continued for the entire incubation period (Fig. 2.9b). In the large-pore microcosms, greater N2O emissions at 75% than at 50% WFPS were also observed for a substantial period of time during the incubation, but only starting from day 5-6. However, during the first ~5 days, greater emissions took place at 50% than at 75% WFPS (Fig. 2.9a). In the small-pore microcosms cumulative N2O emission was greater at 75% than at 50% WFPS (p< 0.05), while WFPS effect was not statistically significant in the large-pore microcosms (Fig. 2.8). F-value and p-value for the treatment effects are provided in Table 2.8. 26 Figure 2.7: CO2 and N2O fluxes during the 21-day incubation in large- and small-pore size microcosms at the two studied WFPS, grouped by moisture content. (a) CO2 emission at 50% WFPS, (b) CO2 emission at 75% WFPS, (c) N2O emission at 50% WFPS, and (d) N2O emission at 75% WFPS. Shown are the treatment means, the error bars represent standard errors (n=5). Symbols * and ** mark significant differences between pore sizes within the same day (p< 0.10 and p< 0.05, respectively). 27 Figure 2.8: Cumulative CO2 and N2O emission during the incubation. Symbols * and ** indicate the differences between pore sizes with 0.10 and 0.05 significance level, respectively. The different lowercase letters represent the differences between moisture contents in the same pore size group. 28 Figure 2.9: N2O fluxes during 21-day incubation in large- and small-pore size microcosms at the two studied WFPSs, grouped by pore-size. Shown are the treatment means, the error bars represent standard errors (n=5). Symbol ** marks the differences between WFPSs within the same day (p< 0.05). Additional information on N2O emissions from bare soil microcosms (i.e., without root residue) under 50% WFPS was presented in Fig. 2.10. Initial N2O emission from bare soils were much lower (< 0.2 mg N2O-N kg-1·soil·day-1) compared to the microcosms with root residues (> 10 mg N2O-N kg-1·soil·day-1). There was a significant difference in N2O emission between large- pore bare soils and small-pore bare soils only at day 3 of the incubation (p< 0.05). 29 Figure 2.10: N2O emission from control soil during the incubation. Symbol ** marks the difference between the pore sizes within the same day (p< 0.05). 2.4 Discussion 2.4.1 Water absorption by decomposing plant roots – the sponge effect The KI solution was preferentially absorbed by the decomposing plant roots, with a minor amount remaining in the soil itself (Fig. 2.4). This indicates the presence of the sponge effect in root residue, which is consistent with previously reported findings of the sponge effect in leaf and stem residues of different plant species (Iqbal et al., 2013; Kravchenko et al., 2017). While the transformation of iodide into organoiodine upon contact with organic material likely also took place (Yamaguchi et al., 2010), the redistribution of the liquid added into the air-dry microcosms by the capillary forces can be regarded as the main driving force for the resultant iodine attenuation patterns. 30 Greater sponge effect in the large-pore soil at 50% WFPS (Fig. 2.4b and d) resulted from the lower water retention capacity of large pores, thus greater matric potential gradient between decomposing plant residue and surrounding soil (Kutlu et al., 2018). However, at 75% WFPS, roots in both large- and small-pore soils had similarly high iodine contents, close to their full saturation (Fig. 2.4c). Kutlu et al. (2018) demonstrated that while the water content of the soybean leaves was greater in the large-pore microcosms rather than in the small-, when soil moisture content ranged from 18–36 % WFPS, the difference disappeared as soil moisture content exceeded 73% WFPS. Consistent with my findings, as water content increased (75% WFPS) the differences in iodine contents between the pore size treatments disappeared. Water distribution gradient from the decomposing roots into soil matrix (Fig. 2.5) reflected the liquid levels within the roots themselves (Fig. 2.4) and were the strongest in 75% WFPS samples, followed by 50% WFPS large pore samples and then the 50% small pore samples. This suggests that the overall gradient in water and iodine levels between the roots and the soil matrix was the main driving force behind the observed trends. While micro-scale patterns in water distribution in the rhizosphere have been assessed before (Carminati et al., 2010), to my knowledge, this is the first time that the water gradients next to decomposing roots were evaluated on a µm scale. Further studies of the micro-scale patterns in water re-distribution within detritusphere are needed, since such patterns can influence microscale redox conditions, microbial activity (e.g., aerobic, anaerobic) hotspots, and thus heterogeneous C and N turnover rates. 31 2.4.2 Root decomposition and CO2 emission Greater root decomposition in large-pore soil at 50% WFPS as compared to the small-pore soil (Fig. 2.6b) is consistent with previously reported aboveground residue decomposition findings. Greater corn leaf volume loss was observed in the large- (> 30 µm) than in the small- (< 10 µm) pore soil at 35-50% WFPS (Negassa et al., 2015; Kravchenko et al., 2017), and greater wheat residue decomposition was associated with 15-60 µm than < 4 µm pores (Strong et al., 2004). Coppens et al. (2007) showed that maximized water content of the plant residue can increase the decomposition rate by PASTIS (Prediction of Agricultural Solute Transport In Soil) model scenario analysis. Cumulative CO2 emissions were not affected by soil WFPS. Consistent with this observation, negligible response of CO2 emission to the soil moisture was reported by Ruser et al. (2006) at 40 - 90% WFPS and by Moyano et al. (2012) at > 40% WFPS. Since the soil WFPS in this study was within an optimal range for microbes, WFPS was probably not a limiting factor for microbial respiration. The influence of pore size on CO2 emissions depended on the soil moisture content. At 50% WFPS, I observed greater CO2 emission from the large- rather than the small- pore microcosms (Fig. 2.7a), consistent with the higher root volume losses. While, no differences between the pore-size treatments were observed at 75% WFPS. The observed higher CO2 emissions from large rather than from small-pore treatments at 50% WFPS contradict other decomposition experiments with soil of contrasting particle sizes, where greater CO2 emissions typically occurred in finer soil materials (Rastogi et al., 2002; Oertel et al., 2016). Greater CO2 emission in the small pore dominated soil was also reported in the studies conducted previously in my research group (Negassa et al., 2015; Toosi et al., 2017). This discrepancy is likely brought by the differences in timings between soil material preparations and 32 incubation experiments. In the process of grinding the large aggregate fraction to procure the small-pore material, the organic carbon originally protected within large aggregates typically becomes available for decomposition (Balesdent et al., 2000). Available C in crushed soil causes a burst of CO2 when it is wetted (Van Veen and Kuikman, 1990; Jarvis et al., 2007). Other studies (e.g., Negassa et al., 2015; Toosi et al., 2017) monitored CO2 emission immediately after wetting, capturing the burst of CO2 in freshly ground soil. Meanwhile, the burst of CO2 was not captured in this study because soil was wetted several days before the incubation, i.e., at planting, and was kept in moist and wet conditions during the 4 days of plant growth. 2.4.3 N2O emission It should be noted that the two studied soil materials did not differ substantially in terms of either total C and N, and/or inorganic N contents (Table 2.1). N2O emissions from the control soil were very low in both materials, and, as expected, tended to be somewhat higher in the small pore than in the large pore treatment (Fig. 2.10), due to greater anaerobic conditions within the former. Presence of decomposing roots increased N2O emission ten to hundred-fold compared to the controls and markedly changed the pattern of differences in N2O emissions between large and small pore materials (Fig. 2.7). These results add to the growing evidence of the importance of interactions among pore architecture, soil moisture, and plant residues for soil biogeochemical processes, including microbial oxygen consumption and denitrification dynamics (Ebrahimi and Or, 2018; Schlüter et al., 2018). The presence of root residue changed the temporal dynamic of soil moisture influence on N2O emissions. After the first 5 days of incubation, the N2O emissions were higher at 75% than at 50% WFPS in both large and small-pore soil microcosms (Fig. 2.9). This result is consistent with a large body of previous work reporting that N2O emission increases along the soil moisture 33 content gradient, reaching maximum at 75% - 100% WFPS (e.g., Khalil and Baggs, 2005; Ciarlo et al., 2007). Denitrification is the main source of N2O production in the anoxic soil matrix at such high moisture levels (Groffman and Tiedje, 1989; McTaggart et al., 2002; Ciarlo et al., 2007; van der Weerden et al., 2012). However, during the first 5 days of incubation, an opposite trend was observed in the large- pore microcosms: N2O emission was significantly higher at 50% than at 75% WFPS. This result can be attributed to the influence of the decomposing roots. At the start of the incubations (first ~5 days), at 50% WFPS, higher amounts of water were absorbed by the root residues in the large- pore than in the small-pore microcosms (Fig. 2.5a). The high moisture levels within the residues enhanced root decomposition (Fig. 2.6b), likely providing greater amounts of available C (Gaillard et al., 1999; Gaillard et al., 2003), and turned the root into a local hotspot of anoxic conditions (Li et al., 2016). The large amounts of N2O produced within the decomposing roots during the first 5 days of incubation then quickly escaped via atmosphere connected pores dominating the large pore microcosms. Later into the incubation (> 5 days) the contribution of the roots to N2O production decreased, and the emitted N2O was probably dominated by the production from within the soil matrix itself. Subsequently, the N2O emissions became higher in the microcosms with higher (75% WFPS) bulk soil moisture level. In the small-pore microcosms at 50% WFPS, the contribution of roots to the initial N2O production and emission was probably lower than in the large-pore microcosms. That could be caused by slower root decomposition (Fig. 2.6b and 2.7a) limiting the sources of C/N required for microbes to produce N2O and weaker sponge effect in the root forming less extreme anoxic conditions within the root (Fig. 2.4b). Therefore, in the small-pore soil, WFPS was the main driving force of the N2O emissions during the entire incubation period. 34 My findings suggest that in soil with a dominance of > 30 µm pores, the contribution of decomposing roots to N2O emission can be substantial and, as a result, the bulk soil WFPS characteristics might not be a reliable N2O emission predictor (Li et al., 2016). These observations concur with results from several other studies. For example, Velthof et al. (2002) reported greater total N2O emission from Brussels sprouts, mustard, and broccoli residues in sandy compared to clay soil. Weak associations between bulk soil moisture content and N2O emissions in residue amended soil is another supporting example: during decomposition of Vicia villosia, no correlation between moisture level and N2O emission was observed at the beginning of incubation (Shelton et al., 2000). Also, N2O emission was not proportional to soil moisture content (40% - 60% WFPS) in the soil where Trifolium pratense L. and Vicia villosa were incorporated (Li et al., 2016). It should be noted that a formal quantification of the contribution of decomposing roots to the overall amounts of emitted N2O was not conducted in this study. Such quantification will be needed to fully assess the potential contribution of decomposing roots to hotspot N2O production and will be the subject of further investigation. Also, young legume roots used in this study tend to have low C:N ratios, likely resulting in maximal N2O productions and emissions (Velthof et al., 2002; Huang et al., 2004). While, quantitatively, my findings may not fully represent the effects from decomposing older roots in the field, they do provide insights on the factors contributing to hot-spot N2O production and emissions from in-situ grown roots. 2.5 Conclusion The study demonstrated that the sponge effect was present in young decomposing soybean roots. Up to 62.6 % greater amounts of the added liquid accumulated within the roots than within the soil. The added liquid formed a distribution gradient around the roots, decreasing with increasing distance from the roots until reaching background soil levels at a distance of ~150 µm. 35 To my best knowledge, this is the first time when the water gradients next to decomposing roots were evaluated on an µm scale using X-ray µCT image analysis. Further studies of the micro-scale patterns in water re-distribution within detritusphere are needed, since such patterns can influence microscale redox conditions, microbial activity (e.g., aerobic, anaerobic) hotspots, and thus heterogeneous C and N turnover rates. At medium soil moisture (50% WFPS) the large-pore dominated soil emitted greater amounts of N2O than the small pore soil, and, surprisingly, even more N2O than the large pore soil at high soil moisture (75% WFPS). This finding suggests that the decomposing root residues acted as hot spots of N2O production, probably due to enhanced sponge effect and associated local anoxic conditions. However, after approximately 5 days of incubation the N2O emission at 50% WFPS became lower than that at 75% WFPS, indicating that the contribution of the decomposing roots to N2O production declined. At high soil moisture (75% WFPS) and in the absence of roots, greater N2O emissions were observed from the soil dominated by small pores. Table 2.4: Summary of F-tests for the statistical models fitted to iodine content data in 7 soil layers. Shown are F-values and p -values for the main and interaction effects. Iodine gradient Effect F Value p-value Pore size 0.00 0.9840 WFPS 9.35 0.0301 Layer 23.34 <.0001 Pore size*WFPS 0.67 0.4519 Pore size*Layer 0.99 0.4544 WFPS*Layer 1.76 0.1513 WFPS*Pore size*Layer 3.96 0.0071 36 Table 2.5: Effect of pore size on iodine content at given WFPSs and soil layers. Statistical differences marked in Fig. 2.5 are based on the p-value presented here. Effect of pore size on Iodine content WFPS Layer F Value p-value WFPS Layer F Value p-value 0 - 48 μm 5.34 0.0483 0 - 48 μm 0.26 0.6227 48 - 96 μm 0.11 0.7457 48 - 96 μm 0.77 0.4046 96 - 144 μm 0.05 0.8263 96 - 144 μm 0.35 0.5686 144 - 192 144 - 192 μm 0.24 0.6337 0.03 0.8664 μm 50% 75% 192 - 240 WFPS1 192 - 240 μm 0.38 0.5538 WFPS1 0.48 0.5074 μm 240 - 480 240 - 480 μm 0.10 0.7557 2.18 0.1766 μm 480 - 720 480 - 720 μm 0.00 0.9754 3.75 0.0873 μm 1 Each line presents the F-test for the pore size effect at the specified WFPS and soil layer. Table 2.6: Summary of F-tests for the statistical models fitted to volume loss data. Shown are F-values and p - values for the main and interaction effects. Root volume loss Effect F Value p-value Pore size 6.34 0.0655 WFPS 1.91 0.2388 Pore size*WFPS 3.49 0.1350 37 Table 2.7: Effect of WFPS at given pore sizes (above) and effect of pore size at given WFPSs (below) on volume loss data. Shown are F-values and p-values for WFPS and pore size effect, when the level of another factor is fixed. Statistical differences marked in Fig. 2.6b and c are based on the p-values presented here. Effect Pore size F Value p-value Large-pore 0.12 0.7485 WFPS1 Small-pore 5.29 0.0829 Effect WFPS F Value p-value 50% WFPS 9.63 0.0361 Pore size2 75% WFPS 0.21 0.6700 1 Each line presents the F-test for WFPS effect at specified pore size. 2 Each line presents the F-test for pore size effect at specified WFPS. Table 2.8: Summary of F-tests for the statistical models fitted to CO2 and N2O flux data. Shown are F-values and p -values for the main and interaction effects. CO2 emission N2O emission Effect F Value p-value F Value p-value Pore size 1.68 0.2081 0.16 0.6947 WFPS 0.88 0.3588 12.83 0.0016 Time 554.34 <.0001 72.82 <.0001 Pore size*WFPS 1.58 0.2215 12.55 0.0016 Pore size*Time 3.44 0.0114 1.80 0.1348 WFPS*Time 11.93 <.0001 16.30 <.0001 Pore size*WFPS*Time 1.84 0.1273 1.24 0.2999 38 Table 2.9: N2O and CO2 emissions with original units. Means and 95% confidence interval was back- transformed from the results of statistical analysis. N2O CO2 (µg N·kg-1soil·day-1) (mg C·kg-1soil·day-1) Moisture Pore size Day Mean CIlower CIupper Mean CIlower CIupper 50 Large 1 93.26 60.06 144.80 44.95 36.76 54.97 50 Large 3 10.78 6.94 16.74 23.71 19.39 28.99 50 Large 7 1.19 0.76 1.84 5.87 4.80 7.18 50 Large 14 0.45 0.29 0.70 2.47 2.02 3.02 50 Large 21 0.69 0.44 1.07 2.09 1.71 2.56 50 Small 1 28.88 19.14 43.59 31.27 25.91 37.75 50 Small 3 3.94 2.61 5.95 13.21 10.95 15.95 50 Small 7 0.36 0.24 0.55 3.91 3.24 4.71 50 Small 14 0.16 0.10 0.24 1.54 1.28 1.86 50 Small 21 0.14 0.09 0.21 1.21 1.00 1.46 75 Large 1 11.22 6.67 18.89 21.15 16.67 26.83 75 Large 3 1.65 0.98 2.78 18.62 14.68 23.62 75 Large 7 2.57 1.53 4.33 5.46 4.30 6.93 75 Large 14 2.26 1.34 3.80 3.41 2.69 4.33 75 Large 21 3.66 2.17 6.16 3.27 2.58 4.15 75 Small 1 122.15 78.67 189.66 31.58 25.83 38.62 75 Small 3 5.93 3.82 9.21 21.16 17.30 25.87 75 Small 7 4.64 2.99 7.21 5.60 4.58 6.85 75 Small 14 3.32 2.14 5.15 3.03 2.48 3.71 75 Small 21 4.24 2.73 6.58 2.04 1.67 2.50 39 ACKNOWLEDGEMENTS This work was funded in part by the National Science Foundation’s Geobiology and Low Temperature Geochemistry Program (Award number 1630399). This material is based upon work supported in part by the Great Lakes Bioenergy Research Center, U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research under Award Number DE- SC0018409. This research used resources of the Advanced Photon Source, a U.S. Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Argonne National Laboratory under Contract No. DE-AC02-06CH11357. I appreciate the support from Michigan State University Environmental Science and Policy Program (ESPP). Also, I thank Michelle Quigley and Maxwell Oerther for help with laboratory analysis. 40 REFERENCES 41 REFERENCES Angers, D.A., Caron, J., 1998. Plant-induced changes in soil structure: processes and feedbacks. Biogeochemistry 42, 55-72. Bai, Z., Liang, C., Bodé, S., Huygens, D., Boeckx, P., 2016. Phospholipid 13C stable isotopic probing during decomposition of wheat residues. Applied Soil Ecology 98, 65-74. Baggs, E., Rees, R., Smith, K., Vinten, A., 2000. Nitrous oxide emission from soils after incorporating crop residues. Soil use and management 16, 82-87. Bateman, E., Baggs, E., 2005. Contributions of nitrification and denitrification to N2O emissions from soils at different water-filled pore space. Biology and fertility of soils 41, 379-388. Begum, N., Guppy, C., Herridge, D., Schwenke, G., 2014. Influence of source and quality of plant residues on emissions of N2O and CO 2 from a fertile, acidic Black Vertisol. Biology and fertility of soils 50, 499-506. Bird, J.A., Kleber, M., Torn, M.S., 2008. 13C and 15N stabilization dynamics in soil organic matter fractions during needle and fine root decomposition. Organic Geochemistry 39, 465-477. Bolinder, M., Angers, D., Bélanger, G., Michaud, R., Laverdière, M., 2002. Root biomass and shoot to root ratios of perennial forage crops in eastern Canada. Canadian Journal of Plant Science 82, 731-737. Carminati, A., Moradi, A.B., Vetterlein, D., Vontobel, P., Lehmann, E., Weller, U., Vogel, H.-J., Oswald, S.E., 2010. Dynamics of soil water content in the rhizosphere. Plant and Soil 332, 163- 176. Carminati, A., Vetterlein, D., 2012. Plasticity of rhizosphere hydraulic properties as a key for efficient utilization of scarce resources. Annals of botany 112, 277-290. Chen, H., Li, X., Hu, F., Shi, W., 2013. Soil nitrous oxide emissions following crop residue addition: a meta‐analysis. Global change biology 19, 2956-2964. Chirinda, N., Olesen, J.E., Porter, J.R., 2012. Root carbon input in organic and inorganic fertilizer- based systems. Plant and Soil 359, 321-333 Ciarlo, E., Conti, M., Bartoloni, N., Rubio, G., 2007. The effect of moisture on nitrous oxide emissions from soil and the N2O/(N2O+ N 2) ratio under laboratory conditions. Biology and fertility of soils 43, 675-681. Coppens, F., Garnier, P., Findeling, A., Merckx, R., Recous, S., 2007. Decomposition of mulched versus incorporated crop residues: modelling with PASTIS clarifies interactions between residue quality and location. Soil Biology and Biochemistry 39, 2339-2350. 42 Deboodt, T., Wildenschild, D., Ideker, J.H., Isgor, O.B., 2019. Use of iodide for improving phase quantification using x-ray tomography. Cement and Concrete Research 116, 102-112. Doane, T.A., Horwáth, W.R., 2003. Spectrophotometric determination of nitrate with a single reagent. Analytical letters 36, 2713-2722. Doube, M., Kłosowski, M.M., Arganda-Carreras, I., Cordelières, F.P., Dougherty, R.P., Jackson, J.S., Schmid, B., Hutchinson, J.R., Shefelbine, S.J., 2010. BoneJ: free and extensible bone image analysis in ImageJ. Bone 47, 1076-1079. Ebrahimi, A., Or, D., 2018. Dynamics of soil biogeochemical gas emissions shaped by remolded aggregate sizes and carbon configurations under hydration cycles. Global change biology 24, e378-e392. Fernandez, G.C., 1992. Residual analysis and data transformations: important tools in statistical analysis. HortScience 27, 297-300. Gaillard, V., Chenu, C., Recous, S., 2003. Carbon mineralisation in soil adjacent to plant residues of contrasting biochemical quality. Soil Biology and Biochemistry 35, 93-99. Gaillard, V., Chenu, C., Recous, S., Richard, G., 1999. Carbon, nitrogen and microbial gradients induced by plant residues decomposing in soil. European Journal of Soil Science 50, 567-578. Gale, W., Cambardella, C., 2000. Carbon dynamics of surface residue–and root-derived organic matter under simulated no-till. Soil Science Society of America Journal 64, 190-195. Groffman, P.M., Tiedje, J.M., 1989. Denitrification in north temperate forest soils: spatial and temporal patterns at the landscape and seasonal scales. Soil Biology and Biochemistry 21, 613- 620. Hansson, K., Kleja, D.B., Kalbitz, K., Larsson, H., 2010. Amounts of carbon mineralised and leached as DOC during decomposition of Norway spruce needles and fine roots. Soil Biology and Biochemistry 42, 178-185. Hobbie, S.E., Oleksyn, J., Eissenstat, D.M., Reich, P.B., 2010. Fine root decomposition rates do not mirror those of leaf litter among temperate tree species. Oecologia 162, 505-513. Huang, Y., Zou, J., Zheng, X., Wang, Y., Xu, X., 2004. Nitrous oxide emissions as influenced by amendment of plant residues with different C: N ratios. Soil Biology and Biochemistry 36, 973- 981. Iqbal, A., Beaugrand, J., Garnier, P., Recous, S., 2013. Tissue density determines the water storage characteristics of crop residues. Plant and Soil 367, 285-299. Jarvis, P., Rey, A., Petsikos, C., Wingate, L., Rayment, M., Pereira, J., Banza, J., David, J., Miglietta, F., Borghetti, M., 2007. Drying and wetting of Mediterranean soils stimulates decomposition and carbon dioxide emission: the “Birch effect”. Tree physiology 27, 929-940. 43 Jin, V.L., Baker, J.M., Johnson, J.M.-F., Karlen, D.L., Lehman, R.M., Osborne, S.L., Sauer, T.J., Stott, D.E., Varvel, G.E., Venterea, R.T., 2014. Soil greenhouse gas emissions in response to corn stover removal and tillage management across the US Corn Belt. BioEnergy Research 7, 517-527. Jung, J.Y., Lal, R., Ussiri, D.A., 2011. Changes in CO 2, 13 C abundance, inorganic nitrogen, β- glucosidase, and oxidative enzyme activities of soil during the decomposition of switchgrass root carbon as affected by inorganic nitrogen additions. Biology and Fertility of Soils 47, 801-813. Keiluweit, M., Wanzek, T., Kleber, M., Nico, P., Fendorf, S., 2017. Anaerobic microsites have an unaccounted role in soil carbon stabilization. Nature communications 8, 1771. Khalil, M., Baggs, E., 2005. CH4 oxidation and N2O emissions at varied soil water-filled pore spaces and headspace CH4 concentrations. Soil Biology and Biochemistry 37, 1785-1794. Kittler, J., Illingworth, J., 1986. Minimum error thresholding. Pattern Recognition 19, 41–47. Köbke, S., Senbayram, M., Pfeiffer, B., Nacke, H., Dittert, K., 2018. Post-harvest N2O and CO2 emissions related to plant residue incorporation of oilseed rape and barley straw depend on soil NO3-content. Soil and Tillage Research 179, 105-113. Kravchenko, A., Guber, A., Razavi, B., Koestel, J., Blagodatskaya, E., Kuzyakov, Y., 2019. Spatial patterns of extracellular enzymes: Combining X-ray computed micro-tomography and 2D zymography. Soil Biology and Biochemistry. Kravchenko, A., Fry, J., Guber, A., 2018. Water absorption capacity of soil-incorporated plant leaves can affect N2O emissions and soil inorganic N concentrations. Soil Biology and Biochemistry 121, 113-119. Kravchenko, A., Toosi, E., Guber, A., Ostrom, N., Yu, J., Azeem, K., Rivers, M., Robertson, G., 2017. Hotspots of soil N2O emission enhanced through water absorption by plant residue. Nature Geoscience 10, 496. Kravchenko, A.N., Negassa, W.C., Guber, A.K., Rivers, M.L., 2015. Protection of soil carbon within macro-aggregates depends on intra-aggregate pore characteristics. Scientific reports 5, 16261. Kumar, K., Goh, K., 1999. Crop residues and management practices: effects on soil quality, soil nitrogen dynamics, crop yield, and nitrogen recovery, Advances in agronomy. Elsevier, pp. 197- 319. Kutlu, T., Guber, A.K., Rivers, M.L., Kravchenko, A.N., 2018. Moisture absorption by plant residue in soil. Geoderma 316, 47-55. Kuzyakov, Y., Hill, P.W., Jones, D.L., 2007. Root exudate components change litter decomposition in a simulated rhizosphere depending on temperature. Plant and Soil 290, 293-305. Lal, R., 2005. World crop residues production and implications of its use as a biofuel. Environment International 31, 575-584. 