TREE SEEDLING MYCORRHIZAL TYPE AND FUNCTIONAL TRAITS INTERACT WITH LIGHT AVAILABILITY TO MEDIATE PLANT-SOIL FEEDBACKS By Katherine Elizabeth Anne Wood A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Forestry – Doctor of Philosophy Ecology, Evolutionary Biology and Behavior – Dual Major 2023 ABSTRACT The seedling recruitment phase is a major demographic bottleneck and is critical for future forest community dynamics. Plant-soil feedbacks (PSFs) are often considered to be key drivers of seedling recruitment. PSFs are a continuous feedback loop in which adults modify properties of the soil beneath their crown, thereby influencing the ability of seedlings to grow and survive in that soil. Mechanisms underlying the strength and direction of PSFs include soil- borne microbes, seedling functional traits that confer defense against or recovery from microbes, and matching/mismatching of mycorrhizal type between juvenile and adult trees. Additionally, the strength and direction of PSFs may shift with light availability, which can modify both microbial abundance and functional traits. In this dissertation, I investigated the role each of these mechanisms and their interactions on tree seedlings PSFs. In Chapter 2, I investigated how shade tolerance may be shaped by, not only responses to light availability, but also by species’ defense and recovery functional traits, soil microbial communities, and interactions of these factors with light availability. I found that shade tolerance may be explained by interactions among soil-borne microbes, seedling functional traits, and light availability, providing a more mechanistic and trait-based explanation of shade tolerance and thus forest community dynamics. In Chapter 3, to determine the extent to which functional traits mediate PSFs via seedling survival, I conducted a field experiment in which I planted seedlings of four temperate tree species across a gradient of light availability and into soil cores collected beneath conspecific (sterilized and live) and heterospecific adults. Results from this chapter indicate that functional trait values in seedlings as young as three weeks vary in response to both soil source and light availability. Furthermore, traits play an important role in mediating effects of local soil sources and light on seedling survivorship, and thus plant traits could play an important role in PSFs. In Chapter 4, to assess the role of mycorrhizal type matching on juvenile trees’ defense/recovery trait response and PSFs, I carried out a greenhouse experiment where I grew seedlings of five temperate tree species under soils cultured by adults of the same species and under three light levels. I found that AM seedlings experience lower survival in soils cultured by AM adults and EM seedlings experience higher survival in soils cultured by EM adults. Additionally, as differences in mycorrhizal colonization and defense/recovery traits between conspecific and heterospecific soils decrease, PSFs are effectively neutralized, providing new insights into how mismatching of mycorrhizal type interacts with traits to influence PSFs. In Chapter 5, to investigate the potential trade-offs between PSFsurvival at low light and PSFbiomass at high light availability, I evaluated biomass data from the parallel factorial blocked field (Chapter 3) and greenhouse (Chapter 4) experiments. AM seedlings experienced negative PSFbiomass that shifted to positive with increasing light availability, and EM seedlings always experienced positive PSFbiomass, irrespective of light level. In addition, I found that measuring PSFsurvival may be more important than PSFbiomass when studying species sensitive to soil-borne microbes and that are expected to grow in low light-environments. Together, these results provide a more mechanistic understanding to the factors underlying PSFs. Tree seedling mycorrhizal type and functional traits appear to interact with light availability to mediate PSFs, thereby influencing seedling regeneration dynamics and subsequent forest community dynamics. ACKNOWLEDGEMENTS I would like to thank my dissertation advisors and committee. Rich Kobe and Sarah McCarthy-Neumann, thank you for believing in me and providing frequent guidance, helping me wrangle complex ideas into more clear stories. I would not have been able to make it over the finish line without your support. Mike Walters, David Rothstein, and Andy Jarosz, thank you for your insights into how my research can apply to broader field observations and natural history. You’ve helped provide much needed context that will help communicate my research to broader audiences. Thank you also to the numerous other researchers and staff who helped me with my projects. Andy vanderYacht, thank you for so many great conversations and your support of my interests in integrating fire ecology and plant-soil feedbacks. To Aki Koyama, thank you for the insightful conversations about mycorrhizae and your near-encyclopedic knowledge of the current literature. To Mark Bunce, Paul Bloese, and Randy Klevickas, thank you for helping me navigate the field and greenhouse, and juggling several projects across multiple locations. To Inés Ibáñez, thank you for your statistics advise and helping me with manuscript revisions. Also, thank you to the many undergraduate research assistants who put in so many long hours and are the reason I have data to present at all. Thank you to the scientists and teachers who have fostered my love of forest ecology along the way. Thank you, Sharon Strauss and Brian Anacker, for providing me with my first research experiences and inspiring me to continue into graduate school. Thank you Marcel Rejmánek for guiding me during my senior thesis and showing me how fun field research can be. Thank you, Walt Carson, for providing me with so many insights into the world of research and broadening my perspectives. Thank you to the Institute for Tropical Ecology and Conservation iv for providing me with my first experiences in tropical rainforests. Thank you to the Organization for Tropical Studies for taking me around Costa Rica to learn about tropical ecology. A very warm thank you to my friends, both in-person and online. To my labmate, Clarice Esch, thank you for being my bother buddy, for providing me with hours of company, and being my shoulder to lean on whenever graduate school was difficult. To Michelle Spicer, for being my graduate school application buddy, and continuing to support me along the way. To Joseph Pini, thank you for being there, through thick and thin, and always believing in me, no matter the obstacles. To Mo, thank you for the spontaneous hang out sessions and listening to me unconditionally. To Matt, Ghrym, Aardon, Mango, Pharma, and all the other teammates I met online, thank you for providing me hours of happiness, even when I lived across the country from you. And finally, the biggest thank you to my mom, Sharman Wood. Thank you for encouraging me from the very beginning. Thank you for cheering me on, no matter what my interest or passion has been. Thank you for providing me with every opportunity to thrive in whatever career or hobby I choose. Thank you for being there, no matter the time or day, to listen to me talk about field work, data analysis, and publication woes. Thank you for waking up at 4am to watch one more practice run of my defense (and then watching it all over again 1 hour later), so I would feel confident. You have been my rock, my constant, throughout all of this. I love you. v TABLE OF CONTENTS CHAPTER 1 Introduction ............................................................................................................. 