44 Lehtinen, T., Schlatter, N., Baumgarten, A., Bechini, L., Krüger, J., Grignani, C., Zavattaro, L., Costamagna, C., Spiegel, H., 2014. Effect of crop residue incorporation on soil organic carbon and greenhouse gas emissions in European agricultural soils. Soil use and management 30, 524-538. Li, X., Sørensen, P., Olesen, J.E., Petersen, S.O., 2016. Evidence for denitrification as main source of N2O emission from residue-amended soil. Soil Biology and Biochemistry 92, 153-160. Littell, R.C., Milliken, G.A., Stroup, W.W., Wolfinger, R.D., Oliver, S., 2006. SAS for Mixed Models, Second Edition. SAS Publishing. McTaggart, I.P., Akiyama, H., Tsuruta, H., Ball, B.C., 2002. Influence of soil physical properties, fertiliser type and moisture tension on N2O and NO emissions from nearly saturated Japanese upland soils. Nutrient Cycling in Agroecosystems 63, 207-217. Meier, I.C., Finzi, A.C., Phillips, R.P., 2017. Root exudates increase N availability by stimulating microbial turnover of fast-cycling N pools. Soil Biology and Biochemistry 106, 119-128. Milliken, G.A., Johnson, D.E., 2009. Analysis of messy data volume 1: designed experiments vol. 1. Chapman & Hall/CRC, London. Moretto, A.S., Distel, R.A., Didone, N.G., 2001. Decomposition and nutrient dynamic of leaf litter and roots from palatable and unpalatable grasses in a semi-arid grassland. Applied Soil Ecology 18, 31-37. Moyano, F.E., Vasilyeva, N.A., Bouckaert, L., Cook, F., Craine, J.M., Don, A., Epron, D., Formanek, P., Franzluebbers, A., Ilstedt, U., 2012. The moisture response of soil heterotrophic respiration: interaction with soil properties. Biogeosciences 9, 1173-1182. Nakagawa, Y., Rosenfeld, A., 1979. Some experiments on variable thresholding. Pattern recognition 11, 191-204. Negassa, W.C., Guber, A.K., Kravchenko, A.N., Marsh, T.L., Hildebrandt, B., Rivers, M.L., 2015. Properties of soil pore space regulate pathways of plant residue decomposition and community structure of associated bacteria. PLoS One 10, e0123999. Oertel, C., Matschullat, J., Zurba, K., Zimmermann, F., Erasmi, S., 2016. Greenhouse gas emissions from soils—A review. Chemie der Erde-Geochemistry 76, 327-352. Ostertag, R., Hobbie, S.E., 1999. Early stages of root and leaf decomposition in Hawaiian forests: effects of nutrient availability. Oecologia 121, 564-573 Ott, R.L., Longnecker, M.T., 2015. An introduction to statistical methods and data analysis. Nelson Education. Parkin, T.B., 1987. Soil microsites as a source of denitrification variability 1. Soil Science Society of America Journal 51, 1194-1199. 45 Patrick, Z., Toussoun, T., Koch, L., 1964. Effect of crop-residue decomposition products on plant roots. Annual Review of Phytopathology 2, 267-292. Paustian, K., Six, J., Elliott, E., Hunt, H., 2000. Management options for reducing CO2 emissions from agricultural soils. Biogeochemistry 48, 147-163. Quigley, M.Y., Rivers, M.L., Kravchenko, A.N., 2018. Patterns and sources of spatial heterogeneity in soil matrix from contrasting long term management practices. Frontiers in Environmental Science 6. Rastogi, M., Singh, S., Pathak, H., 2002. Emission of carbon dioxide from soil. Current Science 82, 510-517. Röver, M., Heinemeyer, O., Munch, J.C., Kaiser, E.-A., 1999. Spatial heterogeneity within the plough layer: high variability of N2O emission rates. Soil Biology and Biochemistry 31, 167-173. Roy, J., Mooney, H.A., Saugier, B., 2001. Terrestrial global productivity. pp.83-94. Elsevier. Ruser, R., Flessa, H., Russow, R., Schmidt, G., Buegger, F., Munch, J., 2006. Emission of N2O, N2 and CO2 from soil fertilized with nitrate: effect of compaction, soil moisture and rewetting. Soil Biology and Biochemistry 38, 263-274. SAS Inc., 2017. Base SAS 9.4 procedures guide: Statistical procedures. SAS Institute. Schlüter, S., Zawallich, J., Vogel, H. J., and Dörsch, P.: Physical constraints for respiration in microbial hotspots in soil and their importance for denitrification, Biogeosciences, 16, 3665-3678. Schmidt, M.W., Torn, M.S., Abiven, S., Dittmar, T., Guggenberger, G., Janssens, I.A., Kleber, M., Kögel-Knabner, I., Lehmann, J., Manning, D.A., 2011. Persistence of soil organic matter as an ecosystem property. Nature 478, 49. Schneider, C.A., Rasband, W.S., Eliceiri, K.W., 2012. NIH Image to ImageJ: 25 years of image analysis. Nature methods 9, 671. Sembiring, H., Raun, W., Johnson, G., Boman, R., 1995. Effect of wheat straw inversion on soil water conservation. Soil science 159, 81-89. Shahbaz, M., Kumar, A., Kuzyakov, Y., Börjesson, G., Blagodatskaya, E., 2018. Interactive priming effect of labile carbon and crop residues on SOM depends on residue decomposition stage: Three-source partitioning to evaluate mechanisms. Soil Biology and Biochemistry 126, 179-190. Shelton, D.R., Sadeghi, A.M., McCarty, G.W., 2000. Effect of soil water content on denitrification during cover crop decomposition. Soil science 165(4), pp 365 -371. Sinsabaugh, R., Reynolds, H., Long, T., 2000. Rapid assay for amidohydrolase (urease) activity in environmental samples. Soil Biology and Biochemistry 32, 2095-2097. 46 Strong, D., Wever, H.D., Merckx, R., Recous, S., 2004. Spatial location of carbon decomposition in the soil pore system. European Journal of Soil Science 55, 739-750. Toosi, E., Kravchenko, A., Guber, A., Rivers, M., 2017. Pore characteristics regulate priming and fate of carbon from plant residue. Soil Biology and Biochemistry 113, 219-230. Tufekcioglu, A., Raich, J.W., Isenhart, T.M., Schultz, R.C., 1998. Fine root dynamics, coarse root biomass, root distribution, and soil respiration in a multispecies riparian buffer in Central Iowa, USA. Agroforestry Systems 44, 163-174. van der Weerden, T.J., Kelliher, F.M., de Klein, C.A., 2012. Influence of pore size distribution and soil water content on nitrous oxide emissions. Soil Research 50, 125-135. Vanlauwe, B., Nwoke, O., Sanginga, N., Merckx, R., 1996. Impact of residue quality on the C and N mineralization of leaf and root residues of three agroforestry species. Plant and Soil 183, 221- 231. Velthof, G.L., Kuikman, P.J., Oenema, O., 2002. Nitrous oxide emission from soils amended with crop residues. Nutrient Cycling in Agroecosystems 62, 249-261. Vessey, J.K., Raper JR, C.D., Henry, L.T., 1990. Cyclic variations in nitrogen uptake rate in soybean plants: uptake during reproductive growth. Journal of experimental botany 41, 1579-1584. Wang, Y., Yu, Z., Li, Y., Wang, G., Liu, J., Liu, J., Liu, X., Jin, J., 2017. Microbial association with the dynamics of particulate organic carbon in response to the amendment of elevated CO2- derived wheat residue into a Mollisol. Science of the Total Environment 607, 972-981. Wildenschild, D., Sheppard, A.P., 2013. X-ray imaging and analysis techniques for quantifying pore-scale structure and processes in subsurface porous medium systems. Advances in Water Resources 51, 217-246. Wildenschild, D., Vaz, C., Rivers, M., Rikard, D., Christensen, B., 2002. Using X-ray computed tomography in hydrology: systems, resolutions, and limitations. Journal of Hydrology 267, 285- 297. 47 CHAPTER 3: O2 depletion and N2O production near decomposing plants directly measured using microsensors† Abstract Plant residues decomposing within the soil matrix are known to serve as hotspots of N2O production. However, the lack of technical tools for microscale in-situ N2O measurements limits understanding of hotspot functioning. Aim of this chapter was to assess performance of microsensor technology for evaluating the temporal patterns of N2O production in immediate vicinity to decomposing plant residues. I incorporated intact switchgrass leaves and roots into soil matrix and monitored O2 depletion and N2O production using electrochemical microsensors along with N2O emission from the soil. I also measured residue’s water absorption and β-glucosidase activity on the surface of the residue - the characteristics related to micro-environmental conditions and biological activity near the residue. N2O production in the vicinity of switchgrass residues began within 0-12 hours after the wetting, reached peak at ~0.6 day and decreased by day 2. N2O was higher near leaf than near root residues due to greater leaf N contents and water absorption by the leaves. However, N2O production near the roots started sooner than near the leaves, in part due to high initial enzyme levels on root surfaces. Electrochemical microsensor provided valuable information on N2O production near leaves and roots, its temporal dynamic, and the factors affecting it. The N2O production from residues measured by microsensors was consistent with the N2O emission from the whole soil, demonstrating the validity of the microsensors for N2O hotspot studies. † Originally published as: Kim, K., Kutlu, T., Kravchenko, A. and Guber, A., 2021. Dynamics of N2O in vicinity of plant residues: a microsensor approach. Plant and Soil, 462(1), pp.331-347. doi: 10.1007/s11104-021-04871-7 48 3.1 Introduction Nitrous oxide (N2O) efflux from agricultural soils is highly spatially and temporally variable (Smith and Tiedje, 1979; Goodroad et al., 1984; Parsons et al., 1991). Plant detritus, i.e., residues of plant roots and aboveground biomass incorporated within the soil matrix, play an important role as originators of ‘hotspots’ of N2O production and are in part responsible for high variability of N2O fluxes (Parkin, 1987; Garcia-Ruiz and Baggs, 2007). Such hotspots occur due to stimulation of microbial activity by the residues, which are abundant sources of carbon (C) and nitrogen (N). The hotspots possess greater microbial biomass, microbial diversity, and enzyme activity compared to the bulk soil, i.e., the soil not directly affected by plant residues (Kuzyakov and Blagodatskaya, 2015). In the bulk soil, oxygen (O2) (Khalil et al., 2004), C (Myrold and Tiedje, 1985; Miller et al., 2008; Senbayram et al., 2012), and N availability (Beauchamp, 1997; Blackmer and Bremner, 1978; Senbayram et al., 2012; Wu et al., 2018) are known to be major factors regulating overall N2O production. In soil micro-environment adjacent to the residues, C, N and O2 availability can be different from that in the bulk soil. Contents of C and N are higher within 4-6 mm distance from the decomposing residues, due to diffusion of decomposition products (Gaillard et al., 1999; Gaillard et al., 2003). Increased microbial respiration stimulated by utilization of the nutrients reduces O2 near the residues (McKenney et al., 2001; Miller et al., 2008; Chen et al., 2013). Water absorption by the residues from the surrounding soil can contribute to lower O2 within the residue and in the surrounding detritusphere (Kravchenko et al., 2017; Kim et al., 2020). The local anoxia stimulates denitrification, thus increases N2O production (Miller et al., 2008; Li et al., 2016; Kravchenko et al., 2017). However, how fast the O2 depletion occurs near the residue, and how it affects the magnitude of N2O production within the residue are still unknown. Moreover, what are 49 the main factors that determine the magnitude of N2O production from these plant residue-induced hotspots is not fully understood. One of the reasons for lingering poor understanding of N2O production from residue- induced hotspots is a common reliance on the use of ground plant materials, which are typically well mixed with the soil, in experimental soil N2O studies. In an intact soil, spatial distribution of residue fragments is highly heterogeneous and is known to be a crucial source of microscale resource heterogeneity (Loecke and Robertson, 2009). Mixing ground residues with the soil changes surface area of the residue, its contact with the soil, and the volume of the soil directly affected by the decomposing residues, leading to potentially significant discrepancies between the experimental results and what actually happens under field conditions (Kravchenko et al., 2018). Hence, work with intact residue fragments is essential for the studies of residue driven N2O hotspots. Another impediment to understanding the drivers of N2O production in the residue-induced hotspots is a lack of tools for measuring O2 depletion and N2O production in close proximity to the residue. As a result, field as well as laboratory studies typically measure only the N2O emitted into the atmosphere from the entire body of the sampled soil, not the N2O produced within the individual hotspots. Electrochemical microsensors allow in-situ non-destructive measurements of gas concentrations with fast response time. They have been used to examine spatial and temporal dynamics of O2 and N2O in biofilms (Nielsen et al., 1990; Dalsgaard and Revsbech, 1992), rhizosphere (Revsbech et al., 1999), soil aggregates (Højberg et al., 1994), sediments (Meyer et al., 2008) and soil profiles (Hansen et al., 2014; Liengaard et al., 2014) in µm to mm scales. Due to high spatial resolution, microsensors can conduct measurements in specific microsites within 50 soil matrix and, potentially, in vicinity to individual fragments of decomposing plant residues. However, they have never been used before for such purpose. Moreover, with few exceptions (Højberg et al., 1994), experiments with N2O and O2 microsensors have been conducted primarily under fully saturated soil conditions (Jørgensen and Elberling, 2012; Hansen et al., 2014; Liengaard et al., 2014). The reason is that (i) the anaerobic environment of fully saturated soils is favorable to denitrification and maximizes N2O, and (ii) the microsensor measurements are more stable and reliable in saturated conditions. However, N2O emissions from the soil can be substantial even in aerobic conditions, partly due to denitrification within decomposing plant residues (Li et al., 2016; Kravchenko et al., 2017). Thus, the use of microsensors in unsaturated soil with incorporated intact residue fragments can generate new insights into this important component of soil N2O emission. In contrast to initial N2O microsensors that required completely anoxic conditions to measure N2O precisely (Revsbech et al., 1988), recent electrochemical N2O microsensors (Unisense A/S, Arhaus, Denmark) can measure both dissolved and gaseous N2O with a detection limit of < 0.5 µmol·L-1 (µM). Yet, the question is how reliable and informative the data from these microsensors will be in vicinity to potential residue-induced N2O hotspots in unsaturated soil. The objective of this study was to evaluate the occurrence and temporal patterns of N2O production in immediate vicinity to switchgrass (Panicum Virgatum) leaf and root residues incorporated into the soil. Within the homogenized soil of my experimental setup, the intact switchgrass residues were expected to serve as the primary nuclei for hotspots of N2O production. I hypothesized that the patterns of N2O production will depend on the residue characteristics. I emulated the situation when the activity of such hotspots is maximized, that is, when dry soil containing the residue is subjected to a wetting event. I also created soil moisture and pore-size 51 distribution settings which were previously found to be optimal for promoting strong N2O production from plant residue-induced hotspots (Kravchenko et al., 2017). This chapter aimed at addressing the following research questions: (i) How soon after the wetting the enhanced N2O production at the surface of the residue begins? (ii) How long the enhanced production lasts? (iii) How well the N2O levels in the vicinity of the plant residue- induced hotspots are related to the O2 levels and N2O emissions from the soil into the atmosphere? (iv) How water absorption and enzyme activity of the plant residues affect the N2O production dynamics? 3.2 Materials and Methods 3.2.1 Soil and plant residues Soil and plant materials were collected from a field where switchgrass was grown in monoculture since 2008 as part of the Great Lakes Bioenergy Research Center (https://lter.kbs.msu.edu/research/long-term-experiments/glbrc-intensive-experiment/) Biofuel Cropping System Experiment located at Kellogg Biological Station (Michigan, U.S.A). The soil of the experimental site is classified as Kalamazoo loam (mesic Typic Hapludalfs) developed on glacial outwash (Oates et al., 2016). A composite soil sample was obtained from 5 randomly selected sites sampled at 5-10 cm depth. The collected soil was sieved through a 6 mm sieve to remove large stones and roots, and air-dried for a week. Air-dried soil was then sieved again to procure 1-2 aggregate fraction for the experiment. The decision to focus on the 1-2 mm aggregate fraction was driven by the previous findings that soil-incorporated plant residues made greater contribution to N2O emissions when surrounded by the soil with prevalence of large pores, a setting that was the best achieved by using the large aggregate fraction (Kravchenko et al., 2017). The soil (1-2 mm fraction) was brought to 52 30% gravimetric water content level and pre-incubated for ~10 days at 20 oC. The purpose of pre- incubation was, first, to reduce the contribution of Birch effect of enhanced microbial activity in wetted soil, magnified in this experiment by previous soil disturbance and sieving (Negassa et al., 2015); and, second, to eliminate seedlings germinated from weeds present within the 1-2 mm soil fraction. The studied 1-2 mm fraction contained 0.97% total C, 0.095% total N, 0.17 mg N/kg of NO3-, and 2.05 mg N/kg of NH4+. Switchgrass (var. Cave-in-rock) biomass was collected from 3 randomly selected sampling sites in October 2019. To minimize the disturbance, soil near switchgrass roots was removed by shovel and the entire plant body was plucked. In the laboratory, the plants were washed with distilled water and dried for ~3 weeks using a botanical press, leaves and roots separately. Dried and flattened leaves and roots were used for further analyses and experiments. 3.2.2 Microsensor experiment Overview: The experimental setup consisted of two boxes (8.7x9.4x3.3 cm3 each) filled with prepared 1-2 mm soil fraction with plant residues placed at a fixed position within each box (Fig. 3.1). To ensure the exact placement of the residues, a removable plastic frame was installed in the center of each box. The frame had a thin rectangular holder (area of 3.75 cm2) in the center. Flattened plant residues were placed within the holder, the frame with the holder was installed in the box. Then two microsensors, one for N2O and one for O2 (N2O-100 and OX-100 electrochemical microsensors with 100 µm tip, Unisense A/S, Aarhus, Denmark), were inserted through the openings on the side of the box. The microsensors were adjusted using the translation stage so as to ensure that the tips of both microsensors were in the immediate vicinity of each other and the surface of the residue (Fig. 3.1c). The openings through which the microsensors were inserted had a rubber cover that eliminated air flow into the box after the microsensors were in 53 place. Lastly, the box was filled with 100 g of the prepared soil to reach ~1 g·cm-3 soil bulk density. Figure 3.1: Experimental setup for microsensor measurements. (a) Flattened residue fragments in the holder. Holder was used to fix the location of residue fragments in the soil box. (b) The experimental setup. It shows boxes containing soil with incorporated residues, microsensors inserted into the boxes, and tubing connecting the air chambers above the soil boxes with the PAS device for measuring air concentrations of N2O and CO2. (c) Schemetic representation of the microsensor locations in vicinity to the residue within the soil box. The top of the box was covered by an air-impermeable chamber with 30 ml headspace volume equipped with outlets for measurements of CO2 and N2O emissions from the surface of the soil. After assembling the chambers, the microsensor monitoring in the dry soil was conducted 54 for approximately 3-4 hours. Then 30 mL of water was slowly added with a syringe from the top of the box to bring soil water content to 30% gravimetric content. Monitoring continued for subsequent ~5 days with every-minute microsensor readings and daily measurements of headspace N2O and CO2 using Photoacoustic Spectroscopy (PAS, INNOVA Air Tech Instruments, Denmark) using static chamber approach with a day interval for 5 days. I referred to the measurements next to the residue by the microsensors as N2O production, and to the readings by PAS from the headspace above the soil samples as N2O emission. Experimental design: The experiment was a randomized complete block design with 8 replications, that is, 8 runs. Each run included two identically equipped and monitored boxes. One box contained leaf residues and the other box contained root residues. In each run, the residues were assigned to the boxes at random. Because of sensor malfunctioning, only the first 4 runs produced useable O2 data. Residue preparation: For leaf monitoring, multiple flattened switchgrass leaf fragments were placed within the holder with minimal overlap. Each fragment was ~20 mm in length and 5- 8 mm in width (a typical width of the switchgrass leaves), with fragments for each experimental run cut from the same plant (Fig. 3.1a). For root monitoring, multiple flattened plant roots with diameters ranging from 0.1 to 1 mm and length of ~20 mm were placed within the holder. Sensor calibration: The sensors were calibrated before and after every experimental run. A two-point calibration was used for OX-100 sensors. The first calibration point (0 µM) was obtained in anoxic solution of sodium ascorbate and NaOH, and the second point (283.03 µM) was obtained in the air-aerated solution of DI water as described in the sensor manual (Oxygen Sensor User Manual, https://www.unisense.com). A four-point calibration at N2O concentrations of 0, 25, 50 55 and 100 µM was used for the N2O-100 sensors. The solutions were prepared by diluting N2O- saturated water, which was obtained by passing pure N2O through DI water. 3.2.3 Microsensor data processing After calibration, the data were filtered to remove the noise in the microsensor readings. The filtering was conducted as following: 1) Medians and standard errors were calculated for every 100-minute interval of the microsensor readings. 2) Only the readings within the range of median ± standard error were selected for further analyses. All O2 observations were adjusted to 320 µM of O2 at the point of water addition, which was the average value of O2 in the dry soil. N2O observations were adjusted so as to be set equal to 0 µM of N2O at the time of water addition. Data processing was performed using pandas library (Available at http://pandas.pydata.org/) in Python 3.6 (Python Software Foundation, available at http://www.python.org/). Three quantitative variables were derived from the adjusted N2O microsensor measurements: peak N2O production, time elapsed from the start of the soil wetting until the peak production, and cumulative N2O productions. In most samples the concentration of N2O increased after water addition and decreased after reaching the peak. In case of multiple peaks, the highest peak was used (e.g., Fig. 3.2 leaf rep4&5). The time between water application and the peak of N2O production was referred to as a ‘lag’. The values of peaks and lags for microsensor data records from all individual soil boxes are marked with black dotted lines in Fig. 3.2. Cumulative N2O production was calculated as the area under the microsensor curves and above zero. For O2, cumulative O2 depletion was calculated as the area under the base O2 concentration (320 µM). Cumulative N2O production was calculated for 1 day, 2 day, and 3 day periods to enable comparisons with N2O emission data, which were collected daily. The N2O production and 56 emission were the highest on day 1 and were substantially reduced in most samples by the end of day 2, thus no cumulative N2O calculations after day 3 were performed for subsequent days. Figure 3.2: Dynamics of N2O concentrations in the soil microcosms with (a) leaf and (b) root residues. Black dotted lines indicate the timepoint at which the lag time and the peak (maximum) concentration were determined. Colored area under each curve presents cumulative N2O production for 2 days. Red circles mark an example of artificial fluctuations in one of the experimental runs. 57 3.2.4 Plant analysis Switchgrass leaves and roots were subjected to water absorption measurements, zymography analysis, and total C and N measurements. Plant materials air-dried in the botanical press as described above were used for all the measurements. Water absorption by residue: Water absorption by leaves and roots incorporated into the soil was measured by placing plant residues (n=3) of the known mass within the wet soil, allowing them to equilibrate, and then determining their weight gains. Specifically, 4 g of dry soil was packed in 5 cm Ø cylinder, and soil water content was adjusted to 30% gravimetric soil water content, the level consistent with that used in the microsensor experiment. Then, ~12.8 mg of dry switchgrass residue was placed in a single layer on the surface of the soil, covered by another 4 g of soil, and more water was added to bring the top soil to 30% gravimetric water content. The amount of residue added to soil was such as to ensure that the surface area of the incorporated material was equal to ~3 cm2. After 4 hours of equilibration, the soil cylinders were disassembled; the residues were retrieved, cleaned from small soil particles attached to the surface with a brush, and weighed. The increase in the residue weight reflected the amount of water absorbed by the residue from the surrounding soil. β-glucosidase activity on residue surfaces: Spatial distribution of the β-glucosidase on the surface of the soil-incorporated leaves and roots was measured using zymography (Spohn et al., 2013; Guber et al., 2019). In a course of zymography, a membrane saturated with an enzyme- specific substrate is placed on the surface of the studied material (e.g., soil). The substrate diffuses from the membrane into the soil where a contact with the enzyme results in the substrate decomposition and a release of the fluorescent product. The map of the products distribution on 58 the membrane is visible in ultraviolet (UV) light and is representative of the enzyme activities on the studied surface. Soil and residue packing procedure for zymography was similar to that of the water absorption experiment. Specifically, a 15 g of 1-2 mm soil fraction was packed in a 4.7x2x3.3 cm3 soil box, and soil water content was adjusted to 30% (gravimetric). Then, ~30 mg of dry switchgrass residue was placed on top, covered by another 15 g of soil, and more water was added to bring the top layer of soil to 30%. Soil boxes were placed into 500 mL Mason jars, with 8 mL of distilled water added on the bottom to prevent soil drying; and incubated for 3 days at 20⁰C in the dark (n=2). Each soil box was taken out of the Mason jar twice during the incubation, on day 1 and day 3, to take zymography images. For that, one side of the box (4.7 x 3.3 cm2) was opened, and a 4x3 cm2 polyamide membrane filter (0.45 µm; Tao Yuan, China) soaked in 6 mM solution of 4-Methylumbelliferyl-β-D-Glucoside (Substrate) was placed on top of the soil surface. Substrate is the fluorogenic solution specific to β-glucosidase, which contains florescent product (4-Methylumbelliferone, MUF) which can be cleaved by enzymes, and fluorescence intensity is then used to calculate β-glucosidase activity. Soil surface with the membrane filter was photographed every 5 minutes for 40 minutes in total under the UV light, using Canon EF 75-300 mm f/4-5.6 III Telephoto Zoom Lens (Canon U.S.A. Inc., U.S.A). For calibration of florescence intensity, 5 µL of MUF standard solutions with concentrations of 1, 2, 5, 10, 50, 100 µM were added to 1 cm2 membranes and photographed in the UV light with camera setting described above. The parameters of nonuniform calibration were calculated as described in Guber et al. (2019). All zymography images were corrected for the background intensity by subtracting the first image. Then corrected zymography images were converted to MUF contents using the calibration parameters. Time series of MUF contents on the 59 images were used to calculate the enzymatic activities in the membrane pixels. The activity was calculated as a maximum slope of linear parts of MUF time series (9 points for 40 mins). The 0.27 cm2 wide area encompassing the residue was used to quantify the enzyme activity on the residues (Fig. 3.3). Figure 3.3: Examples of (a) boxes with soil and plant residues used for soil zymography and (b) resultant zymography images. Yellow dotted rectangles on (a) mark the areas that were subjected to zymography, i.e., membrane placement. The white dotted rectangle on (b) encompasses the area used to calculate the enzyme activity for the incubated plant residue. Total C and N values of the residues: Total C and N of switchgrass leaves and roots were measured using an Elemental Analyzer (ECS 4010 CHNSO Analyzer, Costech Analytical Technologies Inc., U.S.A) (3 replicates). Approximately 15-20 mg of the residue was used for each replicate sample. The residues were cut into small pieces (< 1 mm) using surgical scissors before being packed in the tin caps for the C and N measurements. 