1 CHAPTER 2 Tree seedling shade tolerance arises from interactions between light availability and soil-borne microbes and is mediated by functional traits ........................................................ 8 CHAPTER 3 Tree seedling functional traits mediate plant-soil feedback survival responses across a gradient of light availability ........................................................................................... 29 CHAPTER 4 Tree seedling responses to plant-soil feedback depend on mycorrhizal type of the adults culturing the soil .......................................................................................................... 58 CHAPTER 5 Mycorrhizal type and light availability explain differences in biomass response to plant-soil feedback ................................................................................................................... 92 CHAPTER 6 Conclusion ........................................................................................................... 117 BIBLIOGRAPHY ...................................................................................................................... 123 APPENDIX ................................................................................................................................ 140 vi CHAPTER 1 Introduction The seedling recruitment phase is a major demographic bottleneck and is critical for future forest community dynamics (Gurevitch et al., 2020). Plant-soil feedbacks (PSFs) are often considered to be key drivers of seedling recruitment (Crawford et al., 2019; Putten et al., 2016). PSFs are a continuous feedback loop in which adults modify properties of the soil beneath their crown, thereby influencing the ability of seedlings to grow and survive in that soil (Bever et al., 1997). Mechanisms underlying the strength and direction of PSFs include soil-borne microbes (Bever et al., 2010; Jiang et al., 2020), seedling functional traits that confer defense against or recovery from microbes (Cortois et al., 2016; Xi et al., 2021), and matching/mismatching of mycorrhizal type between juvenile and adult trees (Chen et al., 2019; Kadowaki et al., 2018). Additionally, the strength and direction of PSFs may shift with light availability, which can modify both microbial abundance and functional traits. In this dissertation, I investigated the role of each of these mechanisms and their interactions on tree seedlings PSFs. The putative agents of PSFs are soil-borne microbes, namely pathogens and mycorrhizal fungi (Bever et al., 2010; Jiang et al., 2020). Soil-borne pathogens (including fungi, oomycetes, and bacteria) can cause high seedling mortality (Song & Corlett, 2022), especially in shade (McCarthy-Neumann & Ibáñez, 2013; McCarthy-Neumann & Kobe, 2008; O’Hanlon-Manners & Kotanen, 2004), where wetter conditions enhance microbe reproduction and dispersal (Hersh et al., 2012; Y. Liu & He, 2019; Reinhart et al., 2010). Mycorrhizal fungi are typically thought of as mutualists, providing water and nutrients in exchange for sugars (S. Smith & Read, 2008; Wipf et al., 2019). Mycorrhizal fungi are more abundant in high light availability (Bereau et al., 1 2000; Koorem et al., 2017; Shi et al., 2014). However, at low light availability, where photosynthate production is more limited, they may act parasitically (Konvalinková & Jansa, 2016; McCarthy-Neumann & Ibáñez, 2013). Interactions between soil-borne pathogens and different groups of mycorrhizal fungi can shift the strength and direction of PSFs. At low light availability, the cost of maintaining the mycorrhizal symbiosis may exacerbate the negative effects of pathogens. Mycorrhizae can also confer protection against pathogens, but the degree of protection depends upon mycorrhizal type (Bennett et al., 2017) and may depend upon resource availability (McCarthy-Neumann and Ibáñez 2013). AMF can provide indirect defense against pathogens by competing for space on plant roots (Borowicz, 2001) and EMF can provide direct defense by forming a protective physical sheath on young roots (Laliberté et al., 2015). Also, both AMF and EMF can increase their host plant’s resource acquisition (Liang et al., 2015; Sikes, 2010). Moreover, matching or mismatching of mycorrhizal type (AMF or EMF) between the seedling growing in and the adult tree culturing the soil may influence PSFs. (Here, we refer to species that typically associate with AMF and EMF as “AM species” and “EM species”, respectively). Whereas AM trees typically experience negative PSFs (i.e., inhibition of seedlings around conspecific adults), EM trees more often experience positive PSFs (i.e., facilitation around conspecific adults) (Bennett et al., 2017; Kadowaki et al., 2018). In addition, AM trees have a higher abundance of plant pathogens in their soil (Eagar et al., 2022, 2023) and AM seedlings accumulate soil-borne pathogens faster when growing under AM adults (Chen et al., 2019). However, when there is mismatching of mycorrhizal type (e.g., AM seedlings growing beneath EM trees, and vice-versa), both AM and EM seedlings experience positive or neutral PSFs (Kadowaki et al., 2018). 2 Differences in PSFs may be partially explained by seedling defense and recovery functional traits, which are defined as measurable morphological or physiological attributes affecting plant performance (Violle et al., 2007). Plant functional traits could influence PSFs and vice-versa (P. Ke et al., 2015; Kuťáková et al., 2018; Xi et al., 2021). Functional traits that influence plant defense against and recovery from attack by soil-borne microbes include phenolics, lignin, and nonstructural carbohydrates (NSC). Phenolics and lignin can serve as chemical (Ichihara & Yamaji, 2009) and physical (Augspurger, 1990) defenses against soil- borne microbes. NSC can be mobilized to repair damaged tissues (Dietze et al., 2014). Phenolics, lignin, and NSC are likely affected by soil source (e.g., conspecific versus heterospecific soil). Phenolics production can be induced by mycorrhizal colonization (Wallis & Galarneau, 2020) and potentially by fungal pathogens (Witzell & Martín, 2008). Therefore, phenolics production should subsequently be higher in conspecific soils, where there should be higher colonization by mycorrhizal fungi and infection by effectively-specialized soil-borne pathogens (Benítez et al., 2013; Hersh et al., 2012). It is unclear whether lignin production is influenced by conspecific soils. However, it could be driven by soil nutrient availability, which can be impacted by microbes (J. Li et al., 2020; Luo et al., 2022). NSC should be lower in conspecific soils, due to greater resource allocation to symbionts (Schiestl-Aalto et al., 2019) and recovery against pathogen infection (Martínez-Vilalta, 2014; Saffell et al., 2014). Both PSFs and defense/recovery traits can shift across environmental gradients like light availability (McCarthy-Neumann & Ibáñez, 2013; Smith-Ramesh & Reynolds, 2017). However, most studies have not integrated abiotic factors when evaluating both traits and PSFs (Cortois et al., 2016). Shifts in light can change microbial composition and abundance (Koorem et al., 2017; Y. Liu & He, 2019), which may alter seedlings’ ability to defend against or recover from disease. 