60 3.2.5 Statistical analysis Data analyses for comparisons between the residue types were conducted using PROC MIXED procedure (SAS 9.4, SAS Institute Inc., U.S.A) following recommendations by Milliken and Johnson (2009). For all quantitative variables derived from the microsensor N2O observations, e.g., peak N2O production, lag, and cumulative N2O production, the statistical models consisted of the fixed effect of the residue type (leaves and roots) and random effects of the experimental run and the sensor ID. The statistical model for the analysis of N2O and CO2 emissions in the headspace air above the soil boxes from PAS consisted of residue type, day since water addition, and their interaction as fixed effects; and the experimental run and run by the residue type interaction as random effects. The latter was used as an error term for testing the main effect of the residue type. Repeated measures approach was used to account for repeated measurements of N2O and CO2 from the same soil box during the experiment. The optimal variance-covariance structure was determined as such that produced the lowest AIC and BIC values (Milliken and Johnson, 2009). For both N2O and CO2, first-order autoregressive covariance structure was used in the final model. The statistical model for the analysis of the enzyme activity data consisted of residue type, day, and their interaction. Repeated measures approach was used here as well, using the same model selection appr oach as described above for N2O·CO2 data analysis. The model with unequal variances per day was used as the final model. Since the interaction between the residue type and day was significant, I conducted "slicing", a.k.a simple effect testing, of the interaction by day and by residue type. When the simple effect F-test was found to be statistically significant, comparisons among the days within each residue type were performed using t-test. 61 Relationships among the studied continuous variables, e.g., residue mass and N2O production, N2O production and N2O emission, were studied using regression analysis with PROC REG in SAS. To examine the difference in regression slopes between the two residue types, I used PROC MIXED with model consisting of the residue type effect and the interaction between the residue type and the residue mass, the latter as a continuous variable. Replication and sensor ID were considered as random factors. For all statistical models, the assumption of normality was checked by examining normal probability plots. When the normality assumption was violated, the original data was log- transformed. Equal variance assumption was checked using Levene’s test based on absolute residuals. When violated, I examined potential unequal variance models, and the models with the lowest AIC and BIC values were selected for further analyses. The results were reported as statistically significant when p-value was < 0.05 and as trends when p-value was < 0.10; and marked with * (p< 0.10), ** (p< 0.05), and *** (p< 0.01). 3.3 Results 3.3.1 Microsensor experiments After adding water, O2 concentration near both leaf and root residues decreased immediately or within 12 hours (Fig. 3.4a). In all 4 replications of root residues, O2 levels became stable soon after the initial drop, reaching to ~280 µM. This L-shape trend of O2 dynamics in the samples with root residues was not different from that of the control soil (with no residue added, data from preliminary experiment) (Fig. 3.4a). However, it was not always the case in the samples with leaf residues. In two of the four replications of leaf residue, O2 levels further decreased after the initial immediate decrease and reached minimums of 50-150 µM at ~1.5 day after water 62 addition. In the other two samples, O2 remained stable after the initial decrease, similar to the root samples. Figure 3.4: (a) Microsensor recorded O2 concentrations near switchgrass residues and control soil during the experiment. Water addition started at time 0. (b) Cumulative O2 depletion plotted vs. cumulative N2O production during the first 2 days of the experiment. Dotted line is the linear regression model fitted to the data (p< 0.10, one-tailed test). In 10 out of total 16 replicated samples, N2O concentration near the residues increased immediately after the water addition (Fig. 3.2; Leaf rep 1, 3, 4, 5, 6, and Root rep 1, 3, 5, 6, 8). In the remaining 6 samples (Leaf rep 2, 7, 8, and Root rep 2, 4, 7), N2O concentration started to increase within 12 hours. Typically, the N2O increases occurred almost simultaneously with the 63 drastic drops in O2 concentrations. The time elapsed until reaching the maximum N2O production (i.e., lag) was twice longer for leaf as compared to root residues (0.92 vs. 0.41, p< 0.05, Fig. 3.5d). Both the peak and the cumulative N2O productions were significantly higher in leaf residues compared to root residues (p< 0.05, Fig. 3.5a, b). Peak and cumulative N2O productions were strongly positively correlated (p< 0.01, Fig. 3.6a). Figure 3.5: Boxplot of quantitative measurements from microsensor and Photoacoustic Spectroscopy. (a) Peak N2O production - the maximum N2O concentration observed from 5 days of microcensor recordings, (b) Cumulative N2O production - the area under the microsensor curves and above zero for 2 days, (c) N2O emission –N2O from the surface of the soil measured from headspace, (d) lag N2O production – the time elaped from the start of the soil wetting until the peak production, (e) minumum O2 concentration - the lowest O2 concentration observed from 5 days of microcensor recordings, and (f) O2 depletion for 2 days - the area under the base O2 concentration. ** and * indicate significant differences between leaves and roots (p< 0.05 and 0.10). Black dots are individual observations from each run. The coefficients of variavtion were presented as percentage. 64 Across both residue types, greater mass of the residue resulted in higher peaks of N2O production (Fig. 3.6b). The positive trend was present in the leaf residue, where a 1 g of increase in leaf mass resulted in a 12.3 µM increase in N2O peak production (p< 0.10), however one observation point with an exceptionally high peak N2O production was excluded from this analysis. The residue mass did not affect peak N2O production in the root residues. The residue mass and cumulative N2O production were not significantly correlated, likely due to high variability of the latter (results not shown). N2O production was significantly and positively correlated with cumulative O2 depletion (p< 0.10, Fig. 3.4b). 65 Figure 3.6: (a) Peak N2O production plotted vs. cumulative N2O production for day 1 of the experiment. (b) The mass of plant residue incorporated into soil vs. peak N2O production from the residue. Dotted lines represent the regression models fitted to the entire data set (black) and to leaf data only (green). The regression model for roots (blue) was not statistically significant. Symbols * and ** mark the models statistically significant p< 0.10 and 0.05, respectively. The red circle marks the outlier data point that was not included in the regression analysis. N2O emissions from the soil displayed similar patterns to N2O microsensor observations. The emissions were the highest on day 1 (Fig. 3.7). N2O emissions from the samples with leaves 66 were numerically higher than those from the samples with roots through the first 5 days of incubation, but the difference tended to be statistically significant only on day 1 (p< 0.10). The differences between leaves and roots in terms of CO2 emission rates were not statistically significant. The CO2 and N2O emissions were positively related (p< 0.01, Fig. 3.8), and the relationship between them was stronger in root (R2=0.57) than in leaf (R2=0.41) residue samples. Cumulative N2O production and cumulative N2O mission also were positively correlated to each other for day 1, day 2, and day 3 of the experiment (Fig. 3.9). 67 Figure 3.7: (a) N2O and (b) CO2 emission rates from the soil boxes with incorporated leaf and root residues, measured using Photoacoustic Spectroscopy. Vertical lines represent standard errors. Asterisk * indicates significant differences between leaves and roots (p< 0.10). 68 Figure 3.8: Relationship between CO2 and N2O emission (day 1-5). Dotted lines are regression models for leaf (green) and root (blue). Symbol *** marks the significance of the slopes (p< 0.01). Figure 3.9: Relationship between cumulative N2O productions (measured from soil pore using microsensor) in and emissions (measured from headspace using photoacoustic spectroscopy) from the microcosms (a) for 1 day, (b) 2 days, and (c) 3 days of the experiment. Dotted lines represent linear regression models. All regressions were statistically significant at p< 0.01 and 0.05 (marked with *** and **). There were no significant differences between regression slopes of leaves and roots. 69 3.3.2 Plant analysis Both leaf and root residues absorbed significant amount of water from the surrounding soil (Fig. 3.10a). Leaves absorbed ~1.4 g of water per each g of air-dry biomass, while roots only absorbed 1 g of water per g of biomass (p< 0.10, Fig. 3.10a). Root residues had more C but less N than leaves, resulting in contrasting C:N ratios between the two (24.0 vs. 82.7, Table 3.1). Average β-glucosidase activities from the surface of the residues are presented in Fig. 3.10b. At the first day of the experiment, the enzyme activity was more than 40 times higher on root as compared to leaf surfaces. However, by day 3 the β-glucosidase activity on roots substantially decreased while on leaves it increased, resulting in no statistical differences between the residue types. Figure 3.10: (a) Average water absorption levels by leaf and root residues. The difference between leaves and roots is significant at p< 0.10. (b) Average enzyme (b-glucosidase) activity at the surface of the plant residues at day 1 and 3 of the experiment. Different letters mark significant differences between the days within each residue type (p< 0.10). Symbol ** indicates the significant differences between residue types at a given day (p< 0.05). Vertical lines represent standard errors. 70 Table 3.1: Total carbon and nitrogen contents in the plant residues. Plant type Carbon (w %)*** Nitrogen (w %)*** C:N ratio*** Leaf 44.2 (0.54) 1.90 (0.17) 24.0 (2.5) Root 47.4 (0.40) 0.58 (0.08) 82.7 (12.1) *** indicates significant differences between leaf and roots (p< 0.01). 3.4 Discussion and Conclusion This study demonstrated the utility of the microsensors in monitoring the O2 and N2O levels in immediate vicinity to the soil incorporated plant-residues, which are known originators of N2O hotspots (Parkin, 1987). The study provided the answers to the research questions I posed. In the conditions optimal for both microbial activity and gas diffusion, that is, at 48% water-filled pore space (WFPS) and in abundant presence of large air-filled soil pores of this study, the N2O is quickly emitted out of the soil in the amounts positively associated with its production (Fig. 3.9). Yet, the strength of the association between the N2O near the residues and that emitted into the atmosphere deteriorates over time, reflecting the short-lived nature of the residue hotspots' contribution to the emissions. The origin of the hotspot, i.e., leaves vs. roots, affects the magnitude of the production and emission as well as their temporal dynamic (Figs. 3.2 and 3.7). 3.4.1 Effect of incorporated switchgrass leaves and roots on N2O production and emission Greater amounts of N2O produced (for 2 days) from incorporated leaves than roots (Fig. 3.5b) result from lower C:N ratio and higher N contents of the portions of leaf residues readily available to microbial decomposers. While here I only measured total C and N plant contents (Table 3.1), marked differences between leaves and roots, especially, in terms of total C:N ratio, suggest that similarly contrasting differences likely occurred in labile portions of the plant tissues. While low C:N ratio in residues can promote mineralization, high C:N ratio can lead to N 71 immobilization, reducing N2O emissions (Miller et al., 2008). Also, microorganisms can utilize C and N from leaves more efficiently than those from the roots (Garcia-Ruiz and Baggs, 2007; Partey et al., 2014). Though not measured in this study, contrasting organic chemistry of leaves and roots likely played a role as well. Higher soluble organic C and N in the leaves are known to result in rapid decomposition, while higher concentration of cellulose, bound phenols and lignin phenols in roots can retard decomposition (Birouste et al., 2012; Uselman et al., 2012; Wang et al., 2015). The soil used in this study had low total C, total N, and inorganic N levels compared to the residue, indicating that the N2O production in my soil samples was primarily driven by the microbes that relied on the readily available substrates from the residues, not soil, as their main energy source. Thus, leaves, which provided more available nutrients resulted in greater N2O production. These results are in line with previous studies reporting positive correlations between available C and denitrification (Myrold and Tiedje, 1985; Miller et al., 2008). While it was expected that greater size of the incorporated residue would be associated with greater N2O production (Garcia-Ruiz and Baggs, 2007), that trend was significant only in leaf residue samples (Fig. 3.6b). I attribute this result in part to a narrower range of root mass compared to leaf mass - as the roots used in the study tended to weigh somewhat less than the leaves. Differences in residue chemistry between leaves and roots also likely contributed to the observed differences in correlation strengths. Larger water absorption was another factor that contributed to higher N2O production in the leaves (Fig. 3.10a), as it stimulated development of anoxic conditions and denitrification (Kravchenko et al., 2017). Greater N2O production near the leaves translated into higher N2O emissions during the first day of incubation (Fig. 3.7a). However, afterwards the difference between the leaves and roots disappeared, consistent with an overall decrease in the strength of the association between N2O 72 near the residues and the emitted N2O (Fig. 3.9) and pointing to the reduced importance of the residue's contribution to N2O emissions. It is also possible that the reduction of N2O to N2 contributed to the decreased association between produced and emitted N2O. However, the relatively large (> 30 µm) air-filled pores dominating soil pore-size distribution of the studied soil (Toosi et al., 2017) are unlikely to cause complete anoxic conditions within the soil. The final denitrification product is expected to be N2O rather than N2 when the oxygen in the pore is sufficient (Hwang and Hanaki, 2000). Most of produced N2O likely quickly escaped through the air-filled pores, contributing to the positive relationship between N2O production and emission. It contrasts other works which used fully saturated soils and reported delays in N2O emission (Markfoged et al., 2011). While the CO2 and N2O emissions were positively correlated in both leaves and roots (Fig. 3.8), variations in CO2 emissions explained 57% of variations in N2O emissions in the root residues, while only 41% in the leaf residue samples (Fig. 3.8). Positive correlations between CO2 and N2O emissions reflect stimulated microbial activity in both CO2 and N2O production (Azam et al., 2002; Millar and Baggs, 2004), and the role of C utilization due to increased microbial activity in N2O production (De Catanzaro and Beauchamp, 1985; Millar and Baggs, 2004; Hayashi et al., 2015). Weaker association between CO2 and N2O in the leaf residue samples further highlights the hotspot nature of the N2O production within the soil with incorporated decomposing leaves. Indeed, the anoxic conditions developed in response to greater water absorption by the leaves (Fig. 3.10a) were conducive to denitrification and to resultant N2O production and emission, while unfavorable to CO2 production. 73 3.4.2 Effect of incorporated switchgrass leaves and roots on N2O temporal dynamic On average, the N2O levels near the root residues reached the peak level faster than near the leaves, i.e., within 0.4 day and 0.8 day, respectively (Fig. 3.5d). A more rapid start of N2O production in vicinity of the roots probably resulted from a greater presence of inherent extracellular enzymes on the root surfaces (Fig. 3.10b). The enzymes were produced by both the roots and soil microorganisms when the roots were alive and remained since on the root surfaces (Razavi et al., 2016). Nagahashi and Baker (1984) showed that even after the roots were dead, washed, and dried prior to incubation, β-glucosidase still remained on their surfaces and was not readily removed by washing. Even in fumigated soil the extracellular enzymes retained their activity for ~12 weeks (Schimel et al., 2017). Since β-glucosidase measured in this study is one of the common enzymes produced by both roots and microorganisms, it can be regarded as an indicator of such overall extracellular enzyme presence (Kang et al., 1998; Sinsabaugh et al., 2008; Cayuela et al., 2009). When the roots were rewetted, the inherent enzymes were activated, and immediately started hydrolyzing the residue. The enzymes remaining on the roots likely accelerated root decomposition, and led to subsequently faster initiation of N2O production, as compared to the leaves. This observation is consistent with the earlier findings that preexisting denitrifying enzymes in the soil govern the initiation of denitrification, while it takes ~6 hours until the enzymes are newly synthesized by microbes using energy supplied from surroundings (Smith and Tiedje, 1979). This study supports the importance of inherent enzymes not only in the soil but also on the surfaces of decomposing roots. The activity of inherent enzymes on the roots decreased by day 3, possibly because their consumption rate was greater than the rate of new synthesis (Smith and Tiedje, 1979). Synthesis of new enzymes by microorganisms is possible only when there are sufficient available nutrients 74 (Allison and Vitousek, 2005; Wallenstein and Weintraub, 2008). As indicated from total plant N content (Table 3.1), N was not as readily available in the roots as in the leaves in this tudy, leading to low enzyme activity after the inherent enzymes were consumed. Enzyme activity in the leaves, on the contrary, was low at the beginning and significantly increased by day 3 (Fig. 3.10b), indicating new enzyme production, likely stimulated by higher levels of available N. Longer lag and greater peak of N2O production in leaves seemed to be related to newly synthesized enzyme activity at the surface of the residues. Longer lag in the leaves as compared to roots is also attributable to the presence of epicuticular layer on the leaf surfaces that prevents the water loss (Bragg et al., 2020; Riederer and Schreiber, 2001; Yeats and Rose, 2013). The release of labile substrates might have been delayed by this hydrophobic layer, leading to the delay of the peak in leaf residues. While the peak N2O production is a function of the amount of dissolved organic C and N, lag is likely more a function of the release rate of the dissolved organic matter from the residues. 3.4.3 O2 at plant residue surfaces and its relationship with N2O I observed a weak tendency for O2 near the residues to be more depleted in the leaf than in the root samples (Fig. 3.4a). In all root samples and 2 of the 4 leaf samples, the dynamics of O2 concentrations near the residues was not substantially different from that of the control soil, suggesting that the changes in O2 were caused by an inflow of water into the air-filled pore space of the initially dry soil samples, and not by the plant residue decomposition. It is also possible that O2 sensor tips of these samples was placed in the air-filled large pores, thus could not reflect the overall O2 changes near the residues. Yet, a marked decrease in O2 that took place iN2Of the leaf residue samples following the initial drop suggested that enhanced leaf decomposition did lead to greater O2 depletion near the leaves. Overall, leaves are known to decompose faster than roots due, 75 in part, to their lower lignin:N ratios (Steffens et al., 2015) and lower C:N ratio (Edmonds, 1980; Baggs et al., 2000; Zhang et al., 2008) (Table 3.1). Faster decomposition is associated with greater microbial respiration and growth, thus greater O2 consumption (Chen et al., 2013). Hence, these large decreases after O2 reached the initial short plateau can be due to enhanced O2 consumption occurred during leaf residue decomposition. Greater water absorption is possibly another potential contributor to greater O2 depletion near leaf residues (Fig. 3.10a), since water absorption by the residue fragments can induce higher water contents in their vicinity (~150 µm) (Kim et al., 2020), consequently, reducing O2 concentrations (Kravchenko et al., 2017). The lack of O2 stimulates denitrification and promotes N2O production (Castaldi, 2000; McKenney et al., 2001). I observed a simultaneous occurrence of O2 depletion and N2O production, and a positive correlation between the cumulative O2 depletion and N2O production (p< 0.10, Fig. 3.4b). It contrasts with Rohe et al. (2020) who did not find significant relationship between microsensor O2 observations and denitrification (N2O and N2O+N2). However, Rohe et al. (2020) conducted O2 measurements at local microsites representing only 0.2% of the total soil volume, while they assessed denitrification from the entire soil samples. In this study N2O and O2 were measured in close spatial proximity to each other (< 1 mm distance). The discrepancy between these two studies reflects high spatial variability of N2O production and emphasizes the necessity of smaller-scale approach to understand N2O hotspots. Still, O2 depletion explained only 25% of variation in N2O production near the residues (Fig. 3.4b). One possible reason for the relatively weak association between O2 and N2O is that some of N2O was produced via nitrification, which likely occurred in O2 rich 48% WFPS experimental settings of this study. Although denitrification is responsible for production of more than 50% of N2O in residue-incorporated soils, nitrification is another substantial source of N2O 76 (Li et al., 2016). Another possible reason is the complete denitrification of N2O to N2. Even though my experimental setup was designed to maximize the N2O production and minimize its conversion to N2, complete denitrification always occurs, and it is especially prominent when O2 is depleted to less than 25% of the atmospheric level (Morley and Baggs, 2010). It corresponds to 75 µM O2 concentration in this study, thus in the samples with leaves which had the greatest cumulative O2 depletion (Fig. 3.4b), the relationship between O2 and N2O might have been weakened due to further denitrification. 