3 AMF are more abundant in higher light (Bereau et al., 2000; Koorem et al., 2017; Shi et al., 2014); however, in low light they can act parasitically and thereby decrease seedling survival (Konvalinková & Jansa, 2016). Higher mortality from pathogens typically occurs in low light (McCarthy-Neumann & Ibáñez, 2013; McCarthy-Neumann & Kobe, 2010a), where wetter and cooler conditions enhance microbe reproduction and dispersal (Hersh et al., 2012; Y. Liu & He, 2019; Reinhart et al., 2010). Light availability can also modify trait levels, including reduced production of phenolics (Ichihara & Yamaji, 2009) and lignin (Falcioni et al., 2018; Rogers et al., 2005). Additionally, carbon limitation in shade and lower stored nonstructural carbohydrates (NSC) may constrain recovery from disease (Kobe, 1997; Kobe et al., 2010). There is an increased need to understand the mechanisms underlying forest community dynamics, especially those that regulate the coexistence of tree species and promote species diversity, like PSFs. Negative PSFs (lower performance in conspecific versus heterospecific soils) may have positive effects on forest community diversity by increasing the likelihood that a seedling of a different species will replace an adult tree when it dies. Conversely, positive PSFs (higher performance in conspecific than heterospecific soils) may decrease community diversity by increasing the likelihood that an adult tree is replaced by a seedling of the same species. Together, negative and positive PSFs can mediate species coexistence within forests. However, how defense and recovery traits, in addition to mismatching of mycorrhizal type between juvenile and adult trees, mediate PSFs is relatively unknown. Furthermore, it is unclear how theses relationship might shift across abiotic gradients, such as light availability. This dissertation includes four chapters focusing on different aspects of tree seedling PSFs, including separating roles of mycorrhizal fungi versus soil-borne pathogens, defense and recovery functional traits (phenolics, lignin, and NSC), and mismatching of mycorrhizal type 4 between juveniles and adult trees, all in the context of light availability. A short summary of each chapter follows. Chapter 2: I investigated how shade tolerance may be shaped by, not only responses to light availability, but also by species’ defense and recovery functional traits, soil microbial communities, and interactions of these factors with light availability. I conducted a greenhouse experiment, controlling for AMF and soil-borne pathogen presence/absence and light availability, and measuring defense/recovery traits, for three temperate tree species from the genus Acer that vary in shade tolerance. I found that persistence of tree seedlings under low light availability, which we often interpret as shade tolerance, is not due to light limitation alone, but rather is due to interactions between low light availability and soil-borne microbes. Differences in seedling survival between low and high light only occurred when microbes were present. AMF colonization, phenolics, and NSC generally increased with light availability. Measured amounts of phenolics also were higher when pathogens were present, signifying that phenolics may be an induced defense response. Furthermore, across species, microbe treatment, and light availability, survival increased as phenolics and NSC increased. These results suggest that shade tolerance may be explained by interactions among soil-borne microbes, seedling defense and recovery functional traits, and light availability, providing a more mechanistic and trait-based explanation of shade tolerance and thus forest community dynamics. Chapter 3: To determine the extent to which defense and recovery functional traits mediate PSFs via seedling survival, I conducted a field experiment in which I planted seedlings of four temperate tree species across a gradient of light availability and into soil cores collected beneath conspecific (sterilized and live) and heterospecific adults. I monitored seedling survival twice per week over one growing season, and randomly selected subsets of seedlings to measure 5 mycorrhizal colonization, phenolics, lignin, and NSC levels at three weeks. Results from this study demonstrate that defense and recovery functional trait values in seedlings as young as three weeks vary in response to both soil source and light availability. In general, I found higher measured amounts of mycorrhizal colonization and defense/recovery traits in conspecific than heterospecific soils and in higher light availability. Moreover, seedling survivorship was associated with AMF colonization and phenolics for two species. These results suggest that seedling traits could have an important role in mediating the effects of local soil source and light levels on seedling survivorship, and thus plant traits could play an important role in PSFs. Chapter 4: To assess the role of mycorrhizal type matching on juvenile trees’ trait response and PSFs, I carried out a greenhouse experiment where I grew seedlings of five temperate tree species under soils cultured by adults of the same species and under three light levels. After 12 weeks, I quantified seedling survival, colonization by mycorrhizal fungi (AMF and EMF), and measured their defense and recovery traits (phenolics, lignin, and NSC). I found that negative PSFs experienced by seedlings associating with AMF almost always occurred when they were compared with heterospecific adults associating with EMF. Conversely, positive PSF experienced by EM seedlings occurred when compared to soils cultured by AM adults. For both AM and EM species, the magnitude of effect for PSFs was greatest at low light. Furthermore, soil microbes from conspecific-cultured soils reduced survival for AM species but had no effect on survival for EM species. PSFs for AM seedlings became less negative as percent AMF colonization and defense/recovery traits increased, and PSFs for EM seedlings became less positive as percent AMF colonization and lignin increased. These results suggest that functional traits and increased colonization by mycorrhizal fungi effectively neutralize both negative and 6 positive PSFs, providing new insights into how mismatching of mycorrhizal type interacts with traits to influence PSFs, and thus forest community dynamics. Chapter 5: To investigate the potential trade-offs between PSFsurvival at low light and PSFbiomass at high light availability, I evaluated biomass data from the parallel factorial blocked field (Chapter 3) and greenhouse (Chapter 4) experiments. I found that PSFbiomass was typically negative for AM species and positive for EM species, but these results did not depend upon mycorrhizal matching/mismatching of the juvenile and adult tree. Furthermore, PSFbiomass became more neutral (i.e., less negative) for the AM species as light availability increased. There was also a negative relationship between PSFsurvival in low light availability and PSFbiomass at high light availability. At high light availability, all species experienced positive PSFbiomass, in contrast to low light where AM species experienced negative PSFsurvival and EM species experienced positive/neutral PSFbiomass. This research suggests that measuring PSFsurvival may be more important when studying species sensitive to soil-borne microbes and that are expected to grow in low light-environments. Conversely, measuring PSFbiomass may be more beneficial when seedlings do not experience high mortality. Together, these results help elucidate the mechanisms underlying variation in PSF studies. The final chapter of this dissertation is a synthesis of the findings of the four research chapters, in addition to a discussion of how these results contribute to the scientific field of forest ecology. I demonstratee how soil-borne microbes can mediate differences in shade tolerance. In addition, I show that mismatching of mycorrhizal type and defense/recovery traits may help explain forest regeneration dynamics. Together, my results provide a more mechanistic understanding of seedling recruitment patterns in the context of PSFs, mycorrhizal type, defense/recovery functional traits, and light availability. 7 CHAPTER 2 Tree seedling shade tolerance arises from interactions between light availability and soil-borne microbes and is mediated by functional traits ABSTRACT Shade tolerance is a central concept in forest ecology and strongly influences forest community dynamics. However, the plant traits and conditions conferring shade tolerance are yet to be resolved. I propose that shade tolerance is shaped not only by responses to light but also by a species’ defense and recovery functional traits, soil microbial communities, and interactions of these factors with light availability. I conducted a greenhouse experiment for three temperate species in the genus Acer that vary in shade tolerance. I grew newly germinated seedlings in two light levels (2% and 30% sun) and controlled additions of microbial filtrates using a wet-sieving technique. Microbial filtrate treatments included: <20 µm, likely dominated by pathogenic microbes; 40-250 µm, containing arbuscular mycorrhizal fungi (AMF); combination, including both filtrate sizes; and sterilized combination. I monitored survival for nine weeks and measured fine root AMF colonization, hypocotyl phenolics, stem lignin, and stem+root nonstructural carbohydrates (NSC) at three- week intervals. I found that differences in seedling survival between low and high light only occurred when microbes were present. AMF colonization, phenolics, and NSC generally increased with light. Phenolics were greater with <20 µm microbial filtrate, suggesting that soil-borne pathogens may induce phenolic production; and NSC was greater with 40-250 µm filtrate, suggesting that mycorrhizal fungi may induce NSC production. Across species, microbe 8 treatments, and light availability, survival increased as phenolics and NSC increased. Therefore, shade tolerance may be explained by interactions among soil-borne microbes, seedling traits, and light availability, providing a more mechanistic and trait-based explanation of shade tolerance and thus forest community dynamics. INTRODUCTION Tree seedling mortality responses to understory light availability are an important filter of mature tree species composition and drivers of forest community dynamics (Pacala et al., 1996). Thus, plant survival in low light or shade tolerance (Shirley, 1943) is a central concept in forest ecology. The seedling establishment phase, a major demographic bottleneck, is also critical for future community dynamics (Gurevitch et al., 2020). However, seedling responses to shade are far more complex than responses to light availability alone (Valladares et al., 2016; Valladares & Niinemets, 2008). Understanding the functional traits and biotic and abiotic conditions that convey shade tolerance are key to a more mechanistic understanding of forest community dynamics. A plant’s ability to tolerate low light conditions can be moderated by soil-borne microbes, like soil-borne pathogens and mycorrhizal fungi (Jiang et al., 2020). Soil-borne pathogens (including fungi, oomycetes, and bacteria) can cause high seedling mortality (Song & Corlett, 2022), especially in shade (McCarthy-Neumann & Ibáñez, 2013; McCarthy-Neumann & Kobe, 2008; O’Hanlon-Manners & Kotanen, 2004), where wetter conditions enhance microbe reproduction and dispersal (Hersh et al., 2012; Y. Liu & He, 2019; Reinhart et al., 2010). Arbuscular mycorrhizal fungi (AMF) can provide water and nutrients in exchange for sugars (Wipf et al., 2019). However, they may parasitize tree seedlings and increase mortality in low 9 light (Ibáñez & McCarthy-Neumann, 2016; Konvalinková & Jansa, 2016), despite mutualistic tendencies and greater abundance in high light (Bereau et al., 2000; Koorem et al., 2017; Shi et al., 2014). AMF, pathogens, and light levels can interact to influence seedling mortality. In high light, AMF root colonization is greater (Ibáñez & McCarthy-Neumann, 2016; Konvalinková & Jansa, 2016; Koorem et al., 2017), which can reduce the growth of fungal pathogens, potentially by competing for root space (Borowicz, 2001). AMF may also indirectly ameliorate pathogen effects (Liang et al., 2015), by providing water and nutrients to the host plant (Graham, 2001) and inducing production of defensive traits (Pozo & Azcón-Aguilar, 2007; Zamioudis & Pieterse, 2012) that protect against pathogens (Azcón-Aguilar et al., 2002; Violle et al., 2012). Conversely, in low light, seedling mortality may increase, due to the combined carbon costs of maintaining the AMF mutualism and recovery from pathogen attack. Seedlings could also experience higher mortality from soil-borne microbes due to shade- induced changes in defensive functional trait values, such as reduced phenolics (Ichihara & Yamaji, 2009) and lignin (Falcioni et al., 2018; Rogers et al., 2005). Additionally, carbon limitation in shade and lower stored nonstructural carbohydrates (NSC) may constrain recovery from disease (Kobe, 1997; Kobe et al., 2010). Functional trait values may not only differ within a species based on light availability, but may also vary among species in relation to shade tolerance (Imaji & Seiwa, 2010). Shade tolerant species are typically less vulnerable to mortality by soil-borne microbes than shade intolerant species (Alvarez-Clare & Kitajima, 2007; Augspurger, 1984a; McCarthy-Neumann & Kobe, 2008, 2010a) at least partly because shade tolerant species allocate more carbon to chemical and physical defenses (Coley et al., 1985; 10 Coley & Barone, 1996), and recovery (Kitajima, 1994; Myers & Kitajima, 2007; Poorter et al., 2010). To examine the effects of light availability, soil-borne microbes, tree seedling functional traits, and their interactions on light-dependent seedling survival, I established an experiment to test the following hypotheses: I hypothesized that: 1) Within species, decreased survival under low versus high light only occurs in the presence of soil microbes. 2) Mycorrhizal colonization is lower in the <20 um microbial filtrate where pathogens are likely to be the dominant microbial group. 3) As defensive traits, phenolics and lignin are induced to higher levels in soils where microbes are present (non-sterilized) and in higher light availability. 4) As a recovery trait, NSC is lower in soils where microbes are present and in low light availability. 