3.4.4 Evaluation of microsensors as a tool for hotspot detection Even under well-controlled experimental settings with precisely placed residues, sieved/preincubated soil, and stable temperature and humidity in the laboratory, there was still a substantial variability in microsensor measurements of O2 depletion and N2O production in vicinity of the residues (Fig. 3.5). The main cause of such variability are natural variations in characteristics of the hotspots themselves. The experimental set up was designed so as to minimize the variations, i.e., I used sieved 1-2 mm soil fraction, which was cleaned of particulate organic fragments and stones and pre-incubated. Yet, the sizes, locations, water or air-filled status, and connectivity of pores in vicinity of each plant residue were not controllable. The development and activity of microbial hotspots can be significantly affected by these micro-scale conditions, leading to variations of O2 depletion, N2O production and emission. Another possible cause is variability in micro-environmental conditions surrounding the microsensor tips. For example, after the microsensor was inserted and soil and water were added to the experimental box, it was not possible to confirm whether the sensor's tip ended up being within the water or in the air. The sensor's tip could have been covered by menisci of incoming water or it could have been located within a trapped air between soil aggregates. Depending on the 77 location and the distance to air-filled atmosphere-connected pores, O2 concentration measured by microsensors can be substantially different even within the same soil (Rohe et al., 2020). Placing sensor tips in certain positions within pores to observe the gas changes at the surface of residues is even more challenging, as implied by highly variable O2 depletion pattern in leaves in this study. The microsensor’s 100 µm-scale resolution measures the gas dynamics in a certain pore near the residue, and might not fully represent the dynamics occurring on the entire residue surface. Moreover, concentrations and fluxes in water and air are drastically different and not reflected by the gas partial pressure values – the actual data recorded by the microsensors. This experiment stresses the difficulties in measurements of N2O production, especially in unsaturated conditions of soil at a microscale. Another peculiarity observed in microsensor performances in my study were occasional simultaneous fluctuations in the records from all operating microsensors (Fig. 3.2). The artificial nature of such fluctuations was evident from identical patterns present simultaneously in N2O and O2 microsensors of both boxes. The artificial patterns were small compared to the peak measurements, thus non-detectable during the first 2 days of each experimental run, but they became visible when the microsensor readings decreased approaching atmospheric levels (Fig. 3.2). What induces such fluctuations and how to minimize them requires further investigation. Despite discussed above limitations and difficulties, it should be emphasized that the use of microsensors enabled generating valuable information on N2O production near the residue hotspots and on its temporal dynamic. The obtained information was consistent with the N2O emissions measured using the traditional approach and agreed well with the effects of plant residue characteristics on N2O production expected based on theoretical considerations and published literature. My results emphasize the validity and usefulness of the microsensors for studies of soil 78 N2O production hotspots aimed at understanding mechanisms of micro-scale N2O hotspot production. 79 ACKNOWLEDGEMENTS Chelsea Mamott and GLBRC communication team gratefully helped with figure preparations. I also thank Dr. Dirk Colbry for help with data processing and Maxwell Oerther for assisting in laboratory work. The work was funded in part by the National Science Foundation’s Geobiology and Low Temperature Geochemistry Program (Award 1630399). Support for this research was provided by the Great Lakes Bioenergy Research Center, U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research (Award DE-SC0018409), by the National Science Foundation Long-term Ecological Research Program (DEB 1832042) at the Kellogg Biological Station, and by Michigan State University AgBioResearch. 80 REFERENCES 81 REFERENCES Allison, S.D. and Vitousek, P.M., 2005. Responses of extracellular enzymes to simple and complex nutrient inputs. Soil Biology and Biochemistry. 37, 937-944. Azam, F., Müller, C., Weiske, A., Benckiser, G., Ottow, J., 2002. Nitrification and denitrification as sources of atmospheric nitrous oxide–role of oxidizable carbon and applied nitrogen. Biology and fertility of soils 35, 54-61. Baggs, E., Rees, R., Smith, K., Vinten, A., 2000. Nitrous oxide emission from soils after incorporating crop residues. Soil use and management 16, 82-87. Beauchamp, E., 1997. Nitrous oxide emission from agricultural soils. Canadian Journal of Soil Science 77, 113-123. Birouste, M., Kazakou, E., Blanchard, A. and Roumet, C., 2012. Plant traits and decomposition: are the relationships for roots comparable to those for leaves? Annals of Botany. 109, 463-472. Blackmer, A., Bremner, J., 1978. Inhibitory effect of nitrate on reduction of N2O to N2 by soil microorganisms. Soil Biology and Biochemistry 10, 187-191. Bragg, J., Tomasi, P., Zhang, L., Williams, T., Wood, D., Lovell, J.T., Healey, A., Schmutz, J., Bonnette, J.E. and Cheng, P., 2020. Environmentally responsive QTL controlling surface wax load in switchgrass. Theoretical and Applied Genetics. 133, 3119-3137. Castaldi, S., 2000. Responses of nitrous oxide, dinitrogen and carbon dioxide production and oxygen consumption to temperature in forest and agricultural light-textured soils determined by model experiment. Biology and fertility of soils 32, 67-72. Cayuela, M., Sinicco, T., Mondini, C., 2009. Mineralization dynamics and biochemical properties during initial decomposition of plant and animal residues in soil. Applied Soil Ecology 41, 118- 127. Chen, H., Li, X., Hu, F., Shi, W., 2013. Soil nitrous oxide emissions following crop residue addition: a meta‐analysis. Global change biology 19, 2956-2964. Dalsgaard, T., Revsbech, N.P., 1992. Regulating factors of denitrification in trickling filter biofilms as measured with the oxygen/nitrous oxide microsensor. FEMS Microbiology Letters 101, 151-164. De Catanzaro, J., Beauchamp, E., 1985. The effect of some carbon substrates on denitrification rates and carbon utilization in soil. Biology and fertility of soils 1, 183-187. Edmonds, R.L., 1980. Litter decomposition and nutrient release in Douglas-fir, red alder, western hemlock, and Pacific silver fir ecosystems in western Washington. Canadian Journal of Forest Research 10, 327-337. 82 Gaillard, V., Chenu, C., Recous, S., 2003. Carbon mineralisation in soil adjacent to plant residues of contrasting biochemical quality. Soil Biology and Biochemistry 35, 93-99. Gaillard, V., Chenu, C., Recous, S., Richard, G., 1999. Carbon, nitrogen and microbial gradients induced by plant residues decomposing in soil. European Journal of Soil Science 50, 567-578. Garcia-Ruiz, R., Baggs, E., 2007. N2O emission from soil following combined application of fertiliser-N and ground weed residues. Plant and Soil 299, 263-274. Goodroad, L., Keeney, D., Peterson, L., 1984. Nitrous oxide emissions from agricultural soils in Wisconsin. Journal of Environmental Quality 13, 557-561. Guber, A.K., Kravchenko, A.N., Razavi, B.S., Blagodatskaya, E., Kuzyakov, Y., 2019. Calibration of 2‐D soil zymography for correct analysis of enzyme distribution. European Journal of Soil Science 70, 715-726. Hansen, M., Clough, T.J., Elberling, B., 2014. Flooding-induced N2O emission bursts controlled by pH and nitrate in agricultural soils. Soil Biology and Biochemistry 69, 17-24. Hayashi, K., Tokida, T., Kajiura, M., Yanai, Y., Yano, M., 2015. Cropland soil–plant systems control production and consumption of methane and nitrous oxide and their emissions to the atmosphere. Soil science and plant nutrition 61, 2-33. Højberg, O., Revsbech, N.P., Tiedje, J.M., 1994. Denitrification in soil aggregates analyzed with microsensors for nitrous oxide and oxygen. Soil Science Society of America Journal 58, 1691- 1698. Hwang, S. and Hanaki, K., 2000. Effects of oxygen concentration and moisture content of refuse on nitrification, denitrification and nitrous oxide production. Bioresource Technology. 71, 159- 165. Jørgensen, C.J., Elberling, B., 2012. Effects of flooding-induced N2O production, consumption and emission dynamics on the annual N2O emission budget in wetland soil. Soil Biology and Biochemistry 53, 9-17. Kang, H., Freeman, C., Lock, M., 1998. Trace gas emissions from a north Wales fen-role of hydrochemistry and soil enzyme activity. Water, Air, and Soil Pollution 105, 107-116. Khalil, K., Mary, B., Renault, P., 2004. Nitrous oxide production by nitrification and denitrification in soil aggregates as affected by O2 concentration. Soil Biology and Biochemistry 36, 687-699. Kim, K., Guber, A., Rivers, M., Kravchenko, A., 2020, Contribution of decomposing plant roots to N2O emissions by water absorption, Geoderma 375, 114506. Kravchenko, A., Toosi, E., Guber, A., Ostrom, N., Yu, J., Azeem, K., Rivers, M., Robertson, G., 2017. Hotspots of soil N2O emission enhanced through water absorption by plant residue. Nature Geoscience 10, 496-500. 83 Kravchenko, A., Fry, J. and Guber, A., 2018. Water absorption capacity of soil-incorporated plant leaves can affect N2O emissions and soil inorganic N concentrations. Soil Biology and Biochemistry. 121, 113-119. Kuzyakov, Y., Blagodatskaya, E., 2015. Microbial hotspots and hot moments in soil: concept & review. Soil Biology and Biochemistry 83, 184-199. Li, X., Sørensen, P., Olesen, J.E., Petersen, S.O., 2016. Evidence for denitrification as main source of N2O emission from residue-amended soil. Soil Biology and Biochemistry 92, 153-160. Liengaard, L., Figueiredo, V., Markfoged, R., Revsbech, N.P., Nielsen, L.P., Prast, A.E., Kühl, M., 2014. Hot moments of N2O transformation and emission in tropical soils from the Pantanal and the Amazon (Brazil). Soil Biology and Biochemistry 75, 26-36. Loecke, T.D., Robertson, G.P., 2009. Soil resource heterogeneity in terms of litter aggregation promotes nitrous oxide fluxes and slows decomposition. Soil Biology and Biochemistry 41, 228- 235. Markfoged, R., Nielsen, L.P., Nyord, T., Ottosen, L.D.M., Revsbech, N.P., 2011. Transient N2O accumulation and emission caused by O2 depletion in soil after liquid manure injection. European Journal of Soil Science 62, 541-550. McKenney, D., Drury, C., Wang, S., 2001. Effects of oxygen on denitrification inhibition, repression, and derepression in soil columns. Soil Science Society of America Journal 65, 126- 132. Meyer, R.L., Allen, D.E., Schmidt, S., 2008. Nitrification and denitrification as sources of sediment nitrous oxide production: A microsensor approach. Marine Chemistry 110, 68-76. Millar, N., Baggs, E., 2004. Chemical composition, or quality, of agroforestry residues influences N2O emissions after their addition to soil. Soil Biology and Biochemistry 36, 935-943. Milliken, G.A., Johnson, D.E., 2009. Analysis of messy data volume 1: designed experiments. CRC Press. Morley, N. and Baggs, E., 2010. Carbon and oxygen controls on N2O and N2 production during nitrate reduction. Soil Biology and Biochemistry. 42, 1864-1871. Myrold, D.D., Tiedje, J.M., 1985. Establishment of denitrification capacity in soil: effects of carbon, nitrate and moisture. Soil Biology and Biochemistry 17, 819-822. Nagahashi, G., Baker, A.F., 1984. β-Glucosidase activity in corn roots: Problems in subcellular fractionation. Plant physiology 76, 861-864. Negassa, W.C., Guber, A.K., Kravchenko, A.N., Marsh, T.L., Hildebrandt, B., Rivers, M.L., 2015. Properties of soil pore space regulate pathways of plant residue decomposition and community structure of associated bacteria. PLoS one 10. 84 Nielsen, L.P., Christensen, P.B., Revsbech, N.P., Sørensen, J., 1990. Denitrification and oxygen respiration in biofilms studied with a microsensor for nitrous oxide and oxygen. Microbial ecology 19, 63-72. Oates, L.G., Duncan, D.S., Gelfand, I., Millar, N., Robertson, G.P., Jackson, R.D., 2016. Nitrous oxide emissions during establishment of eight alternative cellulosic bioenergy cropping systems in the North Central United States. Gcb Bioenergy 8, 539-549. Parkin, T.B., 1987. Soil microsites as a source of denitrification variability 1. Soil ScienceSociety of America Journal 51, 1194-1199. Parsons, L.L., Smith, M.S., Murray, R.E., 1991. Soil denitrification dynamics: spatial and temporal variations of enzyme activity, populations, and nitrogen gas loss. Soil Science Society of America Journal 55, 90-95. Partey, S., Preziosi, R., Robson, G., 2014. Improving maize residue use in soil fertility restoration by mixing with residues of low C-to-N ratio: effects on C and N mineralization and soil microbial biomass. Journal of soil science and plant nutrition 14, 518-531. Razavi, B.S., Zarebanadkouki, M., Blagodatskaya, E., Kuzyakov, Y., 2016. Rhizosphere shape of lentil and maize: spatial distribution of enzyme activities. Soil Biology and Biochemistry 96, 229- 237. Revsbech, N., Pedersen, O., Reichardt, W., Briones, A., 1999. Microsensor analysis of oxygen and pH in the rice rhizosphere under field and laboratory conditions. Biology and fertility of soils 29, 379-385. Revsbech, N.P., Nielsen, L.P., Christensen, P.B., Sørensen, J., 1988. Combined oxygen and nitrous oxide microsensor for denitrification studies. Applied and environmental microbiology 54, 2245- 2249. Riederer, M. and Schreiber, L., 2001. Protecting against water loss: analysis of the barrier properties of plant cuticles. Journal of experimental botany. 52, 2023-2032. Rohe, L., Apelt, B., Vogel, H.-J., Well, R., Wu, G.-M., Schlüter, S., 2020. Denitrification in soil as a function of oxygen supply and demand at the microscale. Biogeosciences Discussions, 1-32. Schimel, J., Becerra, C.A., Blankinship, J., 2017. Estimating decay dynamics for enzyme activities in soils from different ecosystems. Soil Biology and Biochemistry 114, 5-11. Senbayram, M., Chen, R., Budai, A., Bakken, L., Dittert, K., 2012. N2O emission and the N2O/(N2O+ N2) product ratio of denitrification as controlled by available carbon substrates and nitrate concentrations. Agriculture, Ecosystems & Environment 147, 4-12. Sinsabaugh, R.L., Lauber, C.L., Weintraub, M.N., Ahmed, B., Allison, S.D., Crenshaw, C., Contosta, A.R., Cusack, D., Frey, S., Gallo, M.E., 2008. Stoichiometry of soil enzyme activity at global scale. Ecology letters 11, 1252-1264. 85 Smith, M.S., Tiedje, J.M., 1979. Phases of denitrification following oxygen depletion in soil. Soil Biology and Biochemistry 11, 261-267. Spohn, M., Carminati, A., Kuzyakov, Y., 2013. Soil zymography–a novel in situ method for mapping distribution of enzyme activity in soil. Soil Biology and Biochemistry 58, 275-280. Steffens, C., Helfrich, M., Joergensen, R.G., Eissfeller, V., Flessa, H., 2015. Translocation of 13C- labeled leaf or root litter carbon of beech (Fagus sylvatica L.) and ash (Fraxinus excelsior L.) during decomposition–A laboratory incubation experiment. Soil Biology and Biochemistry 83, 125-137. Toosi, E., Kravchenko, A., Guber, A., Rivers, M., 2017. Pore characteristics regulate priming and fate of carbon from plant residue. Soil Biology and Biochemistry 113, 219-230. Uselman, S.M., Qualls, R.G. and Lilienfein, J., 2012. Quality of soluble organic C, N, and P produced by different types and species of litter: Root litter versus leaf litter. Soil Biology and Biochemistry. 54, 57-67. Wallenstein, M.D. and Weintraub, M.N., 2008. Emerging tools for measuring and modeling the in situ activity of soil extracellular enzymes. Soil Biology and Biochemistry. 40, 2098-2106. Wang, J.J., Tharayil, N., Chow, A.T., Suseela, V. and Zeng, H., 2015. Phenolic profile within the fine‐root branching orders of an evergreen species highlights a disconnect in root tissue quality predicted by elemental‐and molecular‐level carbon composition. New Phytologist. 206, 1261- 1273. Wu, D., Wei, Z., Well, R., Shan, J., Yan, X., Bol, R. and Senbayram, M., 2018. Straw amendment with nitrate-N decreased N2O/(N2O+ N2) ratio but increased soil N2O emission: A case study of direct soil-born N2 measurements. Soil Biology and Biochemistry. 127, 301-304. Yeats, T.H. and Rose, J.K., 2013. The formation and function of plant cuticles. Plant physiology. 163, 5-20. Zhang, D., Hui, D., Luo, Y., Zhou, G., 2008. Rates of litter decomposition in terrestrial ecosystems: global patterns and controlling factors. Journal of Plant Ecology 1, 85-93. 86 CHAPTER 4: Development of soil micro-environment in plant detritusphere and its contribution to C and N cycling‡ Abstract The detritusphere is one of the most important sources of N2O, however, understanding of how N2O emission from the detritusphere is influenced by soil properties remains elusive. Here, I evaluated the effects of pore architecture and soil moisture on N2O emission during the decomposition of in-situ grown roots of switchgrass, an important bioenergy crop. I combined dual isotopic labeling (15C and 15N) with zymography to gain insights into the location of the microbial N2O production in soils with contrasting pore architectures. The effect of soil pore architecture on N2O emissions was 6 times greater than that of soil moisture. Soil dominated by > 30 μm Ø pores (i.e., large-pore soil) had higher chitinase activity than the soil dominated by < 10 μm Ø pores (i.e., small-pore soil), especially near roots. The chitinase activity on decomposing roots was positively correlated with root-derived N2O emission, indicating that N released by root decomposition was an important source of N2O. Greater N2O and N2 emissions were induced by switchgrass roots in soils dominated by large-pores compared to small-pore soils. The micro-environment developed near decomposing roots of large-pore soil also resulted in positive N2O priming. This study challenged the traditional view on soil moisture as the main factor of N2O production. N2O production and emission was most intensive in microbial activity hotspots (i.e., rhizosphere legacy) in the large pores, where decomposed roots release mineral N as the main N2O source. ‡ Originally submitted as: Kim, K., Gil, J., Ostrom, N., Gandhi, H., Oerther, M., Kuzyakov, Y., Guber, A. & Kravchenko, A. (2021). Soil pore architecture and rhizosphere legacy define N2O production in root detritusphere, Soil Biology and Biochemistry. (In revision) 87 4.1 Introduction Soil pore architecture and soil moisture are crucial environmental influences on production pathways, transport, and emission of N2O due to their regulation of the O2 availability for soil microbes (Bollmann and Conrad, 1998; Butterbach-Bahl et al., 2013; Castellano et al., 2010; Chen et al., 2013). The biological processes of nitrification and denitrification are responsible for most upland soil N2O production (Bracken et al., 2021; Velthof et al., 2002), although other pathways such as dissimilatory nitrate reduction to ammonium (DNRA) and chemodenitrification contribute to N2O (Butterbach-Bahl et al., 2013; Morley and Baggs, 2010). Previous source partitioning studies showed nitrification dominating in relatively dry soils, e.g., at soil water-filled pore space (WFPS) of 40-60%, while denitrification gained importance at higher soil water contents, reaching maximum levels at 70-90% WFPS (Bateman and Baggs, 2005; Bracken et al., 2021; Davidson et al., 1986; Ruser et al., 2006; van der Weerden et al., 2012). Under conditions below 75-80% WFPS, higher water contents and finer-textured soils formulate anaerobic microsites favorable to denitrification (Hénault et al., 2012; Ruser et al., 2006). Consequently, the final product of denitrification is gradually switched from N2O to dinitrogen (N2), reducing the N2O:N2 molar ratio (Schaufler et al., 2010; Zaman et al., 2012). The N2O:N2 ratio reaches to plateau around 80% WFPS (Rohe et al., 2021), where the total emission of N2O+N2 reaches its maximum. Generally, in upland soil, the absolute amount of N2O emission tends to increase along with soil moisture content until 80% WFPS (Schaufler et al., 2010), and as overall soil pore size decreases by compaction (Saggar et al., 2013). However, these general rules are not always applicable, in part due to substantial microscale heterogeneity in soil biological activity and O2 availability, which determine the rate and pathway of microbial processes (Blagodatsky et al., 2011). For example, even in a relatively 88 dry soil, i.e., at 40-60% WFPS, physical constraints of O2 supply to the center of soil aggregates can lead to formation of anoxic microsites and localized, but sizeable, N2O production via denitrification (Schlüter et al., 2018). Likewise, plant residue fragments within a relatively dry soil matrix absorb water from surrounding soil and generate microscale moisture gradients, creating anoxic micro-environments favorable to denitrification (Kim et al., 2020; Kravchenko et al., 2017). Counterintuitively, the influence of such microsites can be enhanced under conditions that are generally considered unfavorable to N2O production via denitrification, i.e., in coarse textured soils, in soils with well-developed pore architecture dominated by macro-pores (> 30 µm Ø), and/or at low soil moisture contents (Kim et al., 2020; Kravchenko et al., 2017; Schlüter et al., 2018). Despite their small spatial extent, such microsites can be responsible for a majority of soil N2O emissions (Kravchenko et al., 2017; Parkin, 1987), limiting the ability to accurately predict and model N2O emissions (Butterbach-Bahl et al., 2013; Groffman et al., 2009). Plant roots play a major role in the formation of soil pores (Angers and Caron, 1998; Bardgett et al., 2014; Bengough, 2012; Jin et al., 2017). Climate change – elevated CO2 and changes in precipitation – can alter soil pore characteristics (Caplan et al., 2017; Hirmas et al., 2018; Wu et al., 2018) and also affects root growth (Hu et al., 2016), patterns in root development (Calleja-Cabrera et al., 2020), and root interactions with soil pore characteristics. The triple relationship between climate change, soil pore characteristics, and roots is complex due to interrelated feedback mechanisms. For instance, roots contribute to N2O emissions by root-derived modification of soil pore architecture and root-derived supply of N for N2O production, with decomposing plant roots often found to be a major N source (Hu et al., 2016). It is integral to assess joint contributions of plant roots and soil pore architecture to N2O emissions, to better understand the triple relationship and its potential feedback loop. 89 I posit that the interactions between decomposing plant roots and soil pores and their effect on N2O production and emission must be studied with roots decomposing at the same locations within the soil matrix where they grew and died. Living roots substantially modify the soil in their vicinity, i.e., rhizosphere, inducing changes in soil structure (Angers and Caron, 1998; Helliwell et al., 2019; Koebernick et al., 2017), hydraulic properties (Carminati et al., 2010), microbial community structure and abundance (Berendsen et al., 2012; Bird et al., 2011; Liang et al., 2016), and nutrient level/availability (Dijkstra et al., 2009; Marschner et al., 1987; Meier et al., 2017). Upon roots' death, the micro-environmental conditions within the rhizosphere (i.e., living root-soil interface and a thin layer of soil affected by living roots) are inherited by the detritusphere surrounding decomposing root residues (i.e., dead-root-soil interface and a thin layer influenced by root decomposition). These conditions can be substantially different from the environment of bulk soil (Kim et al., 2020; Kuzyakov and Razavi, 2019). When root residues are removed from the soil matrix and incorporated into sieved and mixed soil, either as intact layers or as a ground material mixed with the soil, the influence of the unique features of micro-environments in root- based detritusphere is no longer accounted for. This can result in misleading conclusions because soil characteristics in the detritusphere around an in-situ decomposing root are substantially different from the detritusphere around an artificially incorporated root sample. Using in-situ grown roots can enhance the understandings in belowground decomposition by considering the former rhizosphere's soil, i.e., rhizosphere legacy. To address the rhizosphere legacy effects, I worked with a model system of monoculture switchgrass (Panicum Virgatum L.), a prominent grass of North American prairie and a promising bioenergy feedstock (Gelfand et al., 2020). While N fertilizers are often the main source of N2O emissions in conventional agriculture (Tian et al., 2019), their role is less prominent in bioenergy 90 switchgrass systems, which typically receive lower N fertilizer inputs than cash crops, if any (Adler et al., 2007; Ruan et al., 2016). It is partly because switchgrass biomass yields only weakly respond to N availability (Roley et al., 2018; Wang et al., 2019). Rather, decomposing roots are expected to play a major role as N sources for N2O production in switchgrass systems (Adler et al., 2007; Cherubini and Jungmeier, 2010; Monti et al., 2012). Most N2O production in soil is driven by microorganisms (Braker and Conrad, 2011; Butterbach-Bahl et al., 2013; Syakila and Kroeze, 2011); and the overall microbial activity is markedly higher in the rhizosphere and detritusphere as compared to bulk soil, that is, the soil not directly affected by roots or detritus (Marschner et al., 2012). In intact soil with in-situ grown roots, identification of areas directly associated with rhizosphere or detritusphere and quantification of microbial activity dynamics within them are challenging. 2D zymography is a non-destructive technique used for measuring extracellular enzyme activity (Wallenstein and Burns, 2011), which enables the exploration of microbial hotspots at fine spatial scales with high resolutions (Heitkötter and Marschner, 2018; Spohn and Kuzyakov, 2014). Spatial and temporal dynamics of hydrolytic extracellular enzymes, e.g., N-acetylglucosaminidase, have been investigated in root systems and the rhizosphere of various plant species using zymography (Razavi et al., 2016; Sanaullah et al., 2016; Spohn and Kuzyakov, 2013; Spohn and Kuzyakov, 2014; Ma et al., 2017). N- acetylglucosaminidase (a.k.a., chitinase) is involved in both C and N cycling (Tabatabai et al., 2010). Its activity reflects the supply of decomposed organic substances required for N2O production and thus, is potentially associated with N2O emission. In zymography, visualization of chitinase activity is possible by incubating a membrane saturated with a chitinase-specific fluorogenic substrate on the soil surface (Guber et al., 2019). Then, the chitinase activity can be quantified using the fluorescence intensity of the product in the membrane. The ‘zymography’ 91 technique with chitinase-specific substrate can visualize microscale patterns in soil C and N processing and locate hotspots of microbial activity and subsequent N2O production. The N2O production from decomposing organic sources is closely related to the rate of decomposition, and the dynamics of decomposed C and N are closely coupled and affect associated soil-atmosphere exchange processes (Xia et al., 2018). C and N within plant residue fragments and in their immediate vicinity (Gaillard et al., 1999; Nicolardot et al., 2007; Vedere et al., 2020) can be directly used by microbes that produce N2O, turning residue fragments into hotspots of microbial activity and N2O production. Besides decomposed materials, microbial organisms use C and N derived from native SOM. This can cause a priming effect, that is short-term changes in SOM turnover, triggered by the input of added organic substances (Kuzyakov, 2010; 2000). Previously, priming has been measured in terms of C, defined as accelerated (positive priming) or decelerated (negative priming) emissions of CO2 derived from SOM. Recently, N2O priming, that is altered emissions of soil-derived N2O in response to additions of organic substances, has been suggested as an important contributor to soil N2O emission (Daly and Hernandez-Ramirez, 2020; Roman‐Perez and Hernandez‐Ramirez, 2020; Schleusner et al., 2018; Thilakarathna and Hernandez-Ramirez, 2021a; Thilakarathna and Hernandez‐Ramirez, 2021b). Simultaneous investigation of CO2 and N2O priming in the root detritusphere and their dependency on soil pore and moisture conditions can help to better understand changes in SOM stocks and subsequent greenhouse gas emission triggered by plant residues. The goal of this study was to assess the role of decaying plant roots in N2O production and emission from soils with contrasting pore architectures under contrasting soil moisture conditions and to explore the underlying biophysical mechanisms. I focused on how soil physical properties regulate the condition of detritusphere micro-environment, and consequently influcne the 92 microbial activity related to decomposition and N2O production. I used a rhizobox, which is a flat rectangular box containing soil and roots grew along the transparent front panel (Mašková and Klimeš, 2020). Soil rhizoboxes containing in-situ grown roots enabled accounting for the rhizosphere legacy contribution to N2O production and emission. Dual-isotope labeling (13C and 15 N) allowed for simultaneous tracking of plant originated C and N in mineral, organic, and microbial biomass forms in soil. Zymograhic measurements of chitinase activity on the intact surfaces of the rhizoboxes provided microscale maps indicative of microbial activity on the decomposing roots and in the bulk soil. More specifically, I (i) assessed N2O and CO2 emissions from the detritusphere by using 13 15 in-situ grown plant roots labeled with C and N, (ii) quantified the chitinase activity in detritusphere of decomposing roots and in the bulk soil and examine associations between its spatio-temporal dynamic and N2O emissions, (iii) investigated temporal changes in contributions of root- and soil-derived C and N to emitted N2O and CO2 and to microbial biomass during decomposition, and (iv) tested the effects of soil physical conditions (i.e., pore architecture and moisture) on N2O emission and enzyme activity in root-based detritusphere. 4.2 Materials and Methods 4.2.1 Rhizobox preparation Soil for the experiment was collected from monoculture switchgrass plots at the Biofuel Cropping System Experiment (BCSE) of the Long-Term Ecological Research site at W. K. Kellogg Biological Station, Michigan, USA. The soil is Kalamazoo loam (fine-loamy, mixed, mesic, Typic Hapludalf), developed on glacial outwash. Monoculture switchgrass has been grown 93 at the BCSE site since 2008. Composite soil samples for the study were collected from 5 – 10 cm depth and air-dried. Two soils with contrasting pore-size distributions were created for the study: a material with prevalence of > 30 µm Ø pores and a material with prevalence of < 10 µm Ø pores, referred to further as large-pore and small-pore soils, respectively. These pore sizes were selected based on previous studies where different priming effects, CO2 emissions (Toosi et al., 2017), water absorption by plant residues, and N2O emissions (Kravchenko et al., 2017; Kim et al., 2020) were found. Large-pore soil consisted of the sieved 1-2 mm soil fraction which was obtained by sieving air-dry soil through 1 mm and 2 mm mesh. Small-pore soil was prepared as follows: the 1-2 mm soil fraction was subjected to a series of gentle grindings using mortar and pestle, followed by sieving through a 0.053 mm sieve and added to the small-pore soil. The remaining small mineral particles were completely ground using a shatter box until they pass through the 0.053 mm sieve. This approach generated the soils with contrasting pore-size distributions, yet with the comparable mineralogy and microbial structure (Kim et al., 2020; Toosi et al., 2017). Previously conducted X- ray computed micro-tomography scanning of these materials at 2 µm resolution demonstrated that the large-pore and small-pore soils obtained using such procedure are dominated by > 30 µm Ø and < 10 µm Ø pores, respectively (Toosi et al., 2017). Since the goal of the study was to explore the role of decomposing roots in N2O emission, it was imperative to ensure that the N2O originated by decomposing root material could be tracked and distinguished from that originated from the soil organic matter. This was achieved by labeling the plants with 15N via direct fertilization of the roots, as opposed to adding 15N compounds to the soil. This approach required specially constructed rhizoboxes. Rhizoboxes used in this study were transparent plastic containers with dimensions of 4.7 cm x 2 cm x 5.3 cm. Each rhizobox had 8 94 small holes at the bottom to enable roots to grow out of the soil and into the 15N labeling solution below (Fig. 4.1). Each rhizobox also had a removable front panel for zymography measurements (described in 2.5 Zymography analysis). Rhizoboxes were covered with a UV-light screening tape prior to planting to protect the belowground biomass from other photoautotrophs. A total of 66 rhizoboxes were built - 33 of them were filled with large-pore soil and the other 33 with small- pore soil. Approximately 44 g of soil were packed within each rhizobox to achieve a bulk density of 1.15-1.17 g‧cm-3. Later, 30 rhizoboxes for each soil material (large and small) were used to grow switchgrass, and the remaining 6 boxes were kept as is but under the same conditions with switchgrass-containing rhizoboxes during the entire growing period. To enhance germination, switchgrass (var. Cave-In-Rock) seeds were stratified by shaking in 8M sulfuric acid for 5 minutes, followed by rinsing 3 times with distilled water. Residual water from the seeds was removed using paper tissues, and the seeds were spread in an even layer on Whatman #1 filter paper inside a petri dish. Another filter paper was placed on the top, and 5 mL of sterile 0.2% potassium nitrate solution was added. The entire petri dish was sealed with parafilm, covered with a paper bag, and placed at 4 °C in a refrigerator. Seeds that germinated within 3 - 5 days were used for the experiment. One switchgrass seed was placed in each rhizobox, such that half of the seed was buried into the soil. Planted rhizoboxes were placed on a rack tilted at 50⁰ to induce root growth on the removable front panel for future zymography measurements (Fig. 4.1a). Soil moisture was adjusted to ~60% WFPS at planting and maintained at the same level over the growth period by replenishing water. In order to induce the roots to grow out of the box and into the 15N labeling solution (Fig. 4.1b) the boxes were watered from the bottom and a constant soil moisture was maintained during growth by placing wet cloths at the bottom of rhizobox racks and keeping them 95 completely wet during the entire growing period. Switchgrass was grown for 8 weeks in the greenhouse at 24 ⁰C and 16 h of daylight and then subjected to dual labeling (described in 2.2 13C and 15N plant labeling). Over 8 weeks of growth, the switchgrass seedlings developed firm stalks reaching 11 – 30 cm in height; and had several roots coming out of the holes at the bottom of the rhizoboxes. (a) (b) 15N solu'on (c) (d) Figure 4.1: Switchgrass 15N and 13C labeling procedure. (a) Rhizoboxes placed on a rack tilted at 50⁰ to induce root growth on the surface of the box for subsequent zymography. (b) Rhizobox placed on the beaker containing 15N Hoagland solution to label the plant without directly adding 15N to the soil. (c) Rhizoboxes and beakers sealed with aluminum foil to avoid evaporation. (d) Rhizoboxes and beakers placed in the chambers for pulse labeling the plants with 13CO2. 13 4.2.2 C and 15N plant labeling After 8 weeks of growth, the switchgrass was subjected to 13C and 15N labeling. To apply 13 C labeling, 3 13CO2 pulses were provided with a 5-day interval in-between. In each pulse, 110 mg of NaH13CO3 (> 99.9% 13 C) was placed in a plastic chamber (76 L) along with 15-20 96 switchgrass-grown rhizoboxes. The chamber was sealed with air-impermeable silicon and 10 mL of 1 M H2SO4 was injected to complete the acid-carbonate reaction and 13CO2 production (Fig. 4.1d). The amount of NaH13CO3 used in the study was chosen to achieve ~40 atom% of 13CO2 in the labeling chamber headspace, which was higher than 33 atom% suggested in Bromand et al. (2001). For 15N labeling, I used 50% Hoagland solution (diluted with distilled water, ~100 mg N‧L- 1 ) containing N as 15NH415NO3 (98 atom %, Sigma-Aldrich, MO, USA). Each rhizobox was placed on top of a 50 mL glass beaker filled with labeling solution in such a way that the roots, emerging through the holes at the bottom of the rhizobox, were inserted into the liquid inside the beaker (Fig. 4.1b and c). The volume of the solution in the beaker was ~50 mL; the solution level was checked and refilled every day. There was > 5 mm gap between the liquid and the rhizobox, thus the solution from the beaker did not come into direct contact with the rhizobox's soil. Therefore, the only means by which the switchgrass plants could have received 15N was through the root uptake from the solution. This design ensured that the only source of 15N within the rhizoboxes was the 15 N from the plant roots. The rhizoboxes were kept on the beakers with the labeling solution for 2 weeks. After 2 weeks of labeling, i.e., a total of 10 weeks of switchgrass growth, the plants were terminated by cutting aboveground biomass. The rhizoboxes were air-dried at room temperature within a ventilation hood for 3 days to readjust the moisture level prior to incubation. 4.2.3 Incubation experiment and N2O, N2, and CO2 analyses There were 60 root-containing rhizoboxes (plant-terminated, air-dried, see section 2.2) prepared following procedures described in previous sections. Three replicates of such rhizoboxes for each soil pore size material (large-pore or small-pore, batch 0) were subjected to the same analyses as the incubated samples described later. Batch 0 was used to characterize the C and N 97 contents of the rhizobox soil at day 0 (no incubation). The remaining 54 rhizoboxes were subjected to a 39-day incubation under two moisture levels, i.e., 40% WPFS and 70% WFPS. These rhizoboxes containing dead in-situ grown roots were randomly grouped into 3 batches, where each batch consisted of a total of 18 rhizoboxes, 9 from each soil fraction. In each batch, 5 replicated rhizoboxes from each soil were randomly assigned to 40% WFPS and the remaining 4 rhizoboxes were assigned to 70% WFPS. Water needed to reach the designed WFPS was added to each box; and the boxes were kept in 450 mL Mason jars. A small beaker with 10 mL of distilled water was placed within the jar to maintain a high humidity and reduce moisture losses. Jars were tightly sealed and kept in the dark at 21 °C. Batch 1 rhizoboxes were took apart and used for plant and soil analyses on day 3, Batch 2 rhizoboxes were took apart on day 21, and Batch 3 rhizoboxes were used for zymography analysis and took apart on day 39. There were additional 6 rhizoboxes that did not contain plant roots (3 large-pore, and 3 small-pore soils). These rhizoboxes were packed with the same soil materials but switchgrass was not planted in them and kept in the same environment with the rhizoboxes in which switchgrass was grown (placed in the greenhouse and moisture maintained as 60% WFPS). These no-plant controls were incubated with Batch 2. For these control samples, moisture was adjusted to 40%, 55%, and 70% WFPS and the average values of the measurements from the 3 samples was reported. The gas emissions and destructive analysis results were thus averaged across the moisture levels. During the incubation, headspace gas from each jar (Batch 1-3) was collected on days 1, 3, 6, 13, 21, and 39. For gas sampling, approximately 50 mL of headspace gas was pulled out from each jar using a 50 mL Luer-Lock syringe, and injected into 3 separate vials for i) 15N2O; ii) 13CO2 , 5 N2 and 15N2; and iii) CO2 and N2O concentration analyses, respectively. For 15N2O and 13CO2 /15N2 analysis, the headspace gas was injected into 20 mL glass vials at 1 atm pressure. For concentration 98 analysis, the headspace gas was injected into 5.9 mL storage vials (Labco Ltd, Lampeter, U.K.) at 2 atm pressure. The Mason jars were flushed using ultra-pure gas (79% N2, 21% O2, and 0% CO2/N2O) after each sampling. N2O concentrations were analyzed using a gas chromatograph (Agilent Technologies 7890A, Santa Clara, CA, USA) and CO2 concentrations were analyzed with a Licor infrared gas analyzer. Isotopic composition of N2O was measured in stable isotope facility at Michigan State University on an IsoPrime 100 stable isotope ratio mass spectrometer (IRMS) interfaced to a Tracegas inlet system (Elementar, Mt. Laurel, NJ) (Sutka et al., 2003). Using Helium as the carrier gas, the Tracegas inlet removes CO2 and water, before cryofocussing N2O onto a gas chromatographic column, before introduction to the IRMS. The calibration procedure included analyses of isotopically distinct standards of enriched N2O (15N2O, > 98 atom%, Cambridge Isotope Laboratories, Inc. Andover, MA, USA). Each day, at least four of the calibration standards were analyzed covering a range of concentration from 3 to 20 nmole N2O. The minimum detection limit of Isoprime 100 is around 1.5 nmole. CO2 gas samples were measured using the same procedure, except the Tracegas system was set so as not to remove CO2 from the sample. The standards were prepared by mixing laboratory standard gas with different amounts of enriched CO2 (13CO2, 99 atom %, Sigma-Aldrich, MO, USA). N2 concentrations in gas samples were analyzed using a gas chromatograph (Hewlett Packard 5890) with a modified inlet interfaced to an Isoprime IRMS (Elementar) (Roberts et al., 2000). 4.2.4 Zymography analysis A hydrophilic polyamide membrane filter (Tao Yuan, China) 4 cm x 5 cm in size was used for zymography measurements. I used 6 mM 4-Methylumbelliferyl N-acetyl-glucosaminide (MUF-NAG; Sigma-Aldrich, MO, USA) solution as a substrate for chitinase (Spohn and 99 Kuzyakov, 2014). Upon contact with chitinase, the substrate releases florescent product (4- methylumbelliferone) that can be detected under ultraviolet light (Guber et al., 2018). I used Canon EOS Rebel T6 camera with a Canon 75-300 mm f/4-5.6 III Telephoto Zoom Lens to take zymograms (Guber et al., 2021). Batch 3 rhizoboxes were repeatedly measured on days 1, 3, 6, 13, and 21 after the gas collection. The soil surface for zymography of each rhizobox was sprayed with distilled water to bring the soil surface to saturation. The amount of sprayed water was < 5 mg, thus did not significantly increase the soil moisture level. Then the membrane was soaked with MUF-NAG, attached to the soil surface and fixed with a holding frame. The sample was placed in the dark hood with the camera installed at the top and with ultraviolet light source. The membrane was photographed every minute for a total of 50 minutes per sample. The zymograms were calibrated and converted into maps of enzyme activity using MATLAB 9.5 (MathWorks, MA, USA). A calibration was performed by adopting the method developed by Guber et al. (2019). I calculated root and soil enzyme activity by separating the decomposing root and soil areas on each zymogram. For that, a picture of the soil surface, i.e., a reference image, was taken prior to zymography (Fig. 4.2a). Then a series of image processing steps – background removal, automatic adjustment of brightness and contrast, Gaussian blur (pixel 2.0), and auto local thresholding (default option) in Fiji (Schindelin et al., 2012) – was applied to the reference images (Fig. 4.2). The resultant binary image was then subjected to particle analyzer imbedded in BoneJ plugin (Doube et al., 2010) to exclusively select root particles. The final reference image provided information on the location of the roots on the rhizobox surfaces (Fig. 4.2f and Fig 4.3c). The areas of the decomposing root and the soil were then multiplied by the average enzyme activity values calculated from the zymograms to obtain separated enzyme activity values for the decomposing roots and for the soil (Fig. 4.3d and e). It should be noted that 100 recent efforts in in-situ rhizosphere mapping using zymography tools demonstrated that the spatial extent of microbial hotspots around roots does not exceed 50-250 µm (Khosrozadeh et al., 2021 in review). In the current study, a portion of these very narrow active layers of soil surrounding decomposing roots was likely classified as root surfaces. Thus, what I report as chitinase activity on the decomposing roots likely includes a portion of the detritusphere that was former rhizosphere in immediate proximity (< 500 µm) to the roots. Figure 4.2: Processes to obtain the reference image of roots. (a) Raw image taken prior to zymography measurement. (b) Raw image cut into region of interest. (c) Image after background removal, default adjustment of brightness and contrast, and Gaussian blur. (d) Image after default threshold applied. (e) Result of particle analyzer that separates all particles detected in the image. (f) Final reference image obtained by selecting root particles. Chitinase activity during root decomposition was analyzed and reported as i) total chitinase activity (Fig. 4.3b), which is the mean chitinase activity on the entire zymography surface, ii) root chitinase activity, which is the mean chitinase activity on the decomposing root surfaces only (Fig. 4.3d), and iii) soil chitinase activity, which is the mean chitinase activity on the soil surface (i.e., 101 bulk soil) (Fig. 4.3e). That is, mean chitinase activity in area of interest (root or soil, presented as brown color in Fig. 4.3d and e) was used for decomposing root and soil chitinase activities. Note that the root chitinase activity is not the activity of the living roots but that of the decomposing roots. 10 (a) (b) pmol‧cm-2‧min-1 0 (c) (d) (e) Area of interest Figure 4.3: Illustration of zymography procedures. (a) Example of a raw image that captures the florescence developed from enzyme-substrate reaction. The white dashed box in (a) represents the region of interest used for further zymography analysis. (b) A map of enzyme activity calculated from the set of raw images taken every 5 minutes. (c) A final reference image in which black and white indicates the root and soil, respectively. (d) Chitinase activity on the surface of decomposing roots. (e) Chitinase activity on the surface of soil. In images (d) and (e), brown color represents the area of interest, and no enzyme activity was presented transparently. 4.2.5 Plant and soil analysis Rhizoboxes of batch 1, 2, and 3 were subjected to destructive analysis on days 3, 21, and 39 of incubation, respectively. Roots and soil in the rhizoboxes were separated by using forceps. For dissolved organic C (DOC), total dissolved nitrogen (TDN), microbial biomass, ammonium 102 (NH4+) and nitrate (NO3-), 10 g of fresh, wet soil was weighed immediately after collection and added to a vial with 50 mL of 2 M KCl solution (1:5 soil: solution ratio). The resultant soil suspension was homogenized using an orbital shaker set to 180 rpm for 24 h and centrifuged with 5000 rpm for 10 min. The supernatant was filtered through a 0.45 μm membrane. Regular (i.e., not isotope-specific) DOC and TDN were analyzed using Shimadzu TOC-Vcph C analyzer with a total nitrogen module (Shimadzu, Tokyo, Japan). Inorganic N (NH4+ and NO3-) were determined spectrophotometrically at 630 and 530 nm, using salicylate-cyanurate method (Sinsabaugh et al., 2000) and vanadium method (Doane and Horwáth, 2003), respectively. Dissolved organic N (DON) was calculated by subtracting inorganic N from TDN. Microbial biomass was measured using a modified version of the fumigation-extraction method (Vance et al., 1987). Fumigated and non- fumigated KCl (2M) soil extracts were analyzed using Shimadzu TOC-Vcph C analyzer, and the difference between fumigated and non-fumigated samples was divided by Kec (0.45) and Ken (0.54) to obtain microbial biomass C and N , respectively (Brookes et al., 1985; Vance et al., 1987). However, most estimates of microbial biomass N had considerable uncertainty due to instrument failure, thus only microbial biomass C (MBC) was reported. Remaining soil and roots were air-dried for 2 days and homogenized to measure total C and N contents using Costech elemental combustion system (Costech Analytical Technologies Inc., CA, USA). Additionally, in the rhizoboxes from batch 3 I also collected and analyzed total C and N of soil particles directly attached to the roots, i.e., rhizosphere soil, separately. Measurements of δ13C and δ15N in solid samples, i.e., soil and plant tissues, were performed in stable isotope facility at Michigan State University using an Isoprime Vision IRMS interfaced to a Vario Isotope Cube elemental analyzer (Elementar). Isotopic values of KCl extracts were determined by freeze-drying 103 the solution and analyzing the isotopic values of the power using EA-IRMS in Environmental Molecular Sciences Laboratory (EMSL) at Pacific Northwest National Laboratory. 4.2.6 Calculation methods Relative differences in isotopic composition to the standards (δ) were converted to atom% to calculate switchgrass root-derived C and N in all pools (gas, microbial biomass, dissolved organic C, and total dissolved N). Atom% = {[(δ‧1000-1+1) ‧RR]-1 +1}-1 ‧ 100 Eq. 1 where δ (‰) is relative difference in isotope ratios and RR is the ratio of the heavy isotopes to the light isotopes (13C/12C and 14N/15N). The fraction from the labeled residue (fres) in certain pool is calculated as: fres = (Atom%trt - Atom%control) ‧ (Atom%root - Atom%control) Eq. 2 where Atom%trt is atom% (atom%13C or 15N) of the treatments in the certain pool, Atom%control is the atom% of unamended sample of corresponding pools (13C or 15 N natural abundance), and Atom%root is the atom% of the original labeled roots. Then the mass or concentration of the pool derived from the labeled residue (Cres) was calculated as: Cres = Ctotal * fres Eq. 3 Where Ctotal is the total mass or concentration of the pool. The pool derived from soil (Csoil) was calculated as: Csoil = Ctotal * (1 - fres) Eq. 4 104 Delta (δ) value of microbial biomass (δΜΒ) was calculated from the δ values of fumigated and non- fumigated extracts as: δΜΒ = [δF ‧ CF - δNF ‧ CNF]‧ CMB -1 Eq. 5 Where δF and δNF are δ values of fumigated and non-fumigated samples; CF and CNF are concentrations (C or N) of fumigated and non-fumigated samples; and CMB is a difference between CF and CNF. The rates of CO2 and N2O priming of decomposing roots were calculated by subtracting the average CO2 and N2O emission rate from unplanted control soils from each observation of soil- derived CO2 and N2O emissions of the treatments. Cumulative CO2 and N2O priming effects, which are defined as the differences in cumulative soil-derived N2O and CO2 emissions between treatments and no-plant controls, were calculated as the sum of the product of the priming rates and the time intervals from day 1 to day 21. Cumulative priming was determined by each rhizobox, thus the rhizoboxes belonged to batch 1 (took apart for destructive analysis on day 3) were not included for calculation. 4.2.7 Statistical analysis Data were analyzed using a mixed-model approach implemented in PROC MIXED procedure of SAS 9.4 (SAS Institute Inc., NC, USA). For gas flux data, including total, root- derived, soil-derived and fractions of CO2‧N2O, total N2, and rates of CO2‧N2O priming; and for zymography data, including total, root, and soil chitinase activity, the statistical model consisted of fixed effects of soil (large- vs. small-pore dominated soils), soil moisture (40 vs. 70 %WFPS), time, and their interactions. The statistical model also included the random effect of individual rhizoboxes nested within materials and moistures. Time was treated as a repeated measure factor, 105 with individual rhizoboxes used as a subject of repeated measurements. The repeated measures analysis was conducted using the approach outlined in Milliken and Johnson (2002). For cumulative gas emissions, including cumulative CO2 and N2O, and cumulative priming effects, and for total C and N derived from roots the statistical model consisted of fixed effects of material, moisture, and their interactions. For measurements from destructive analysis, including total C, total organic N, MBC, and inorganic N, the statistical model consisted of fixed effects of material, moisture, time, and their interactions. The relationships between pairs of continuous variables (e.g., between CO2 and N2O or between chitinase activity and emitted gases) were explored by fitting statistical models that included the mentioned above categorical factors, the linear effects of the studied continuous variable, and the interactions between them. The latter enabled assessing the differences in the relationships between the continuous variables at different levels of the soil pore or soil moisture effects. When the regression slopes were not significantly different between the soil pore sizes (p> 0.05), results of fitting a model with a common regression slope were reported. For all datasets, the normality assumption was evaluated by checking normal probability plots, and when violated, the data were natural log-transformed (Milliken and Johnson, 2009). The homogeneity of variances assumption was tested by Levene’s test based on the absolute values of model residuals. When the homogeneity of variances assumption was found violated, the analysis with heterogeneous variances was conducted (Milliken and Johnson, 2002). Statistical significances were indicated as *** (p< 0.01) and ** (p< 0.05), with tendency as * (p< 0.10). 