5) Across all species and when microbes are present, survival increases as phenolics, lignin, and NSC increase. MATERIALS AND METHODS I conducted a fully factorial blocked-design greenhouse experiment at the Michigan State University Tree Research Center in Lansing, Michigan, USA (42.7 ºN, 84.5 ºW). The experiment consisted of three species, four microbial communities (<20 µm filtrate, representing pathogenic microbes; 40-250 µm filtrate, representing AMF; combined filtrate (both <20 µm and 40-250 µm); and sterilized combined filtrate) and two light levels (2% and 30% full sun, representing shade and light gap environments). Individual pots were set up on six different benches (three 11 per light level), where all treatment combinations were represented. I planted 80 seedlings per treatment combination for a total of 1,920 seedlings. I monitored seedlings every three days for survival, and randomly selected subsets for trait measurements at three, six, and nine weeks. Species selection I selected three biogeographically widespread, co-occurring tree species within the genus Acer: saccharum, rubrum, and negundo. These species have similar seed sizes (Osunkoya et al., 1994), but vary in shade tolerance (Burns & Honkala, 1990a; Niinemets & Valladares, 2006a). Light availability I grew seedlings at two light levels (2% and 30% full sun). I created light treatments by covering six greenhouse benches (three per treatment) with an inner layer of black shade cloth and an outer layer of reflective knitted poly-aluminum shade cloth (BFG Supply, Burton, Ohio, USA). I confirmed light levels using PAR (photosynthetically active radiation) measurements at each bench with a LI-COR 250A quantum sensor (LI-COR, Lincoln, Nebraska, USA) on a uniformly overcast day. Soil collection and preparation of soil inocula Soils were collected from Alma College’s Ecological Field Station in Vestaburg, Michigan, USA (43.4 °N, 84.9 ºW), in a 100-ha mixed-hardwood forest stand containing a 3-ha subplot with mapped and tagged trees. In August 2016, I randomly selected three adult trees per species. I selected adults that were at least two crown diameters away from other study species to reduce potential cross-culturing of soil. I collected soil (top 15 cm) within 1 m of each focal tree stem, maintained as separate replicates throughout the remainder of the experiment (as recommended by Rinella and Reinhart 2018). I prepared soil by dicing roots and sifting soil through a 1 cm mesh sieve, retaining all roots that passed through the 1 cm sieve, as they may 12 harbor host-specific microbial communities. Soil samples were stored at 4 °C for up to 2 months before preparation of soil inocula filtrates. I created four microbial communities from sifted field soil using a wet-sieving method (Callaway et al., 2011; Klironomos, 2002; König et al., 2016; Liang et al., 2015; Pizano et al., 2014). For each extraction, I agitated 50 g of soil in a blender with 250 mL of deionized water at high speed for 60 sec, then passed the slurry through three analytical sieves (250-, 40-, and 20 μm) using a high-pressure water hose, for a total volume of 800 mL. To minimize contamination between treatments, I cleaned the sieves ultrasonically for 5 min between each extraction. The 250 µm sieve collected larger roots and coarser soil. I floated material retained by the 40 μm sieve on the surface of a 60% sucrose solution and centrifuged it at 688 g for 20 min. I collected material in the water and at the water-sucrose interface on 47 mm Whatman no.1 filter paper, surface sterilized it with 10% NaOCl for 10 sec, and washed it with distilled water under a filtration vacuum. I divided each filter paper into eight equal pieces. I collected the filtrate that passed through the 20 μm sieve and separated it into eight 100 mL containers. In sum, I created three microbial communities based on filtrate size classes: <20 µm, 40-250 µm, and combined (containing both <20 µm and 40-250 µm filtrates). To test for abiotic effects of conspecific cultured soils due to nutrients or allelopathy, I combined filtrates (<20- and 40-250 µm) and sterilized by steam autoclaves (at 121 °C for two hours). I also quantified and compared AMF colonization in the sterilized and <20 µm filtrates against a control containing only filter paper and deionized water to test the effectiveness of the sterilization for reducing microbes. I found no AMF colonization in soils treated with the sterilized (t = 2.97, df =176, p = 0.003) or <20 µm filtrates (t = 3.79, df = 178, p < 0.001), 13 relative to a distilled water control (Figure 2.2A). Filtrates were kept refrigerated at 4 °C for up to 48 hours, before adding them to the greenhouse pots. In the greenhouse, I filled 1,920 (655 cm3) deepots (Stuewe and Sons, Tangent, Oregon, USA) with sterilized commercial topsoil (Hammond Farms Landscape Supply, Lansing, Michigan, USA). To aid seedling germination, I also topped each pot with 2 cm of 85% sterilized commercial soil mix, containing peat moss, perlite, and vermiculite (Fafard 4P Mix, Sun Gro Horticulture, Agawan, Massachusetts, USA). In pilot trials of this experiment, I found over 50% seedling mortality in the first two weeks, without the addition of the commercial soil mix (personal observation). I steam sterilized the topsoil and soil mix by autoclaving twice at 121 °C for two hours, with a 48-hour incubation period between cycles. Within 48 hours of wet sieving, I added microbe treatments to soil, keeping filtrate from each adult and species separate. To enable microbial communities to sporulate and AMF hyphae to establish, I cultured the soil with Allium as bait plants, before planting Acer seedlings (Al- Yahya’ei et al., 2011; Klironomos et al., 1999). In January 2017, each pot was planted with three germinating Allium tuberosum seedlings. After two months, I removed aboveground A. tuberosum seedling biomass. One week after Allium removal, I planted Acer seedlings whose hypocotyls had emerged within the three previous days. I purchased seeds from Sheffield’s Seed Company (Locke, New York, USA). To minimize microbes from non-experimental soil sources, I surface sterilized the seeds with 0.6% NaOCl both prior to cold stratification and germinating in perlite. I watered the seedlings three times per week with 50 mL deionized water. 14 Survival and functional trait measurements I grew the Acer seedlings for nine weeks. I recorded emergence and survival every three days and assigned date of death as the first census with total leaf and stem tissue necrosis. I harvested a random subset of 20 seedlings for each species and each treatment at three, six, and nine weeks, to quantify AMF colonization, phenolics, lignin, and NSC. I chose these times for our sub-harvests, because in a previous greenhouse experiment, mortality curves for tree seedlings subjected to soil-borne pathogens often increased at week three and peaked between four to six weeks after germination (McCarthy-Neumann & Ibáñez, 2012). To quantify percent AMF colonization, I stained seedling roots with a 5% Shaeffer black ink in vinegar solution (Vierheilig et al., 1998) and counted fungal structures (e.g., arbuscules, coils, vesicles, hyphae) along 100 intersections under the microscope (McGonigle et al., 1990). To quantify phenolics, I analyzed hypocotyl samples, using a microplate-adapted colorimetric total phenolics assay with Folin-Ciocalteu reagent (Ainsworth & Gillespie, 2007). To quantify lignin, I analyzed root and stem samples with an Ankom 200 fiber analyzer (ANKOM Technologies, Macedon, NY, USA), using the acid-fiber detergent fiber filter technique. To quantify NSC, I analyzed stem samples, using a standardized enzyme method for sugar and starch extraction and quantification (Landhäusser, Chow, Dickman, et al., 2018). Statistical analyses To test the influence of light availability and microbial community on tree seedling survival, I used an individual based counting process in a Cox survival model (Burnham & Anderson, 2002; McCarthy-Neumann & Ibáñez, 2012). Data for each seedling 𝑖 and each time 𝑡, 𝑁!", were coded as 0 until the seedling was found dead, 𝑁!" = 1. I used a count process to model 15 the number of events (mortality, 𝑁!") until the experiment ended at nine weeks. I modeled the likelihood as: and the process as: 𝑁!" ~ 𝑃𝑜𝑖𝑠𝑠𝑜𝑛(λ!") λ𝑖𝑡 = ℎ"𝑒($"), where parameters were estimated as a function of the hazard (ℎ), which is the intrinsic rate of mortality due to individual age or time within the experiment), and of risk (𝜇), which is the extrinsic rate of mortality due to light availability and microbial community. Simulations (3 chains) were run until convergence of the parameters was ensured (25,000 iterations) and then run for another 50,000 iterations, from which the posterior parameter values (Figure A2.1, A2.2) and predicted survival (Figure A2.3) were estimated. Predicted survival values were used to assess whether there were differences in how species responded to microbe treatments and light. I then used predicted survival values and their associated uncertainty to test if there were differences in how species responded to low versus high light and in different microbe treatments. Differences that did not include zero in their 90% credible intervals were considered statistically significant (Kruschke, 2014). I used linear mixed effects models to evaluate the effects of light availability and microbial community on seedling traits. I ran individual models for each species and trait, where light level and microbial community were treated as fixed effects, and harvest time, bench (nested within light level), and adult tree were treated as random effects. I then used linear mixed effects models to assess the effects of traits on seedling survival in the three weeks following trait measurement (i.e., traits collected at three weeks were used to predict seedling survival from three to six weeks, and traits at six weeks were used to predict 16 seedling survival from six to nine weeks). I treated AMF colonization, phenolics, lignin, and NSC as fixed effects, and included light level as a covariate. Species was treated as a random effect. I evaluated relative support for each of our linear mixed effects models using multi- modal inference with corrected Akaike’s Information Criterion (Burnham & Anderson, 2002). Models with DAICc < 6 of the best-approximating model were considered plausible (Richards, 2008). I performed all analyses in R 3.5.1 (R Core Team, 2020). I used the rjags package (Plummer, 2019) to fit survival models and to run predicted survival and contrast simulations. I used the built-in “lmer” function to fit linear mixed effects models and tested significance of main effects using the “Anova” function in the car package (Fox & Weisberg, 2019). Model selection for linear mixed effects models was determined with the “step” function in the lmerTest package (Kuznetsova et al., 2017). Post-hoc Tukey pairwise comparisons of significant main effects were made using the “emmeans” and “joint_tests” functions in the multcomp package (Hothorn et al., 2008; Lenth, 2020). RESULTS Differences in seedling survival in low versus high light occurred only when microbes were present Differences in seedling survival in low light versus high light, a measure of shade tolerance, appeared only in the presence of soil biota (Figure 2.1A). In low light, survival decreased up to 14% with the <20 µm filtrate, 12% with the 40-250 µm filtrate, and 17% with the combined filtrate. A. negundo experienced the largest decreases in survival (17% reduction in 17 the combined filtrate), whereas A. saccharum had the lowest (11% reduction in the combined filtrate). Furthermore, A. saccharum survival decreased only in the presence of combined filtrate. Microbial treatments influenced seedling survival in only three of the nine cases in high light, but in eight of nine in low light (Figure 2.1B). At low light, the presence of soil biota relative to sterilized soil decreased survival 12-26% for all species and filtrates, except for A. saccharum with 40-250 µm filtrate. Additionally, negative soil biota effects were more common with the <20 µm filtrate, decreasing survival for all three species, regardless of light level. Figure 2.1 Difference in predicted survival (mean ± 90% credible interval) in A) low versus high light and B) relative to sterilized soil. For each species and microbial communities, at the end of 9-weeks. Negative values indicate decreased survival in treatment versus A) high light or B) sterilized soil. Statistically significant differences (90% CI do not overlap 0) are indicated with *. 18 Seedling functional trait values varied with light availability and microbial community Across all treatments, percent root colonization by AMF was 11-18% greater in high than low light for A. saccharum and A. negundo (Figure 2.2A). There was no AMF colonization in the <20 µm filtrate, and colonization was similar in the 40-250 µm and combined filtrates. Overall, A. negundo (76%) had the highest AMF colonization, compared to A. saccharum (66%) and A. rubrum (45%). Phenolic content (nmol Gallic acid equivalents per mg dry extract) increased with light availability across almost all microbe filtrate treatments (Figure 2.2B); the only exception was A. saccharum with the <20 µm filtrate. For A. rubrum and A. negundo, phenolic content was negligible in low light with the sterilized and combined filtrates. Overall, A. saccharum had the highest phenolic content (0.23 nmol/mg), compared to A. negundo (0.14 nmol/mg) and A. rubrum (0.06 nmol/mg). Percent dry mass lignin was greater in high than low light but depended upon species and microbial filtrate (Figure 2.2C). For A. saccharum, lignin increased 5% across microbe filtrates, in high versus low light. For A. negundo, lignin increased 27% but only in the 40-250 µm filtrate. Across species, A. saccharum had the highest percent dry mass lignin (12%), compared to A. rubrum (8%) and A. negundo (6%). Percent dry mass NSC was greater in high than low light for all species and was generally greatest with the 40-250 µm filtrate (Figure 2.2D). Additionally, for A. rubrum and A. negundo, NSC decreased 40-70% in the <20 µm and combined microbe filtrates, in low versus high light. Across species, A. saccharum had the highest dry mass NSC (12%), compared to A. negundo (11%) and A. rubrum (8%). 19 Figure 2.2 A) percent colonization AMF (%), B) phenolic content (nmol Gallic acid equivalents per mg dry extract), C) percent dry mass lignin (%), and D) percent dry mass NSC (%). Means within each panel not sharing a letter are statistically different by the Tukey test at alpha = 0.05. Best-fit model terms are overlayed on each panel. 20 Figure 2.2 (cont’d) 21 Seedling functional trait values are associated with survival Survival increased with both phenolics (c2 = 5.93, df = 1, p = 0.015; Figure 2.3A) and NSC (c2 = 7.72, df = 1, p = 0.005; Figure 2.3B). However, there was no significant relationship between light availability and survival (p > 0.05), either alone or interacting with phenolics and NSC. Figure 2.3 Percent seedling survival as a function of A) phenolic content (nmol Gallic acid equivalents per mg dry extract) and B) percent dry mass NSC (%). Each point represents a mean of trait values at a harvest time (3 or 6 weeks) and survival for seedlings in the 3 weeks after harvest. 