106 4.3 Results 4.3.1 Switchgrass growth and its effects on soil characteristics At plant termination after 10 weeks of growth, the root biomass (av. 426 mg per rhizobox) and the aboveground biomass (av. 130 mg per rhizobox) were similar between the two soils with small and large pores (p> 0.10, data not shown). Total C, N, atom%13C, and atom%15N of the roots did not differ between the soils (p> 0.10, Table 4.1). The area occupied by roots on the side of the rhizobox subjected to zymography constituted on average 7.5% of the soil surface and were similar in large- and small-pore soils (p> 0.10, data not shown). Immediately after plant termination, the soil of contrasting pore sizes did not differ in terms of total C and organic N, inorganic N, and isotopic signatures of C and N (p> 0.10) (Table 4.1). The large-pore soil had lower MBC (384 mg C kg-1 soil) than the small-pore soil (571 mg C kg-1 soil) (p< 0.05, Table 4.1). Table 4.1: C and N characteristics of soil and roots after plant growth and labeling. Standard deviations in parenthesis. Letters indicate significant differences between large- and small-pore soils (p< 0.05). Switchgrass roots Large-pore Small-pore Total C (mg C kg-1 dry matter) 446 (26.5) 444 (24.7) Total N (mg N kg-1 dry matter) 29.7 (10.3) 28.8 (9.7) Atom%13C 1.61 (0.31) 1.63 (0.29) Atom%15N 8.83 (1.04) 9.10 (0.74) Soil after plant growth Large-pore Small-pore -1 Total C (mg C kg soil) 11.1 (0.4) 11.2 (0.02) Total N (mg N kg-1 soil) 1.16 (0.03) 1.17 (0.03) + -1 NH4 (mg N kg soil) 5.36 (1.08) 3.68 (1.46) NO3- (mg N kg -1 soil) 0.23 (0.10) 0.43 (0.25) Atom%13C 1.09 (0.00) 1.09 (0.00) 15 Atom% N 0.43 (0.02) 0.42 (0.03) Microbial Biomass C (mg C kg-1 soil) 384a (45) 571b (44) 107 4.3.2 N2O, CO2, and N2 emissions Total, root-derived, and soil-derived N2O emissions were greater in large-pore soil compared to small-pore soil, and greater at 70% WFPS compared to 40% WFPS throughout the incubation (Fig. 4.4a and b, Table 4.2). Soil pore size strongly regulated total N2O emissions (p< 0.01). After 21 days of incubation the large-pore soil had 11 and 28 times higher cumulative total N2O emissions compared to small-pore soil at 40% and 70% WFPS, respectively (Fig. 4a and b). Contrarily, N2O emission from the control rhizoboxes without roots was not statistically different in the two soil materials. Soil moisture of 70% WFPS generated 3.5 and 1.4 times higher total N2O emissions compared to 40% WFPS in large-pore soil and small-pore soil, respectively. Table 4.2: Summary of statistical analysis for gas emissions. Effect Description Pore Soil materials of different pore size distribution (Large vs. Small) Moisture Treatments of moisture contents (40% WFPS vs. 70% WFPS) Location Roots (rhizosphere) vs. bulk soils Time Day of incubation (day 0, 3, 13, 21) * Interaction between effects Total Total Fraction Fraction Soil- N2O CO2 Root- N2O CO2 of N2O of CO2 derived priming Priming derived from from N2O N2O roots roots Effect Pore <.0001 0.0783 <.0001 0.3876 <.0001 <.0001 <.0001 0.0007 Moisture <.0001 0.2306 0.1403 0.1605 <.0001 0.0002 <.0001 0.6810 Pore*moisture 0.0030 0.5296 0.9174 0.8156 0.0289 0.0027 <.0001 0.0528 Time <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 Pore*Time <.0001 0.0959 0.1188 0.1587 0.0019 <.0001 0.0054 <.0001 Moisture*Time 0.2883 0.0540 0.3330 0.2592 0.1921 0.2811 0.2512 0.3594 Pore*Moisture*Time 0.1677 0.2562 0.4311 0.0169 0.5670 0.7014 0.2344 <.0001 108 The fraction of N2O derived from decomposing roots also differed between the two soils. While in the large-pore soil the root-derived N2O accounted for > 60% of the total emitted N2O for 3-21 days of incubation, in the small-pore soil it peaked at 60% on day 3 and continuously decreased to < 40% on day 21 (Fig. 4.4c). Although the higher moisture level increased the total amount of emitted root-derived N2O, it did not affect the fraction of the emitted N2O that was derived from the root residues. In contrast to N2O emissions, CO2 emissions were weakly controlled by pore sizes and not regulated by moisture levels (Fig. 4.5). In all treatments, the maximum CO2 emission rate was reached on days 3-6 and then decreased dramatically. The large-pore soil tended to have greater CO2 emissions compared to the small-pore soil (p< 0.10) but only at a later stage of the incubation (Fig. 4.5a). Unlike N2O, CO2 emissions were not affected by soil moisture (p> 0.10). Cumulative CO2 emissions from planted rhizoboxes were ~2 times greater than that from unplanted control soils (Fig. 4.6c and d). 109 (a) (b) (c) Total N2O emission (a, b) Pore *** Moisture *** Time *** Pore x Moisture *** Pore x Time *** Moisture x Time *** Pore x Moisture x Time * Frac%on from roots (c) Pore *** Time *** Pore x Time * Figure 4.4: Dynamics of N2O emission during the incubation: (a) total N2O emission rate at 40% WFPS and (b) 70% WFPS; and (c) the fraction of root-derived N2O across both WFPS treatments. Error bars are standard errors of the mean. ‘Large’ and ‘Small’ in the legends indicate the soils dominated by large (> 30 µm Ø) and small (< 10 µm Ø) pores, respectively. Controls are the rhizoboxes that were not planted and did not contain switchgrass roots. The gray box presents significant results of the factors and their interactions. Red asterisks * and *** mark statistical significance (p< 0.10 and 0.01). P-values of the factors and their interactions were presented in Table 4.2 (supplementary material). 110 (a) (b) * *** (c) Total CO 2 emission (a, b) Pore * Time *** Pore x Time * Moisture x Time * Frac%on from roots (c) Time *** Pore x Moisture x Time ** Figure 4.5: Temporal dynamics of CO2 emission and its fraction derived from the decomposing roots. Total CO2 emission rate (a) at 40% WFPS and (b) at 70% WFPS, and (c) fraction of root-derived CO2. Error bars are standard errors of the mean. Gray boxes present significant results of the factors and their interactions. Red asterisks *, ** and *** mark statistically significant differences between large- and small-pore soils (p< 0.10, 0.05, and 0.01). 111 Large-pore Small-pore (a) (b) c a a *** a NS a (Posi,ve) b b Priming Effect *** ** b S S S S trol WFP WFP trol WFP WFP Con 4 0 % 7 0 % Con 40% 70% (c) (d) b a ab a ** b a (Nega,ve) NS a a NS Priming Effect NS S S S S trol WFP WFP trol WFP WFP Con 4 0 % 7 0 % Con 40% 70% Figure 4.6: Cumulative N2O emissions from the (a) large- and (b) small-pore soil; and cumulative CO2 emission from the (c) large- and (d) small-pore soil after 21 days of incubating the rhizoboxes with in- grown switchgrass roots. Control refers to unplanted soil boxes incubated under each WFPS. Differences between soil-derived gas emission and the control gas emission (priming effect) are presented as red arrows. Red asterisks ** and *** mark the cases where the priming effect was significantly different from 0 (p< 0.05 and 0.01). NS stands for ‘Not Significant’. Letters indicate the significant differences between root- derived cumulative gas emissions (white, p< 0.05) and priming effect (red, p< 0.10) at a given moisture level. Error bars are standard errors of the mean. Note the different y-axis scales between (a) and (b). 112 Cumulative N2O priming was governed only by pore sizes (p< 0.01, marked as red arrows in Fig. 4.6a and b). However, the rate of N2O priming over 21 days was influenced by both pore size (p< 0.01) and moisture level (p< 0.05), and the moisture effect was more noticeable in large- pore soils (Fig. 4.7). The cumulative N2O priming was positive in all treatments except for small pore soil at 40% WFPS, and 20-25 times greater in large pore soils than in the small pore soils (Fig. 4.6a and b). Cumulative CO2 priming (marked as red arrows in Fig. 4.6c and d) tended to be different between pore sizes (p< 0.10) and moisture levels (p< 0.10), and was not statistically different from 0 or even negative depending on the treatment (Fig 4.6c and d). (a) (b) Pore *** Moisture ** Pore*Moisture ** a *** Time ** *** ** ** *** a b b b b a a a a a a a a a a a a a a Figure 4.7: Dynamics of the N2O priming effect at (a) 40% WFPS and (b) 70% WFPS in the large- and small-pore dominated soils. Red asterisks ** and *** indicate significant differences between large- and small-pores (p< 0.05 and 0.01). Letters indicate the significant differences between moisture levels at given soil materials and time (p< 0.05). CO2 emissions were positively correlated with N2O emissions, however, the strength of the correlation depended on the soil. Only 7% of variation in N2O emissions was explained by the 113 variation in CO2 emissions in large-pore soil, while 68% of the variation in N2O emissions was explained by the variation in CO2 emissions in small-pore soil (Fig. 4.8). N2 emissions from root- containing rhizoboxes were not significantly different from the control rhizoboxes without roots. The difference in N2 emissions between planted and unplanted soils was greater than 0 on day 6 in the large-pore soil (Fig. 4.9). R2=0.07 *** Controls R2=0.68 *** Figure 4.8: Relationship between total CO2 and total N2O emission rates in the two studied soils and moisture levels. Regression line of small-pore dominated soils was presented in green. Blue plots present the average emission rates of the unplanted controls. R2 was reported separately for large-pore and small- pore soils (p< 0.01). Regression slope from the large pores was not presented due to low R2. 114 (a) Large-pore (b) Small-pore ** ** Figure 4.9: Dynamics of N2 emission during the root decomposition in (a) large-pore soils and (b) small- pore soils. Shown are the differences between the planted soil and soil without plants (control). Asterisks ** mark the cases where the difference between the treatment and the control were greater than zero (p< 0.05). Error bars are standard errors of the mean. Gray lines show the actual N2 emission (average of the two moisture levels) from planted soils. 4.3.3 Chitinase activity dynamics during root decomposition Chitinase activities on root surfaces and in bulk soil were consistently higher in the large- pore soil compared to the small-pore soil, and at 70% WFPS compared to 40% WFPS (p< 0.05, Fig. 4.10). Chitinase activity of the decomposing root was significantly higher than soil chitinase activity regardless of the soil and moisture level (p< 0.01, Table 4.3). There were distinctive temporal trends of chitinase activity depending on the spatial location. While root chitinase activity increased drastically and reached a maximum on day 3-6 (Fig. 4.10b and e), the chitinase activity in bulk soil was the highest during 6-13 days of the incubation (Fig. 4.10c and f). At 40% WFPS, large-pore soil and small-pore soil had maximum root enzyme activity of 1.2 and 0.4 pmol‧min- 115 1 ‧cm-2, respectively. At 70% WFPS, activity was higher than at 40% WFPS, with 2.4 pmol‧min- 1 ‧cm-2 in large-pore soil and 1.1 pmol‧min-1‧cm-2 in small-pore soil. Root Bulk soil 10 (a) (b) (c) 40% WFPS *** ** c b * ** *** ** *** b b b ab b b a a pmol‧cm-2 ‧min-1 (d) (e) b (f) b b ** * *** b 70% WFPS b b b *** b b ab b a a a ab b ab ab a a 0 Figure 4.10: Dynamics of chitinase activity on root and soil surfaces in the large- and small-pore dominated soils at the two studied soil moisture levels. Example zymograms are shown in (a) and (d); root chitinase activity in (b) and (e), and soil chitinase activity in (c) and (f) for 40% and 70% WFPS, respectively. Asterisks *, **, and *** indicate the significant differences between the two soils at a given day of incubation and moisture content, at significance levels of 0.1, 0.05, and 0.01, respectively. The letters mark differences between time (incubation day) within given soil and moisture. Error bars are standard errors of the mean. 116 Table 4.3: Summary of statistical analysis for chitinase activity. Effect p-value Pore 0.0022 Moisture 0.013 Location 0.0001 Time <.0001 Pore*Time <.0001 Moisture*Time 0.0094 Location*Time 0.0006 Both root-derived N2O and CO2 emission rates were positively correlated with root chitinase activity (Fig. 4.11a and b). While the root chitinase activity explained 22% of the variability in root-derived N2O emission rate (p< 0.01), it explained only 3% of the variability in root-derived CO2 emissions (p< 0.05). Neither soil-derived N2O nor CO2 emission correlated with soil chitinase activity (Fig. 4.11c and d). The regression slopes were not significantly different between the pore sizes and moisture levels. 117 (a) (b) (c) (d) Figure 4.11: Relationships between chitinase activity and emissions of root-derived N2O or CO2. (a) Correlation between root chitinase activity and root-derived N2O emission rate, and (b) root- derived CO2 emission rate. (c) Correlation between soil chitinase activity and soil-derived N2O emission rate, and (d) soil-derived CO2 emission rate. Black dash lines represent linear regression models. N.S indicates that the regression slope was not significantly different from zero. 4.3.4 Soil inorganic N, plant-derived C and N, and microbial biomass C Although there were no differences in NH4+ and NO3- contents between the large- and small-pore soils after plant termination before incubation, the large-pore soil tended to have higher inorganic N content once the root decomposition started. In the large-pore soil, NH4+ was significantly higher than that in the small-pore soil on day 3 (10.6 vs. 5.9 mg N‧kg-1 soil) and on day 39 (1.7 vs. 1.1 mg N‧kg-1 soil) (Fig. 4.12a). Likewise, NO3- was significantly higher in the 118 large-pore soil compared to small-pore soil on day 21 (2.2 vs. 1.3 mg N‧kg-1 soil) (Fig. 4.12b). Ammonium rapidly increased at the early stage of the root decomposition (~day 3) and decreased afterward. Nitrate increased continuously throughout the decomposition, reaching 20.4 mg N‧kg-1 soil on day 39. No effect of moisture level was detected on NH4+ or NO3- contents. BC (a)NH (b) NH4+4+ (b) NO33- - (c) NO *** (a) MBC (b) NH 4+ (c) NO 3- d aa *** d c d b ** c ** d c b c c ** c ** c b b c c **b a c b d c ** a c **a a b c d c a ** a a b c b a a c b a Figure 4.12: Dynamics of soil (a) NH4+ and (b) NO3- during rhizobox incubation. Asterisks ** and *** mark significant differences between large- and small-pore dominated soils at the given day of incubation at p of 0.05 and 0.01, respectively. Letters mark the differences between days of incubation in the given soil (p< 0.05). Error bars are standard errors of the mean. 119 While there were no differences in DOC between the two soils on day 0, the effect of pore size on DOC content became larger as the incubation proceeded (Fig. 4.13a). The difference in DOC between the pore sizes was the greatest on day 39, with 98 vs. 52 mg C‧kg-1 soil for large- and small-pore soils, respectively. Moisture level did not affect the DOC. DON increased quickly and reached 74 and 60 mg N‧kg-1 soil within 3 days for the large- and small-pore soils, respectively (Fig. 4.13b). It further increased until day 21 and decreased back to the same level that was present on day 3, which was still higher than the initial level. Consistent with the DOC, DON was not affected by moisture level, but was greater in large-pore soils than in small-pore soils (p< 0.01). MBC on day 39 was 100-140 mg C‧kg-1, and 15 % (small-pore) ~25% (large-pore) was derived from the root C (Fig. 4.14). While total MBC was not affected by the pore size and moisture level, the root-derived MBC was higher in the large-pore soils than in small-pore soils at the end of the incubation (day 39) (p< 0.05). (a) DOC (b) DON *** ** Pore *** Time * Pore *** Pore x Time ** Time *** Figure 4.13: Dynamics of (a) dissolved organic C (DOC) and (b) dissolved organic N (DON) during the rhizobox incubation. Asterisks *, **, and *** indicate the significant differences between large-pore and small-pore soil (on the figure) and the significant effect of the factor at p levels of 0.10, 0.05, and 0.01. 120 (a) MBC (b) DOC 40% WFPS 70% WFPS 40% WFPS 70% WF Pore *** Pore ** Pore *** ge all ge all ge all ge Lar Sm Lar Sm Lar Sm Lar Figure 4.14: Microbial biomass C (MBC) derived from soil organic matter (gray) and labeled decomposing root residues (blue) on day 39 of the incubation. Asterisk ** indicates significant effect of the root-derived MBC at 0.05 error probability level. Error bars are standard errors of the mean. Rhizosphere soil, that is, the soil particles directly attached to the in-situ grown and decomposing roots, had higher C and N content and fraction derived from the roots compared to the bulk soil at the end of the incubation (p< 0.01, Fig. 4.15). Total soil C and N content was greater in the large-pore soils compared to small-pore soils, but the fraction originated from the roots was not affected by pore sizes. Moisture level did not affect total soil C and N in the rhizosphere, nor the fraction of C and N from decomposing roots (Fig. 4.15c and d). 121 (a) (b) Loca'on *** Loca'on *** Pore ** Pore *** (c) Loca'on *** (d) Loca'on *** Moisture x Loca'on ** Figure 4.15: (a) Total soil C and (b) total soil N, and fractions of (c) root-derived total C and (d) root- derived total N in the root-associated soil and in the bulk soil on day 39 of the incubation. Symbol ** and *** indicates significant difference of the factor at 0.05 and 0.01 error probability level. Error bars are standard errors of the mean. 122 4.4 Discussion 4.4.1 In-situ grown roots in root decomposition studies The current shortage of decomposition studies that use in-situ grown roots is likely driven by several experimental limitations: 1) changes of the root mass during decomposition are not trackable, since it is not possible to measure the initial mass of the in-situ roots; 2) locations of the root residues during decomposition are not known, leading to difficulties in locating hotspots of microbial activity and N2O production; and 3) tracking the fate of root-originated N, for example 15 with a commonly used approach of N stable isotope labeling, can be a challenge because of difficulties in providing sufficiently high level of label within the roots without contaminating (15N label by fertilization) the soil. Kim et al. (2020) overcame the first two limitations by employing X-ray computed microtomography before and after incubating the experimental microcosms, which allowed measurements of initial root volumes, determination of root volume losses during decomposition, and visualization of root positions within the samples. The current study overcame the third limitation by employing an innovative system of supplying plant roots with N without soil contamination and ensuring that roots are the only source of 15N in the soil. 4.4.2 Soil pore size as a key driver of N2O emissions Consistent with expectations and previous studies (Schaufler et al., 2010; Shelton et al., 2000), during in-situ root decomposition, greater total N2O emissions and both greater root- and soil-derived components were observed at 70% WFPS than at 40% WFPS (Fig. 4.4, 4.5, and Table 4.2). High soil moisture content leads to lower O2 availability and faster NO3- diffusion, promoting N2O production via denitrification (Bollmann and Conrad, 1998; Butterbach-Bahl et al., 2013; Chen et al., 2013; Myrold and Tiedje, 1985). However, my results suggest that prevailing soil pore 123 architecture can play a much more important role in stimulating N2O production and transport to the soil surface than soil moisture (Fig. 4.16). The cumulative amount of N2O emitted from the large-pore soil was ~21 times greater than that emitted from small-pore soil, while the cumulative amount emitted at 70% WFPS, i.e., the WFPS of commonly observed maximal N2O emissions (Davidson, 1991; Schmidt et al., 2000) – was only 3.5 times greater than that at 40% WFPS. That is, the effect of soil pore architecture was more than 6 times greater than the effect of soil moisture. Figure 4.16: Graphical representation of the micro-environments in the rhizoboxes dominated with large- pore soil (left) and small-pore soil (right). Large-pore dominated soil has higher fungal growth, enzyme activity, and microbial activity near roots due to rhizosphere legacy. It leads to faster decay of roots and SOM in large-pore dominated soil, where larger amount of both root-driven and soil-driven organic substances can be used as substrates for N2O emission. Sponge effect (water absorption by decaying roots) is more prominent in large-pore dominated soil because of its lower water retention. Anaerobic detritusphere (near roots) and aerobic surrounding large pores together make optimum condition for denitrification and increase both root-driven and soil-driven N2O emission. At first glance, greater emissions of N2O from soils dominated by >30 µm Ø (large) pores than from soils dominated by < 10 µm Ø (small) pores seem paradoxical – if the overall anaerobicity were the controlling factor for N2O production, soils dominated by small pores with their greater pore tortuosity and resultant O2 shortages would produce more N2O (Groffman and Tiedje, 1989; Zaman et al., 2012). Yet, higher N2O emissions from large-pore than from small- 124 pore dominated soils have been reported before (Gu et al., 2013; Kaiser and Heinemeyer, 1996; Kravchenko et al., 2017; Weitz et al., 2001). It has been proposed that the reason for lower N2O emissions is a greater extent of complete denitrification to N2 in small-pore dominated soils, which is driven by their low O2 levels and slow rates of gas diffusion preventing an escape of produced N2O (Gu et al., 2013; Kaiser and Heinemeyer, 1996; Kravchenko et al., 2017; Weitz et al., 2001). Surprisingly, my findings challenge this explanation. The total N2 emission from the small-pore soil with decomposing plant roots did not exceed that of the no-plant control soil (Fig. 4.9b). On the contrary, in the large-pore soil, a significant increase in N2 emission relative to the control was observed after 6 days of root decomposition (Fig. 4.9a); an increase that was concurrent with high N2O emissions (Fig. 4.4). These results suggest that greater N2O to N2 reduction in soils dominated by small pores can only be a minor contributor to their ~21-fold lower N2O emissions. Moreover, I speculate that the differences in N2O diffusion rates in the large- and small-pore dominated soils were not as substantial as generally expected. N2O is highly water soluble and spatial heterogeneity in distribution patterns of soil water can markedly decrease its diffusion in gaseous phase (Shcherbak and Robertson, 2019). Water films and menisci in the large-pore soil material of this study occupy a large portion of the >30 µm pore space (Kutlu et al., 2018), likely slowing N2O diffusion (Fig. 4.16). Instead, I suggest that greater water absorption by tissues of plant residues (Iqbal et al., 2013; Myrold et al., 1981; Quemada and Cabrera, 2002) in large-pore dominated soil creates conditions favorable to denitrification, leading to enhanced production of both N2O and N2. Decomposing plant residues absorb water from adjacent soil pores, creating local micro-gradients of moisture within surrounding soil (Kim et al., 2020; Kutlu et al., 2018). This “sponge effect” (Kravchenko et al., 2017; Kim et al., 2020) is more pronounced when the soil around the residue 125 is dominated by large rather than small pores, due to lower water retention by soil with larger pores. Such local anoxic conditions within decomposing roots and in the adjacent former rhizosphere soil were accompanied by high available organic content (Fig S5 and S6) and active microbial communities, serving as respective sources and drivers of N2O production (Fig. 4.16). I speculate that, at the same time, some large pores in close spatial proximity to the decomposing roots were atmosphere-connected, thus instrumental to eventual emission of produced N2O and N2. Consistent with this explanation, > 60% of emitted N2O was root-derived for the entire incubation duration (starting from day 3) in the large-pore soil, while roots' contribution was much lower in small-pore soils, both near decomposing roots and in the bulk soil (Fig. 4.4d). My findings (i) emphasize the importance of root residues in soils dominated by large (> 30 µm) pores with respect to increased N2O emissions and (ii) elucidate a potential mechanism behind such increases (Fig. 4.16). Specifically, root residues act as powerful hotspots of N2O production, while diffusion of produced N2O via large pores magnifies the quantities emitted. Previous studies that investigated the interactive role of soil pore structure and moisture content on N2O emissions found marginal effect of pore structure on soil-derived N2O (Mangalassery et al., 2013; Rohe et al., 2021). Given that these studies used soil without dead roots, it reinforces my results that decomposing roots enhance the role of pore structure and are the main producer of N2O, facilitating the use of both root- and soil-derived N. On the other hand, caution is needed when applying the suggested mechanism in field soils. Given the experimental condition of 40% and 70% WFPS soil water content, my result may be applicable particularly in upland soils. Note that this study used soil with artificially contrasting pore characteristics to maximize the hypothesized pore effects. Thus, in field conditions where there is more wide range of pore sizes, the importance of pores and root existence might play a 126 less role than what is observed in this study. Likewise, natural root senescence and decay occur asynchronously in field switchgrass systems, while all part of the roots started to decay simultaneously in this study. Hence, the greenhouse gas emission observed in this study shows the maximized contribution of roots. Although extreme, the studied condition can be comparable to what is occurring in the field during fall senescence or after biomass harvest. 