22 DISCUSSION My results support that survival of newly germinated tree seedlings in low versus high light, or “shade tolerance,” may be due to interactions between low light and soil-borne microbes and be mediated by defense and recovery functional traits. Previous studies have demonstrated that seedling functional traits are influenced by light and microbes and that functional traits can influence growth and survival at low light (Falster et al., 2018), but have not linked resources, traits, and survival, as in this study. Across species, overall seedling survivorship and insensitivity to shading corresponded with shade tolerance categorizations. A. saccharum (shade tolerant) had the highest overall survival and was least sensitive to the microbial filtrates, compared to A. rubrum (intermediate) and A. negundo (intolerant). In this study, A. negundo had the highest AMF colonization of the three study species (Figure 2.2A) and had the largest decreases in low-light survival with added microbial filtrates (Figure 2.1). This aligns with prior studies which have shown that shade intolerant species experience greater mortality from disease in shade (Augspurger, 1984a; Pizano et al., 2014) and higher growth when grown in high light (Xi et al., 2023). Similarly, shade tolerant species showed no significant growth responses to microbial filtrate treatments, in contrast to pioneer species that were more sensitive to the habitat from which soil microbial filtrates were collected (Pizano et al. 2017). Soil-borne microbes explain variation in tree seedling survival responses to light There is good support that the 20 µm filtrate in our study is primarily composed of fungi and bacteria. We found no evidence of mycorrhizal colonization in our <20 µm soil filtrate (Figure 2.2), suggesting that this treatment is mainly composed of non-mycorrhizal fungi and bacteria (also consistent with Klironomos 2002). Additionally, there is support that the 40 µm 23 filtrate is primarily composed of the mycorrhizal community associated with soils cultured by conspecific adults. In a prior study that investigated this methodology, Wagg et al. (2014) found that soil passing through 250 µm sieves contained ~80% of the mycorrhizal community, and an additional ~20% of the mycorrhizal community passed through the 50 µm sieves. In the Wagg et al. (2014) study, filtrate <25 µm in size, effectively had no nematodes, <10% mycorrhizal fungi, ~70% other fungi and ~90% bacteria of the original soil community. In addition, several other studies have used these filtrate size classes to isolate and investigate the roles of soil-borne pathogens (McCarthy-Neumann and Kobe, 2008; König et al., 2016) and AMF (Klironomos, 2002; Callaway et al., 2011; Liang et al., 2015; Pizano et al., 2017). Differences in low versus high light survival appeared only when soil-borne microbes were present (Figure 2.1A), supporting hypothesis 1. With microbes present, survival decreased for all species, with the largest differences occurring in low light (Figure 2.2B). My results are consistent with previous research demonstrating that tree seedlings have higher mortality in shade and that the major cause of seedling death arises from disease (Augspurger, 1984a, 1984b; Vaartaja, 1962). My results are also consistent with Liang et al. (2015), who utilized the wet- sieving method and found that pathogens were associated with decreased biomass and survival, AMF were associated with increased biomass, and combined filtrate treatments canceled each other out for both biomass and survival; however, Liang et al. (2015) did not consider interactions with light availability. I did not see increases in seedling survival when AMF were present, despite high percent colonization of seedling roots. AMF might increase seedling survival due to overall higher resource availability and also through indirect defense against pathogens via competition for root space (Borowicz, 2001; Liang et al., 2015). This result was in contrast to the study by Liang et 24 al. (2015), in which they found higher seedling survival with AMF. However, benefits of AMF in this study may have manifested in growth (Gehring, 2003), which I did not measure. Young seedlings still relying on maternal seed reserves and high resource availability in the greenhouse may have diminished the importance of AMF (Forero et al., 2019; Heinze et al., 2020; Kulmatiski & Kardol, 2008). Moreover, AMF may have acted indirectly by enhancing production of phenolics and NSC, which were both positively correlated with AMF colonization (Figure A2.4). I also did not see any negative relationship between AMF colonization and NSC in low light, which may suggest that AMF act parasitically when photosynthates are limited. Amounts of functional traits varied with light availability and soil-borne microbes For all seedling species examined, phenolic content increased when microbes were present, supporting part of hypothesis 3. Both AMF (Pozo & Azcón-Aguilar, 2007; Vierheilig, 2004; Whipps, 2004) and pathogens (Nicholson & Hammerschmidt, 1992; Witzell & Martín, 2008) can induce phenolics production. Seedling phenolics also consistently increased with light availability (Figure 2.2B), further supporting hypothesis 3 and suggesting alleviation of photosynthate constraints on chemical defense production (Ballaré, 2014). Similarly, percent dry mass lignin increased with light availability and was highest in the 40-250 µm filtrate (Figure 2.2C), also supporting hypothesis 3. Higher lignin in the 40-250µm filtrate and lower lignin in the <20µm and combined filtrates for A. rubrum suggest that AMF have a positive effect while pathogens have a negative effect on lignin production. In partial support of hypothesis 4, I found that seedling NSC decreased when pathogens were present (Figure 2.2D), consistent with NSC reserves acting as a carbon buffer after damage (Gleason & Ares, 2004; Kobe et al., 2010; McPherson & Williams, 1998; Myers & Kitajima, 2007). However, contrary to predictions, NSC increased when AMF were present, but only when 25 not combined with the pathogen filtrate. The positive association between AMF colonization and NSC is consistent with Y.-L. Li et al. (2022). Functional traits were associated with greater seedling survival. Partly supporting hypothesis 5, I found that both phenolics (chemical defense) and NSC (carbon buffer precluding recovery from damage) had positive associations with tree seedling survival. These results are consistent with previous studies that have speculated higher allocation to defensive traits increases survival of shade tolerant seedlings in low light conditions (Alvarez- Clare & Kitajima, 2007; Augspurger, 1984a; Augspurger & Kelly, 1984; Kitajima, 1994; Vaartaja, 1962). I found no evidence of an association between lignin and survival. This was in contrast to previous studies that have posited that differences in lignin development impact seedling susceptibility to pathogens (Lee et al., 2019; Sattler & Funnell-Harris, 2013; Zhu et al., 2021). In this study, A. saccharum, the most shade tolerant species, had up to 50% greater lignin than the other two species (Figure 2.2D), but these differences in lignin values did not manifest in differences in survival. Caveats and future research There are several areas upon which future research could build on this work. I used sucrose-centrifugation to separate AMF spores from most other microbes and debris, and a bleach sterilization step to kill potential pathogens. While I was able to see high AMF colonization in the 40-250 µm and combined filtrates, and no colonization in the <20 µm or sterilized filtrates (Figure 2.2A), I cannot eliminate the possibility that I excluded some AMF in smaller filtrate sizes and included additional microbes and debris in the 40-250 µm filtrate. A more robust method would include further isolating pure AMF spores, as with Calloway (2011) 26 and Pizano et al. (2017). Alternatively, by adding a genetic analysis of the microbial inoculum, I could have determined more accurately which microbial groups were present, which would enhance our understanding of the results. I