4.4.3 Relationships between chitinase activity, root decomposition, and N2O emissions Immediately after plant termination, the activity of chitinase was similar between the large- and small-pore soils, either near the roots or in the bulk soil (Fig. 4.10). However, after the onset of incubation, the chitinase activity in the large-pore soil increased dramatically, in contrast to only a minor increase in small-pore soil. The high activity in the large-pore soil was maintained during 21 days of root decomposition (Fig. 4.10). Chitinase activity can be viewed as a precurser of organic matter decomposition since it is involved in both C and N mineralization and cycling (Sinsabaugh and Moorhead, 1994; Tabatabai et al., 2010). Greater chitinase activity and consequently, decomposition of organic substances, took place in the large-pore than in the small- pore soil (Fig. 4.16). Note that switchgrass plants grew within two contrasting soils from their germination and, subsequently, their roots were decomposed in-situ. Thus, by the time of plant termination (10 weeks after sowing), the microbial community structure in the two soils might have been already different, because dominant microbial organisms and their activity depend on pore space characteristics (Gupta and Germida, 2015). Fungi are often more abundant in sand and grow better in pores of >10 μm (Chenu and Cosentino, 2011; Kögel‐Knabner et al., 2008; Witzgall et al., 2021), while bacteria are often preferentially located in < 10 μm pores (Schlüter et al., 2018). Thus, the 127 large-pore soil was likely more favorable to fungal growth, while small-pore soil favored bacterial growth. Since switchgrass roots commonly form symbiotic relationships with arbuscular mycorrhizal fungi (Jach‐Smith and Jackson, 2020), I speculate that fungal biomass accumulated during switchgrass growth played an important role in cycling of root-derived N; and greater amounts of labeled 13C and 15N processed by living switchgrass plants might have been assimilated into fungal biomass and hyphae. Higher level of chitin-rich fungal necromass (dead biomass and hyphae) and its processing in large-pore soils may explain substantially higher chitinase activity and concomitant increased N2O contribution from root-derived N (Fig. 4.4c) in large-pore soils. Micro-environmental conditions after plant termination were also beneficial for microbial decomposers, especially fungi, in large-pore than in small-pore dominated soils (Fig. 4.16). Adequate supply of O2 (Keiluweit et al., 2017) and greater organic inputs from roots of growing plants (Quigley et al., 2018) are regarded as the main drivers of enhanced microbial activity and high microbial turnover within large pores (Kravchenko et al., 2021). Specifically, I anticipate that a succession of fungi from mycorrhiza to saprotrophs occurred after plant termination (Herman et al., 2012), which is responsible for chitinase production and efficient decomposition of chitin-rich necromass from the fungal biomass and hyphae in the former rhizosphere. Saprotrophs prefer well- aerated large-pore soil and require less N compared to mycorrhiza (Leake et al., 2003), thus excess N from mycorrhizal fungal necromass might have been used for denitrification. Enhanced fungal activity likely was responsible for the observed 5-7-fold higher chitinase activity in the bulk soil of the large- than of the small-pore soil that developed within just the first few days of incubation (Fig. 4.10c and 8f). Fungi have higher contribution to chitinase production in soil compared to bacteria (De Boer et al., 1999; Miller et al., 1998; Yanai and Toyota, 2005), thus the chitinase activity measured in this study, to a certain extent, reflect the level of fungal activity. 128 Higher NH4+ during the first few days of root decomposition (Fig. 4.12a) in the large-pore soil was another indicator of faster decomposition and N mineralization taking place. A positive correlation between the chitinase activity measured directly on the decomposing roots (i.e., root chitinase activity) and the rate of emission of root-derived N2O (R2 = 0.22) supports the link between decomposition of chitin contained in roots and N2O production (Fig. 4.11a). I speculate that the rapid decomposition of roots in large-pore soils contributed to a great supply of inorganic N. Nitrate mineralized from decomposed organic substances (e.g., chitin) can be immediately utilized by denitrifying microbes (Fan et al., 2014), contributing to more pronounced root-derived N2O production in large-pore soils (Fig. 4.6). 4.4.4 Longevity of high N2O emissions The longevity of N2O emissions in the current experiment may be attributed to the rhizosphere legacy, that is, of roots being decomposed in-situ. Rhizosphere formed in the large- pore soil of this study represented an ideal micro-environment for microorganisms (Chau et al., 2011; Vieira et al., 2020). For a period of time from few days to almost 10 weeks, rhizosphere in the large-pore soil might have continuously received root exudates and other rhizodeposits such as lysates, dead fine roots, and mucilage (Lynch and Whipps, 1990; Nguyen, 2003). Upon plant termination, not only the decaying roots but also the previously deposited organic inputs in the former rhizosphere's soil were the subjects of intense decomposition by active and abundant microbial communities, acting in the optimal micro-environment of large pores (Kravchenko et al., 2021). Additionally, microbial biomass from organisms that previously assimilated the root exudates during live plant growth can be lysed after plant termination and subjected to decomposition afterwards. Root-mediated pH changes in the rhizosphere (Blossfeld et al., 2013; Hinsinger et al., 2003) can also change pathways and magnitudes of N2O production. For instance, 129 decreased pH can lead to heterotrophic denitrification as the main N2O production pathway (Duan et al., 2019) at decomposition stage. My results strongly suggest that micro- and mesocosm incubations with mixed-in root residues might underestimate the durations and intensities of N2O emissions. High N2O emissions from the large-pore soil were maintained over 21 days of root decomposition and only slightly decreased by day 39. In contrast, in the small-pore soil the N2O emissions decreased much faster (Fig. 4.4a and b), accompanied by a decrease in the proportion of root-derived N2O (Fig. 4.4c). Apparently, the root-based hotspots of N2O production in the soil dominated by large pores can last longer and substantially contribute to the total emissions for more than a month after plant termination. This finding contradicts previous reports where N2O reached its base level in 7-14 days after the start of incubation (Kravchenko et al., 2017; 2018). Similarly, short-termed flashes of N2O emissions were reported in other plant residue decomposition studies, where N2O emission rates reached base levels within 14-32 days (Begum et al., 2014; Fan et al., 2014; Köbke et al., 2018; Li et al., 2016). In the previous work, as in most prior studies, plant residues were incorporated to sieved soil; and when switchgrass roots were placed into the large-pore soil material, similar to those used in this study, peaks of N2O emissions lasted only 2-4 days (Kim et al., 2021). 4.4.5 Decoupled N2O and CO2 production in large-pore dominated soils CO2 emission is one of the general indicators of microbial activity, and since many soil processes – including residue decomposition, nitrification, and denitrification – are mediated by soil microbes, in residue amended soils CO2 and N2O emissions are often found to be positively correlated (Azam et al., 2002; Millar and Baggs, 2004). Yet, such correlation was present only in the small-pore (R2=0.68), but not in the large-pore (R2 =0.07) soils (Fig. 4.8). At the same level of 130 CO2 emission, i.e., same general microbial activity, large-pore soil emitted 10-20 times more N2O compared to the small-pore soil (Figs. 4 and 6). This decoupling of CO2 and N2O emissions suggests that different groups of microorganisms were involved in N2O production in large- and small-pore dominated soil, and the groups active in the large-pore soils were those that produced N2O but not CO2. 4.4.6 N2O priming Roots, in-situ decomposing in the large-pore soil, were not only a major source of N2O; they also stimulated N2O production from decomposition of intrinsic SOM via positive N2O priming (Fig. 4.6a and b, Fig. 4.16). The phenomenon of positive N2O priming has been studied only recently (Roman‐Perez and Hernandez‐Ramirez, 2020) and its exact mechanisms are still poorly understood. Roman-Perez and Hernandez-Ramirez (2020) reported that high SOM-C and SOM-N availability generated a positive N2O priming effect and suggested a successional shift in priming mechanisms from stoichiometry to N-mining. Since the availability of SOM in the rhizosphere can increase due to root exudates during plant growth (Li et al., 2021), the rhizosphere legacy might have maximized the magnitude of N2O priming. My N2O and CO2 priming results further support the notion of decoupling of N2O and CO2 production in presence of in-situ decomposing switchgrass roots, the effect especially pronounced in the large-pore soils, but present in the small-pore soil as well. Positive N2O priming coincided with negative or negligible CO2 priming (Fig. 4.6). Negative C priming is observed when the added organic matter is recalcitrant (Guenet et al., 2010; Zhang et al., 2019), but in other cases, addition of highly available substrate also causes preferential substrate utilization leading to negative priming (Kuzyakov and Bol, 2006). Since C:N ratios of the studied young switchgrass roots were relatively low (~15), it is possible that they were the preferred source of C to soil microorganisms 131 over SOM, resulting in reduced SOM decomposition. Yet, some microorganisms, specifically those responsible for N2O production from both root and soil N sources and those responsible for chitinase production during root decomposition, were apparently strongly stimulated. Such joint stimulation took place on the root surfaces and in their close proximity, probably in the former rhizosphere, rather than in the bulk soil. Notably, the temporal patterns of N2O priming (Fig. 4.7) closely followed those of the chitinase activity on the decomposing roots (Fig. 4.10b and e), while there were no similarities with the chitinase activity patterns in bulk soil (Fig. 4.10c and f). Also, soil chitinase activity and soil-derived N2O emission were not related to each other (Fig. 4.11c), while chitinase activity on the root surfaces and plant-derived N2O emission were positively correlated (Fig. 4.11a). Previous reports state that the extracellular enzymes produced in response to newly added organic matter can be efficient at SOM decomposition and can lead to positive priming (Fontaine et al., 2003; Wu et al., 1993). Thus, the greater extent of N2O priming in the soil dominated by large pores (Fig 4.6a, b, and Fig. 4.7) is likely driven by higher enzyme activity near decomposing roots there (Fig. 4.10), which stimulated the decomposition of not only roots but also N-containing SOM, resulting in higher contents of mineral N in large-pore soils (Fig. 4.12). Strategical procurement of N but not C from SOM by microorganisms involved in N2O production apparently can take place in the former rhizosphere soil during root residue decomposition. 4.5 Conclusion The effect of soil pore architecture on N2O production from soil with in-situ decomposing switchgrass roots was 6 times higher than that of the soil moisture. Markedly greater quantities of N2O were emitted from the soil dominated by the large (>30 µm) pores than from the soil dominated by small (< 10 µm) pores, while only modestly higher emissions were observed in 132 incubations at 70% as compared to 40% WFPS. Longevity of N2O emission hotspots in the large- pore soils exceeded that of the small-pore soils, as well as that of previously published incubation studies. Greater and prolonged N2O emissions in the large-pore soils resulted from greater N2O production, and not from the lower N2O to N2 conversion, as previously expected. The enhanced emissions were driven by higher microbial activity, possibly fungi, as attested by marked surge in chitinase activity. Chitinase activity in the large-pore soils increased at a greater rate and remained higher than in the small-pore soil, suggesting that the high N2O emissions were related to faster root decomposition and microbial turnover in the former rhizosphere. Root-derived N was the main source of N2O, especially in the large-pore soils; however, in-situ root decomposition also accelerated mining of SOM-N for N2O and N2O priming. Apparently, the location for such priming was the former rhizosphere soil in close proximity to the decaying roots. Microbial processes leading to N2O production in the soil dominated by large pores become decoupled from the CO2 generating microbial activities, suggesting a different group of microbes being involved in N2O production, possibly, strategically procuring N from the root residues and the soil. I demonstrated that the rhizosphere soil carries a strong legacy effect as it turns into detritusphere after plant death and decomposition. Ignoring this effect by mixing root residues with destructed soil may underestimate the durations and intensities of N2O emissions. My study emphasized the pore architecture as a crucial factor that determines the magnitude and longevity of N2O emission hotspots. It revealed that the role of pores in pathways of climate change can be more important than previously perceived. 133 ACKNOWLEDGEMENTS Support for this research was provided by the Great Lakes Bioenergy Research Center, U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research (Award DE-SC0018409), by the National Science Foundation Long-term Ecological Research Program (DEB 1832042) at the Kellogg Biological Station, and by Michigan State University AgBioResearch, and the “RUDN University program 5-100”. I appreciate the support from Samuel Decamp, Kevin Kahmark, and David Weed for sample analyses and related lab works. I also want to thank Lisa Tiemann for sharing the seed stimulation/germination protocol. 134 REFERENCES 135 REFERENCES Adler, P.R., Grosso, S.J.D. and Parton, W.J., 2007. Life‐cycle assessment of net greenhouse‐gas flux for bioenergy cropping systems. Ecological Applications. 17, 675-691. Angers, D.A. and Caron, J., 1998. Plant-induced changes in soil structure: processes and feedbacks. Biogeochemistry. 42, 55-72. Aulakh M., Doran J., Walters D., Mosier A., Francis D., 1991. Crop residue type and placement effects on denitrification and mineralization. Soil Science Society of America Journal. 55, 1020– 1025. Azam, F., Müller, C., Weiske, A., Benckiser, G. and Ottow, J., 2002. Nitrification and denitrification as sources of atmospheric nitrous oxide–role of oxidizable carbon and applied nitrogen. Biology and Fertility of Soils. 35, 54-61. Bardgett, R.D., Mommer, L. and De Vries, F.T., 2014. Going underground: root traits as drivers of ecosystem processes. Trends in Ecology & Evolution. 29, 692-699. Bateman, E. and Baggs, E., 2005. Contributions of nitrification and denitrification to N2O emissions from soils at different water-filled pore space. Biology and Fertility of Soils. 41, 379- 388. Begum, N., Guppy, C., Herridge, D. and Schwenke, G., 2014. Influence of source and quality of plant residues on emissions of N2O and CO2 from a fertile, acidic Black Vertisol. Biology and Fertility of Soils. 50, 499-506. Bengough, A., 2012. Water dynamics of the root zone: rhizosphere biophysics and its control on soil hydrology. Vadose Zone Journal. 11, vzj2011.0111. Berendsen, R.L., Pieterse, C.M. and Bakker, P.A., 2012. The rhizosphere microbiome and plant health. Trends in Plant Science. 17, 478-486. Bird, J.A., Herman, D.J. and Firestone, M.K., 2011. Rhizosphere priming of soil organic matter by bacterial groups in a grassland soil. Soil Biology and Biochemistry. 43, 718-725. Bird, J.A., Kleber, M. and Torn, M.S., 2008. 13C and 15N stabilization dynamics in soil organic matter fractions during needle and fine root decomposition. Organic Geochemistry. 39, 465-477. Blagodatsky, S., Grote, R., Kiese, R., Werner, C. and Butterbach-Bahl, K., 2011. Modelling of microbial carbon and nitrogen turnover in soil with special emphasis on N-trace gases emission. Plant and Soil. 346, 297-330. Blair, N., Faulkner, R., Till, A. and Sanchez, P., 2005. Decomposition of 13C and 15N labelled plant residue materials in two different soil types and its impact on soil carbon, nitrogen, aggregate stability, and aggregate formation. Soil Research. 43, 873-886. 136 Blossfeld, S., Schreiber, C.M., Liebsch, G., Kuhn, A.J. and Hinsinger, P., 2013. Quantitative imaging of rhizosphere pH and CO2 dynamics with planar optodes. Annals of Botany. 112(2), 267-276. Bollmann, A. and Conrad, R., 1998. Influence of O2 availability on NO and N2O release by nitrification and denitrification in soils. Global Change Biology. 4, 387-396. Bracken, C.J., Lanigan, G.J., Richards, K.G., Müller, C., Tracy, S.R., Grant, J., Krol, D.J., Sheridan, H., Lynch, M.B., Grace, C. and Fritch, R., 2021. Source partitioning using N2O isotopomers and soil WFPS to establish dominant N2O production pathways from different pasture sward compositions. Science of the Total Environment. 781, 146515. Braker, G. and Conrad, R., 2011. Diversity, structure, and size of N2O-producing microbial communities in soils—what matters for their functioning? Advances in Applied Microbiology. 75, 33-70. Bromand, S., Whalen, J.K., Janzen, H.H., Schjoerring, J.K. and Ellert, B., 2001. A pulse-labelling method to generate 13C-enriched plant materials. Plant and Soil, 235(2): 253-257. Brookes, P., Landman, A., Pruden, G. and Jenkinson, D., 1985. Chloroform fumigation and the release of soil nitrogen: a rapid direct extraction method to measure microbial biomass nitrogen in soil. Soil Biology and Biochemistry. 17, 837-842. Butterbach-Bahl, K., Baggs, E.M., Dannenmann, M., Kiese, R. and Zechmeister-Boltenstern, S., 2013. Nitrous oxide emissions from soils: how well do we understand the processes and their controls? Philosophical Transactions of the Royal Society B: Biological Sciences. 368, 20130122. Calleja-Cabrera, J., Boter, M., Oñate-Sánchez, L. and Pernas, M., 2020. Root growth adaptation to climate change in crops. Frontiers in Plant Science. 11, 544. Cantarel, A.A., Bloor, J.M. and Soussana, J.F., 2013. Four years of simulated climate change reduces above‐ground productivity and alters functional diversity in a grassland ecosystem. Journal of Vegetation Science. 24, 113-126. Caplan, J.S., Giménez, D., Subroy, V., Heck, R.J., Prior, S.A., Runion, G.B. and Torbert, H.A., 2017. Nitrogen‐mediated effects of elevated CO 2 on intra‐aggregate soil pore structure. Global Change Biology. 23, 1585-1597. Carminati, A., Moradi, A.B., Vetterlein, D., Vontobel, P., Lehmann, E., Weller, U., Vogel, H.-J. and Oswald, S.E., 2010. Dynamics of soil water content in the rhizosphere. Plant and Soil. 332, 163-176. Castellano, M.J., Schmidt, J.P., Kaye, J.P., Walker, C., Graham, C.B., Lin, H. and Dell, C.J., 2010. Hydrological and biogeochemical controls on the timing and magnitude of nitrous oxide flux across an agricultural landscape. Global Change Biology. 16, 2711-2720. Chen, H., Li, X., Hu, F. and Shi, W., 2013. Soil nitrous oxide emissions following crop residue addition: A meta‐analysis. Global Change Biology. 19, 2956-2964. 137 Chen, R., Senbayram, M., Blagodatsky, S., Myachina, O., Dittert, K., Lin, X., Blagodatskaya, E. and Kuzyakov, Y., 2014. Soil C and N availability determine the priming effect: microbial N mining and stoichiometric decomposition theories. Global Change Biology. 20, 2356-2367. Chenu, C. and Cosentino, D., 2011. Microbial regulation of soil structural dynamics. The Architecture and Biology of Soils: Life in Inner Space. 37-70. Cherubini, F. and Jungmeier, G., 2010. LCA of a biorefinery concept producing bioethanol, bioenergy, and chemicals from switchgrass. The International Journal of Life Cycle Assessment. 15, 53-66. Daly, E.J. and Hernandez-Ramirez, G., 2020. Sources and priming of soil N2O and CO2 production: Nitrogen and simulated exudate additions. Soil Biology and Biochemistry. 149, 107942. Davidson, E.A., 1991. Fluxes of nitrous oxide and nitric oxide from terrestrial ecosystems. Microbial Production and Consumption of Greenhouse Gases: Methane, Nitrogen Oxides, and Halomethanes., 219-235. Davidson, E.A., Swank, W.T. and Perry, T.O., 1986. Distinguishing between nitrification and denitrification as sources of gaseous nitrogen production in soil. Applied and Environmental Microbiology. 52, 1280-1286. De Boer, W., Gerards, S., Gunnewiek, P.K. and Modderman, R., 1999. Response of the chitinolytic microbial community to chitin amendments of dune soils. Biology and Fertility of Soils, 29(2): 170-177. de Graaff, M.-A., Six, J., Jastrow, J.D., Schadt, C.W. and Wullschleger, S.D., 2013. Variation in root architecture among switchgrass cultivars impacts root decomposition rates. Soil Biology and Biochemistry. 58, 198-206. Dijkstra, F.A., Bader, N.E., Johnson, D.W. and Cheng, W., 2009. Does accelerated soil organic matter decomposition in the presence of plants increase plant N availability? Soil Biology and Biochemistry. 41, 1080-1087. Doane, T.A. and Horwáth, W.R., 2003. Spectrophotometric determination of nitrate with a single reagent. Analytical Letters. 36, 2713-2722. Doube, M., Kłosowski, M.M., Arganda-Carreras, I., Cordeliéres, F., Dougherty, R.P., Jackson, J., Schmid. B,, Hutchinson. J,R,, Shefelbine. S.J., 2010, BoneJ: free and extensible bone image analysis in ImageJ. Bone 47:1076-9. Drew, E., Murray, R., Smith, S. and Jakobsen, I., 2003. Beyond the rhizosphere: growth and function of arbuscular mycorrhizal external hyphae in sands of varying pore sizes. Plant and Soil. 251, 105-114. 138 Duan, P., Song, Y., Li, S. and Xiong, Z., 2019. Responses of N2O production pathways and related functional microbes to temperature across greenhouse vegetable field soils. Geoderma. 355, 113904. Emery, S.M., Kinnetz, E.R., Bell-Dereske, L., Stahlheber, K.A., Gross, K.L. and Pennington, D., 2018. Low variation in arbuscular mycorrhizal fungal associations and effects on biomass among switchgrass cultivars. Biomass and Bioenergy. 119, 503-508. Fan, F., Yin, C., Tang, Y., Li, Z., Song, A., Wakelin, S.A., Zou, J. and Liang, Y., 2014. Probing potential microbial coupling of carbon and nitrogen cycling during decomposition of maize residue by 13C-DNA-SIP. Soil Biology and Biochemistry. 70, 12-21. Fan, P. and Guo, D., 2010. Slow decomposition of lower order roots: a key mechanism of root carbon and nutrient retention in the soil. Oecologia. 163, 509-515. Fontaine, S., Mariotti, A. and Abbadie, L., 2003. The priming effect of organic matter: a question of microbial competition? Soil Biology and Biochemistry. 35, 837-843. Gaillard, V., Chenu, C., Recous, S. and Richard, G., 1999. Carbon, nitrogen and microbial gradients induced by plant residues decomposing in soil. European Journal of Soil Science. 50, 567-578. Gelfand, I., Hamilton, S.K., Kravchenko, A.N., Jackson, R.D., Thelen, K.D. and Robertson, G.P., 2020. Empirical Evidence for the Potential Climate Benefits of Decarbonizing Light Vehicle Transport in the US with Bioenergy from Purpose-Grown Biomass with and without BECCS. Environmental Science & Technology. 54, 2961-2974. Groffman, P.M. and Tiedje, J.M., 1989. Denitrification in north temperate forest soils: spatial and temporal patterns at the landscape and seasonal scales. Soil Biology and Biochemistry. 21, 613- 620. Groffman, P.M., Butterbach-Bahl, K., Fulweiler, R.W., Gold, A.J., Morse, J.L., Stander, E.K., Tague, C., Tonitto, C. and Vidon, P., 2009. Challenges to incorporating spatially and temporally explicit phenomena (hotspots and hot moments) in denitrification models. Biogeochemistry. 93, 49-77. Gu, J., Nicoullaud, B., Rochette, P., Grossel, A., Hénault, C., Cellier, P. and Richard, G., 2013. A regional experiment suggests that soil texture is a major control of N2O emissions from tile- drained winter wheat fields during the fertilization period. Soil Biology and Biochemistry. 60, 134- 141. Guber, A., Blagodatskaya, E., Juyal, A., Razavi, B.S., Kuzyakov, Y. and Kravchenko, A., 2021. Time-lapse approach: correcting deficiencies of 2D soil zymography. Soil Biology and Biochemistry. 108225. Guber, A., Kravchenko, A., Razavi, B.S., Uteau, D., Peth, S., Blagodatskaya, E. and Kuzyakov, Y., 2018. Quantitative soil zymography: mechanisms, processes of substrate and enzyme diffusion in porous media. Soil Biology and Biochemistry. 127, 156-167. 139 Guber, A.K., Kravchenko, A.N., Razavi, B.S., Blagodatskaya, E. and Kuzyakov, Y., 2019. Calibration of 2‐D soil zymography for correct analysis of enzyme distribution. European Journal of Soil Science. 70, 715-726. Guenet, B., Leloup, J., Raynaud, X., Bardoux, G. and Abbadie, L., 2010. Negative priming effect on mineralization in a soil free of vegetation for 80 years. European Journal of Soil Science. 61, 384-391. Gupta, V.V. and Germida, J.J., 2015. Soil aggregation: Influence on microbial biomass and implications for biological processes. Soil Biology and Biochemistry. 80, A3-A9. Heitkötter, J. and Marschner, B., 2018. Soil zymography as a powerful tool for exploring hotspots and substrate limitation in undisturbed subsoil. Soil Biology and Biochemistry. 124, 210-217. Helliwell, J.R., Sturrock, C.J., Miller, A.J., Whalley, W.R. and Mooney, S.J., 2019. The role of plant species and soil condition in the structural development of the rhizosphere. Plant, Cell & Environment, 42(6): 1974-1986. Hénault, C., Grossel, A., Mary, B., Roussel, M. and Léonard, J., 2012. Nitrous oxide emission by agricultural soils: a review of spatial and temporal variability for mitigation. Pedosphere, 22(4): 426-433. Herman, D.J., Firestone, M.K., Nuccio, E. and Hodge, A., 2012. Interactions between an arbuscular mycorrhizal fungus and a soil microbial community mediating litter decomposition. FEMS Microbiology Ecology. 80, 236-247. Hinsinger P., Plassard C., Tang C., Jaillard B., 2003. Origins of root-mediated pH changes in the rhizosphere and their responses to environmental constraints: a review. Plant and Soil. 248, 43-59. Hirmas, D.R., Giménez, D., Nemes, A., Kerry, R., Brunsell, N.A. and Wilson, C.J., 2018. Climate- induced changes in continental-scale soil macroporosity may intensify water cycle. Nature. 561, 100-103. Hu, X., Liu, L., Zhu, B., Du, E., Hu, X., Li, P., Zhou, Z., Ji, C., Zhu, J. and Shen, H., 2016. Asynchronous responses of soil carbon dioxide, nitrous oxide emissions and net nitrogen mineralization to enhanced fine root input. Soil Biology and Biochemistry. 92, 67-78. Hupe, A., Schulz, H., Bruns, C., Haase, T., Heß, J., Joergensen, R.G. and Wichern, F., 2018. Even flow? Changes of carbon and nitrogen release from pea roots over time. Plant and Soil. 431, 143- 157. Iqbal, A., Beaugrand, J., Garnier, P. and Recous, S., 2013. Tissue density determines the water storage characteristics of crop residues. Plant and Soil. 367, 285-299. Jach‐Smith, L.C. and Jackson, R.D., 2020. Inorganic N addition replaces N supplied to switchgrass (Panicum virgatum) by arbuscular mycorrhizal fungi. Ecological Applications. 30, e02047. 140 Jenkinson, D., Fox, R. and Rayner, J., 1985. Interactions between fertilizer nitrogen and soil nitrogen—the so‐called ‘priming’effect. Journal of soil Science. 36, 425-444. Jin, K., White, P.J., Whalley, W.R., Shen, J. and Shi, L., 2017. Shaping an optimal soil by root– soil interaction. Trends in Plant Science. 22, 823-829. Kaiser, E.-A. and Heinemeyer, O., 1996. Temporal changes in N2O-losses from two arable soils. Plant and Soil. 181, 57-63. Kallenbach, C.M. and Grandy, S.A., 2015. Land‐use legacies regulate decomposition dynamics following bioenergy crop conversion. GCB Bioenergy. 7, 1232-1244. Koebernick, N. et al., 2017. High‐resolution synchrotron imaging shows that root hairs influence rhizosphere soil structure formation. New Phytologist, 216(1): 124-135. Keiluweit, M., Wanzek, T., Kleber, M., Nico, P. and Fendorf, S., 2017. Anaerobic microsites have an unaccounted role in soil carbon stabilization. Nature communications. 8, 1-10. Khosrozadeh, S., Guber, A.K., Kravchenko, A.N., Ghaderi, N., Blagodatskaya, E., 2021. Soil Oxidoreductase Zymography: Visualizing the Spatial Distribution of Peroxidases and Phenol Oxidases Activity at the Root-Soil Interface. In review Kim, K., Guber, A., Rivers, M. and Kravchenko, A., 2020. Contribution of decomposing plant roots to N2O emissions by water absorption. Geoderma. 375, 114506. Kim, K., Kutlu, T., Kravchenko, A. and Guber, A., 2021. Dynamics of N2O in vicinity of plant residues: a microsensor approach. Plant and Soil. Köbke, S., Senbayram, M., Pfeiffer, B., Nacke, H. and Dittert, K., 2018. Post-harvest N2O and CO2 emissions related to plant residue incorporation of oilseed rape and barley straw depend on soil NO3-content. Soil and Tillage Research. 179, 105-113. Kögel‐Knabner, I., Guggenberger, G., Kleber, M., Kandeler, E., Kalbitz, K., Scheu, S., Eusterhues, K. and Leinweber, P., 2008. Organo‐mineral associations in temperate soils: Integrating biology, mineralogy, and organic matter chemistry. Journal of Plant Nutrition and Soil Science. 171, 61-82. Kravchenko, A., Fry, J. and Guber, A., 2018. Water absorption capacity of soil-incorporated plant leaves can affect N2O emissions and soil inorganic N concentrations. Soil Biology and Biochemistry. 121, 113-119. Kravchenko, A., Guber, A., Gunina, A., Dippold, M. and Kuzyakov, Y., 2021. Pore‐scale view of microbial turnover: Combining 14C imaging, μCT and zymography after adding soluble carbon to soil pores of specific sizes. European Journal of Soil Science. 72, 593-607. Kravchenko, A., Toosi, E., Guber, A., Ostrom, N., Yu, J., Azeem, K., Rivers, M. and Robertson, G., 2017. Hotspots of soil N2O emission enhanced through water absorption by plant residue. Nature Geoscience. 10, 496-500. 141 Kutlu, T., Guber, A.K., Rivers, M.L. and Kravchenko, A.N., 2018. Moisture absorption by plant residue in soil. Geoderma. 316, 47-55. Kuzyakov, Y., 2010. Priming effects: interactions between living and dead organic matter. Soil Biology and Biochemistry. 42, 1363-1371. Kuzyakov, Y. and Bol, R., 2006. Sources and mechanisms of priming effect induced in two grassland soils amended with slurry and sugar. Soil Biology and Biochemistry. 38, 747-758. Kuzyakov, Y., Friedel, J. and Stahr, K., 2000. Review of mechanisms and quantification of priming effects. Soil Biology and Biochemistry. 32, 1485-1498. Kuzyakov Y., Razavi B.S. 2019. Rhizosphere size and shape: Temporal dynamics and spatial stationarity. Soil Biology and Biochemistry 135, 343-360. Leake, J.R., Donnelly, D.P. and Boddy, L., 2003. Interactions Between Ecto-mycorrhizal and Saprotrophic Fungi. In: M.G.A. van der Heijden and I.R. Sanders (Editors), Mycorrhizal Ecology. Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 345-372. Li, H., Bölscher, T., Winnick, M., Tfaily, M.M., Cardon, Z.G. and Keiluweit, M., 2021. Simple Plant and Microbial Exudates Destabilize Mineral-Associated Organic Matter via Multiple Pathways. Environmental Science & Technology. 55, 3389-3398. Li, X., Sørensen, P., Olesen, J.E. and Petersen, S.O., 2016. Evidence for denitrification as main source of N2O emission from residue-amended soil. Soil Biology and Biochemistry. 92, 153-160. Liang, C., Jesus, E.d.C., Duncan, D.S., Quensen, J.F., Jackson, R.D., Balser, T.C. and Tiedje, J.M., 2016. Switchgrass rhizospheres stimulate microbial biomass but deplete microbial necromass in agricultural soils of the upper Midwest, USA. Soil Biology and Biochemistry. 94, 173-180. Liu, C., Lu, M., Cui, J., Li, B. and Fang, C., 2014. Effects of straw carbon input on carbon dynamics in agricultural soils: a meta‐analysis. Global Change Biology. 20, 1366-1381. Liu, S., Razavi, B.S., Su, X., Maharjan, M., Zarebanadkouki, M., Blagodatskaya, E. and Kuzyakov, Y., 2017. Spatio-temporal patterns of enzyme activities after manure application reflect mechanisms of niche differentiation between plants and microorganisms. Soil Biology and Biochemistry. 112, 100-109. Lynch, J. and Whipps, J., 1990. Substrate flow in the rhizosphere. Plant and soil, 129(1): 1-10. Ma, X., Razavi, B.S., Holz, M., Blagodatskaya, E. and Kuzyakov, Y., 2017. Warming increases hotspot areas of enzyme activity and shortens the duration of hot moments in the root-detritusphere. Soil Biology and Biochemistry. 107, 226-233. Mangalassery, S., Sjögersten, S., Sparkes, D.L., Sturrock, C.J., Mooney, S.J., 2013. The effect of soil aggregate size on pore structure and its consequence on emission of greenhouse gases. Soil and Tillage Research 132, 39-46. 142 Marschner, H., Römheld, V. and Cakmak, I., 1987. Root‐induced changes of nutrient availability in the rhizosphere. Journal of Plant Nutrition. 10, 1175-1184. Marschner, P., Marhan, S. and Kandeler, E., 2012. Microscale distribution and function of soil microorganisms in the interface between rhizosphere and detritusphere. Soil Biology and Biochemistry. 49, 174-183. Mašková, T. and Klimeš, A., 2020. The effect of rhizoboxes on plant growth and root: shoot biomass partitioning. Frontiers in Plant Science, 10: 1693. Meier, I.C., Finzi, A.C. and Phillips, R.P., 2017. Root exudates increase N availability by stimulating microbial turnover of fast-cycling N pools. Soil Biology and Biochemistry. 106, 119- 128. Millar, N., Baggs, E.M., 2004. Chemical composition, or quality, of agroforestry residues influences N2O emissions after their addition to soil. Soil Biology and Biochemistry. 36, 935–943. Miller, M., Palojärvi, A., Rangger, A., Reeslev, M. and Kjøller, A., 1998. The use of fluorogenic substrates to measure fungal presence and activity in soil. Applied and Environmental Microbiology, 64(2): 613-617. Milliken, G.A. and Johnson, D.E., 2002. Analysis of messy data, volume III: analysis of covariance, CRC Press. Milliken, G.A. and Johnson, D.E., 2009. Analysis of messy data volume 1: designed experiments, CRC Press. Mitchell, E., Scheer, C., Rowlings, D., Conant, R.T., Cotrufo, M.F. and Grace, P., 2018. Amount and incorporation of plant residue inputs modify residue stabilisation dynamics in soil organic matter fractions. Agriculture, Ecosystems & Environment. 256, 82-91. Monti, A., Barbanti, L., Zatta, A. and Zegada‐Lizarazu, W., 2012. The contribution of switchgrass in reducing GHG emissions. GCB Bioenergy. 4, 420-434. Morley, N. and Baggs, E., 2010. Carbon and oxygen controls on N2O and N2 production during nitrate reduction. Soil Biology and Biochemistry, 42(10): 1864-1871. Myrold, D., Elliott, L., Papendick, R. and Campbell, G., 1981. Water Potential‐Water Content Characteristics of Wheat Straw. Soil Science Society of America Journal. 45, 329-333. Myrold, D.D. and Tiedje, J.M., 1985. Diffusional constraints on denitrification in soil. Soil Science Society of America Journal. 49, 651-657. Nguyen, C., 2003. Rhizodeposition of organic C by plants: mechanisms and controls. Agronomie, 23(5-6): 375-396. 143 Nicolardot, B., Bouziri, L., Bastian, F. and Ranjard, L., 2007. A microcosm experiment to evaluate the influence of location and quality of plant residues on residue decomposition and genetic structure of soil microbial communities. Soil Biology and Biochemistry. 39, 1631-1644. Nuccio, E.E., Hodge, A., Pett‐Ridge, J., Herman, D.J., Weber, P.K. and Firestone, M.K., 2013. An arbuscular mycorrhizal fungus significantly modifies the soil bacterial community and nitrogen cycling during litter decomposition. Environmental Microbiology. 15, 1870-1881. Pachauri, R.K., Allen, M.R., Barros, V.R., Broome, J., Cramer, W., Christ, R., Church, J.A., Clarke, L., Dahe, Q. and Dasgupta, P., 2014. Climate change 2014: synthesis report. Contribution of Working Groups I, II and III to the fifth assessment report of the Intergovernmental Panel on Climate Change, IPCC. Parkin, T.B., 1987. Soil microsites as a source of denitrification variability. Soil Science Society of America Journal. 51, 1194-1199. Quemada, M. and Cabrera, M., 2002. Characteristic moisture curves and maximum water content of two crop residues. Plant and Soil. 238, 295-299. Quigley, M.Y., Negassa, W.C., Guber, A.K., Rivers, M.L. and Kravchenko, A.N., 2018. Influence of pore characteristics on the fate and distribution of newly added carbon. Frontiers in Environmental Science. 6, 51. Razavi, B.S., Zarebanadkouki, M., Blagodatskaya, E. and Kuzyakov, Y., 2016. Rhizosphere shape of lentil and maize: spatial distribution of enzyme activities. Soil Biology and Biochemistry. 96, 229-237. Roberts, B.J., Russ, M.E. and Ostrom, N.E., 2000. Rapid and Precise Determination of the δ18O of Dissolved and Gaseous Dioxygen via Gas Chromatography− Isotope Ratio Mass Spectrometry. Environmental Science & Technology. 34, 2337-2341. Rohe, L., Apelt, B., Vogel, H.J., Well, R., Wu, G.M., Schlüter, S., 2021. Denitrification in soil as a function of oxygen availability at the microscale. Biogeosciences 18(3), 1185-1201. Roley, S.S., Duncan, D.S., Liang, D., Garoutte, A., Jackson, R.D., Tiedje, J.M. and Robertson, G.P., 2018. Associative nitrogen fixation (ANF) in switchgrass (Panicum virgatum) across a nitrogen input gradient. PloS One. 13, e0197320. Roman‐Perez, C.C. and Hernandez‐Ramirez, G., 2020. Sources and priming of N2O production across a range of moisture contents in a soil with high organic matter. Soil Biology and Biochemistry. 149, 107942. Ruan, L., Bhardwaj, A.K., Hamilton, S.K. and Robertson, G.P., 2016. Nitrogen fertilization challenges the climate benefit of cellulosic biofuels. Environmental Research Letters, 11(6): 064007 Rummel, P.S., Well, R., Pfeiffer, B., Dittert, K., Floßmann, S. and Pausch, J., 2020. Nitrate uptake and carbon exudation–do plant roots stimulate or inhibit denitrification? Plant and Soil. 1-17. 144 Ruser, R., Flessa, H., Russow, R., Schmidt, G., Buegger, F. and Munch, J., 2006. Emission of N2O, N2 and CO2 from soil fertilized with nitrate: effect of compaction, soil moisture and rewetting. Soil Biology and Biochemistry. 38, 263-274. Saggar, S., Jha, N., Deslippe, J., Bolan, N.S., Luo, J., Giltrap, D.L., Kim, D.G., Zaman, M. and Tillman, R.W., 2013. Denitrification and N2O: N2 production in temperate grasslands: Processes, measurements, modelling and mitigating negative impacts. Science of the Total Environment, 465: 173-195. Sanaullah, M., Razavi, B.S., Blagodatskaya, E. and Kuzyakov, Y., 2016. Spatial distribution and catalytic mechanisms of β-glucosidase activity at the root-soil interface. Biology and Fertility of Soils. 52, 505-514. Schaufler, G., Kitzler, B., Schindlbacher, A., Skiba, U., Sutton, M. and Zechmeister‐Boltenstern, S., 2010. Greenhouse gas emissions from European soils under different land use: effects of soil moisture and temperature. European Journal of Soil Science. 61, 683-696. Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, T., Preibisch, S., Rueden, C., Saalfeld, S. and Schmid, B., 2012. Fiji: an open-source platform for biological-image analysis. Nature Methods. 9, 676-682. Schleusner, P., Lammirato, C., Tierling, J., Lebender, U. and Rütting, T., 2018. Primed N2O emission from native soil nitrogen: A 15N‐tracing laboratory experiment. Journal of Plant Nutrition and Soil Science. 181, 621-627. Schlüter, S., Eickhorst, T. and Mueller, C.W., 2018. Correlative imaging reveals holistic view of soil microenvironments. Environmental Science & Technology. 53, 829-837. Schlüter, S., Henjes, S., Zawallich, J., Bergaust, L., Horn, M., Ippisch, O., Vogel, H.-J. and Dörsch, P., 2018. Denitrification in soil aggregate analogues-effect of aggregate size and oxygen diffusion. Frontiers in Environmental Science. 6, 17. Schlüter, S., Zawallich, J., Vogel, H.-J. and Dörsch, P., 2019. Physical constraints for respiration in microbial hotspots in soil and their importance for denitrification. Biogeosciences. 16, 3665- 3678. Schmatz, R., Recous, S., Aita, C., Tahir, M.M., Schu, A.L., Chaves, B. and Giacomini, S.J., 2017. Crop residue quality and soil type influence the priming effect but not the fate of crop residue C. Plant and Soil. 414, 229-245. Schmidt, U., Thöni, H. and Kaupenjohann, M., 2000. Using a boundary line approach to analyze N2O flux data from agricultural soils. Nutrient Cycling in Agroecosystems. 57, 119-129. Shahbaz, M., Kuzyakov, Y. and Heitkamp, F., 2017. Decrease of soil organic matter stabilization with increasing inputs: mechanisms and controls. Geoderma. 304, 76-82. Shcherbak, I. and Robertson, G.P., 2019. Nitrous Oxide (N2O) Emissions from Subsurface Soils of Agricultural Ecosystems. Ecosystems. 22, 1650-1663. 145 Shelton, D.R., Sadeghi, A.M. and McCarty, G.W., 2000. Effect of soil water content on denitrification during cover crop decomposition. Soil Science. 165, 365-371. Sinsabaugh, R. and Moorhead, D., 1994. Resource allocation to extracellular enzyme production: a model for nitrogen and phosphorus control of litter decomposition. Soil Biology and Biochemistry. 26, 1305-1311. Sinsabaugh, R., Reynolds, H. and Long, T., 2000. Rapid assay for amidohydrolase (urease) activity in environmental samples. Soil Biology and Biochemistry. 32, 2095-2097. Spohn, M. and Kuzyakov, Y., 2013. Distribution of microbial-and root-derived phosphatase activities in the rhizosphere depending on P availability and C allocation–Coupling soil zymography with 14C imaging. Soil Biology and Biochemistry. 67, 106-113. Spohn, M. and Kuzyakov, Y., 2014. Spatial and temporal dynamics of hotspots of enzyme activity in soil as affected by living and dead roots—a soil zymography analysis. Plant and Soil. 379, 67- 77. Sutka, R., Ostrom, N., Ostrom, P., Gandhi, H. and Breznak, J., 2003. Nitrogen isotopomer site preference of N2O produced by Nitrosomonas europaea and Methylococcus capsulatus Bath. Rapid Communications in Mass Spectrometry. 17, 738-745. Syakila, A. and Kroeze, C., 2011. The global nitrous oxide budget revisited. Greenhouse Gas Measurement and Management. 1, 17-26. Tabatabai, M., Ekenler, M. and Senwo, Z., 2010. Significance of enzyme activities in soil nitrogen mineralization. Communications in Soil Science and Plant Analysis. 41, 595-605. Thilakarathna, S.K. and Hernandez-Ramirez, G., 2021a. Primings of soil organic matter and denitrification mediate the effects of moisture on nitrous oxide production. Soil Biology and Biochemistry. 108166. Thilakarathna, S.K. and Hernandez‐Ramirez, G., 2021b. How does management legacy, nitrogen addition, and nitrification inhibition affect soil organic matter priming and nitrous oxide production? Journal of Environmental Quality.50, 78-93. Tian, L., Cai, Y. and Akiyama, H., 2019. A review of indirect N2O emission factors from agricultural nitrogen leaching and runoff to update of the default IPCC values. Environmental Pollution. 245, 300-306. Tian J., Dungait J., Lu X., Yang Y., Hartley I., Zhang W., Mo J., Yu G., Zhou J., Kuzyakov Y., 2019. Long-term nitrogen addition modifies microbial composition and functions for slower carbon cycling and sequestration in tropical forest soil. Global Change Biology. 25, 3267-3281. Toosi, E., Kravchenko, A., Guber, A. and Rivers, M., 2017. Pore characteristics regulate priming and fate of carbon from plant residue. Soil Biology and Biochemistry. 113, 219-230. 146 Trinsoutrot, I., Recous, S., Mary, B. and Nicolardot, B., 2000. C and N fluxes of decomposing 13C and 15N Brassica napus L.: effects of residue composition and N content. Soil Biology and Biochemistry. 32, 1717-1730. van der Weerden, T.J., Kelliher, F.M. and de Klein, C.A., 2012. Influence of pore size distribution and soil water content on nitrous oxide emissions. Soil Research. 50, 125-135. Vance, E.D., Brookes, P.C. and Jenkinson, D.S., 1987. An extraction method for measuring soil microbial biomass C. Soil Biology and Biochemistry. 19, 703-707. Vedere, C., Gonod, L.V., Pouteau, V., Girardin, C. and Chenu, C., 2020. Spatial and temporal evolution of detritusphere hotspots at different soil moistures. Soil Biology and Biochemistry. 150, 107975. Velthof, G.L., Kuikman, P.J. and Oenema, O., 2002. Nitrous oxide emission from soils amended with crop residues. Nutrient cycling in agroecosystems, 62(3): 249-261. Wallenstein, M.D. and Burns, R.G., 2011. Ecology of Extracellular Enzyme Activities and Organic Matter Degradation in Soil: A Complex Community‐Driven Process. Methods of Soil Enzymology. 9, 35-55. Wang, S., Sanford, G.R., Robertson, G.P., Jackson, R.D. and Thelen, K.D., 2019. Perennial bioenergy crop yield and quality response to nitrogen fertilization. BioEnergy Research. 1-10. Wang, X., Tang, C., Severi, J., Butterly, C.R. and Baldock, J.A., 2016. Rhizosphere priming effect on soil organic carbon decomposition under plant species differing in soil acidification and root exudation. New Phytologist. 211, 864-873. Weitz, A.M., Linder, E., Frolking, S., Crill, P. and Keller, M., 2001. N2O emissions from humid tropical agricultural soils: effects of soil moisture, texture and nitrogen availability. Soil Biology and Biochemistry. 33, 1077-1093. Witzgall, K., Vidal, A., Schubert, D.I., Höschen, C., Schweizer, S.A., Buegger, F., Pouteau, V., Chenu, C. and Mueller, C.W., 2021. Particulate organic matter as a functional soil component for persistent soil organic carbon. Nature Communications. 12, 1-10. Wu D., Senbayram M., Zang H., Ugurlar F., Aydemir S., Brueggemann N., Kuzyakov Y., Bol R., Blagodatskaya E., 2018. Effect of biochar origin and soil pH on greenhouse gas emissions from sandy and clay soils. Applied Soil Ecology. 129, 121-127. Wu, J., Brookes, P. and Jenkinson, D., 1993. Formation and destruction of microbial biomass during the decomposition of glucose and ryegrass in soil. Soil Biology and Biochemistry. 25, 1435- 1441. Wu, Q., Zhang, C., Yu, Z., Zhang, J., Zhu, C., Zhao, Z., Xiong, J. and Chen, J., 2018. Effects of elevated CO 2 and nitrogen addition on organic carbon and aggregates in soil planted with different rice cultivars. Plant and Soil. 432, 245-258. 147 Xia, L., Lam, S.K., Wolf, B., Kiese, R., Chen, D. and Butterbach‐Bahl, K., 2018. Trade‐offs between soil carbon sequestration and reactive nitrogen losses under straw return in global agroecosystems. Global Change Biology. 24, 5919-5932. Xu, J., Liu, S., Song, S., Guo, H., Tang, J., Yong, J.W., Ma, Y. and Chen, X., 2018. Arbuscular mycorrhizal fungi influence decomposition and the associated soil microbial community under different soil phosphorus availability. Soil Biology and Biochemistry. 120, 181-190. Xu, Y., Ding, F., Gao, X., Wang, Y., Li, M. and Wang, J., 2019. Mineralization of plant residues and native soil carbon as affected by soil fertility and residue type. Journal of Soils and Sediments. 19, 1407-1415. Yanai, Y. and Toyota, K., 2006. Effects of soil freeze-thaw cycles on microbial biomass and organic matter decomposition, nitrification and denitrification potential of soils, Proc. Int. Symp. of JSPS Core to Core Program between Hokkaido University and Martin Luther University Symptom of Environmental Change in Siberian Permafrost Region, Halle-Wittenberg, Eds. Hatano R and Guggenberger G, p 177-191, Hokkaido University Press, Sapporo. Zaman, M., Nguyen, M.L., Šimek, M., Nawaz, S., Khan, M.J., Babar, M.N. and Zaman, S., 2012. Emissions of nitrous oxide (N2O) and di-nitrogen (N2) from the agricultural landscapes, sources, sinks, and factors affecting N2O and N2 ratios. Greenhouse gases-emission, measurement and management. (Ed. Guoxiang Liu). 1-32. Zhang, Z., Wang, W., Qi, J., Zhang, H., Tao, F. and Zhang, R., 2019. Priming effects of soil organic matter decomposition with addition of different carbon substrates. Journal of Soils and Sediments. 19, 1171-1178. 148 CHAPTER 5: Conclusion Soil is a critical terrestrial compartment acting as sources and sinks for GHG emissions. Detritusphere is part of the lithosphere of high environmental significance - C and N derived from the detritus are incorporated into the soil, and vigorous microbial activity stimulated by the detritus leads to the emission of GHGs. Understanding mechanisms of microbial processes taking place in detritusphere can thus provide strategies to mitigate GHG emissions and accrue C and N stocks. Microbial processes and activities in detritusphere are a function of soil environmental conditions there, which can be distinctly different from those of the bulk soil. Still, characteristics and developments of those micro-environments have yet to be thoroughly investigated. This study sought to elucidate the micro-environmental characteristics in plant detritusphere and understand the effect of those micro-environmental conditions on C and N dynamics. I applied experimental techniques including X-ray CT, zymography, stable isotope labeling, and in-situ growing of roots to explore detritusphere micro-environments. X-ray CT allowed a quantitative measurement of soil moisture and decomposition, while zymography allowed spatial and temporal evaluation of extracellular enzyme activity. Stable isotope (13C and 15 N) labeling enabled tracking of plant-derived C and N in the atmosphere, soil, and microbial biomass. Using in-situ grown roots for decomposition allowed the comprehensive investigation of the detritusphere inherited from the rhizosphere. The findings of this study improved the understanding of temporal and spatial changes in micro-environments near decomposing plant residues and demonstrated their potential influences on overall GHG emissions and their variability. The study showed that the detritusphere micro- environment is characterized as 1.5 – 2 times high moisture contents compared to the bulk soil due to moisture absorption by plant residues. It revealed that the moisture gradient is formulated within 149 ~150 µm from the decomposing residue, and slope of the gradient depends on the pore characteristics and moisture contents of the surrounding soil. This study also found that the detritusphere micro-environment has low O2 content and high extracellular enzyme activity due to moisture absorption by plant materials. Residue type greatly influenced these micro-environmental conditions. Leaf materials led to greater O2 depletion compared to root materials 48 hours after the water is added into the system. It is partly attributed to the greater water absorption by leaves rather than roots, and also higher content of dissolved organic matters from the leaves. High moisture content, low O2 content, and high extracellular enzyme activity in detritusphere led to faster decomposition and greater N2O emission. This study also revealed that soil pore characteristics and root growth (rhizosphere legacy) together determine the intensity of hotspot microbial activity, priming effects, and GHG emissions. In contrary to the traditional knowledge that fine-textured soil is more favorable to denitrification and N2O emission than coarse-textured soil, coarse soil materials could emit more N2O when the plant roots grow and decompose at the same place. Complex additive effect of fungal activity, root exudates, and water absorption were discussed in main chapters. Finally, this study revealed that the plant-derived and soil-derived N2O emission occur at the same location – detritusphere, not the bulk soil. Micro-environmental conditions favorable to microbial activity promotes not only the process of plant materials, but also native soil organic matters. My dissertation contributed to the characterization of micro-environmental conditions in detritusphere, and their relevance to C and N cycling. It stresses the importance of hotspot micro- environments in predicting GHG emission and related microbial processes and urges further research to understand the full mechanism and incorporate those in GHG prediction models. 150