recommend that future research utilize the wet- sieving method in conjunction with spore isolation and/or genetic analyses. Likewise, by culturing the pots with Allium, I may have inadvertently increased AMF presence, disproportionate to pathogens, or increased the relative abundance of microbes that specialize with Allium, rather than Acer species. Although often thought of as generalists, AMF can show some host specificity (Kajihara et al., 2022; H. Yang et al., 2012), which could influence post-culturing microbial communities. Furthermore, by removing the aboveground biomass and leaving the belowground root structures intact, I may inadvertently changed nutrient dynamics within the pots. Thus, I recommend that future experiments culture AMF and other microbe communities with the host species of interest (in this case, Acer species). In future experiments, I recommend using the host species as a bait plant in the culturing step. While I investigated three species within a single genus to make broader generalizations, subsequent studies trying to generalize these results should include more species across additional levels of shade tolerance. Furthermore, although the three Acer species used in this study are similar in seed size, A. saccharum, the most shade tolerant species, has a larger relative seed size than A. negundo, the least shade tolerant species in this study. Thus, differences in seed size may have confounded effects of shade tolerance. Also, effects of light availability could be caused by changes in microclimate (e.g., soil temperature and moisture) and not directly due to irradiance. Similarly, effects of light may be mediated by photoreceptors and jasmonates, not just assimilate availability through higher photosynthetic rates (Ballaré & Austin, 2019; Pierik & Ballaré, 2021). 27 Additionally, I utilized shade cloth in the greenhouse to create shaded conditions, not vegetation shade. Future studies should consider teasing apart these mechanisms in the field, rather than the greenhouse, to provide more realistic seedling responses. Furthermore, although I was interested in light availability and shade tolerance, other environmental variables, such as nutrient or water availability, also could influence seedling survivorship (McCarthy-Neumann & Kobe, 2019). Implications for forest community dynamics This study provides a needed first step in developing a mechanistic understanding of how soil-borne microbes impact seedling shade tolerance, explained through functional traits. Although fast-growing shade intolerant species may be expected to outcompete shade tolerant species in high light (Pacala et al., 1996), shade intolerant species can be limited by the negative interactive effects of soil-borne microbes at low light (Y. Liu & He, 2019; McCarthy-Neumann & Kobe, 2010a), restricting their recruitment niche to areas with higher light and fewer soil- borne microbes. In this paper, I have demonstrated the importance of interactions between soil- borne microbes and light availability in determining tree seedling survival. Furthermore, I have related both intra- and interspecific differences in survival to functional traits, supporting a more trait-based and mechanistic approach to understanding forest community dynamics. A modified version of this chapter has been published in Frontiers in Ecology and Evolution. The original publication is available at https://www.frontiersin.org/articles/10.3389/fevo.2023.1224540/full. 28 CHAPTER 3 Tree seedling functional traits mediate plant-soil feedback survival responses across a gradient of light availability ABSTRACT Though not often examined together, both plant-soil feedbacks (PSFs) and functional traits have important influences on plant community dynamics and could interact. For example, seedling defense and recovery traits could impact seedling survivorship responses to soils cultured by conspecific versus heterospecific adults. Furthermore, levels of defense and recovery functional traits could vary with soil culturing source. In addition, these relationships might shift with light availability, which can affect trait values, microbe abundance, and whether mycorrhizal colonization is mutualistic or parasitic to seedlings. To determine the extent to which defense and recovery functional traits mediate PSFs via seedling survival, I conducted a field experiment. I planted seedlings of four temperate tree species across a gradient of light availability and into soil cores collected beneath conspecific (sterilized and live) and heterospecific adults. I monitored seedling survival twice per week over one growing season, and I randomly selected subsets of seedlings to measure seedling defense and recovery traits (i.e., mycorrhizal colonization and phenolics, lignin, and nonstructural carbohydrates) levels at three weeks. Though evidence for PSFs was limited, Acer saccharum seedlings exhibited positive PSFs (i.e., higher survival in conspecific than heterospecific soils). In addition, soil microbes had a negative effect on A. saccharum and Prunus serotina seedling survival, with reduced survival in live versus sterilized conspecific soil. In general, I found higher trait values (measured 29 amounts of a given trait) in conspecific than heterospecific soils and higher light availability. Additionally, A. saccharum survival increased with higher levels of phenolics, which were higher in conspecific soils and high light. Quercus alba survival decreased with higher colonization by arbuscular mycorrhizal fungi. I demonstrate that defense and recovery trait values (i.e., measured amounts of colonization by mycorrhizal fungi, phenolics, lignin, and NSC) in seedlings as young as three weeks vary in response to soil source and light availability. Moreover, seedling survivorship was associated with trait values for two species, despite both drought and heavy rainfall during the growing season that may have obscured survivorship-trait relationships. These results suggest that seedling traits could have an important role in mediating the effects of local soil source and light levels on seedling survivorship and thus plant traits could have an important role in PSFs. INTRODUCTION Though often examined separately, both plant-soil feedbacks (PSFs) and functional traits are important in plant community dynamics (Cadotte et al. 2015, Crawford et al. 2019, McGill et al. 2006, van der Putten et al. 2013, Yang et al. 2018). PSFs are a continuous feedback loop whereby plants modify properties of the soil they are growing in and influence the performance of future plants growing in that soil (Bever et al., 1997). These feedbacks subsequently affect community composition, which in turn influences soil properties, and so on. The net effect of interactions results in positive (better performance in conspecific soils), negative (better performance in heterospecific soils), or neutral PSFs. The putative agents of PSFs are soil-borne microbes, like mycorrhizae and pathogens (Bever et al., 2010; Jiang et al., 2020). Arbuscular mycorrhizal fungi (AMF) are often mutualistic, exchanging water and nutrients for photosynthates (Wipf et al., 2019). Soil-borne 30 pathogens, including fungi, oomycetes, and bacteria, can cause the death of entire seedling cohorts (Mangan et al., 2010; Terborgh, 2012), and pathogens with higher effective specialization are more abundant in conspecific soils (Benítez et al., 2013; Hersh et al., 2012). Mycorrhizal colonization is frequently higher in conspecific soils and in soils cultured by adult trees of the same mycorrhizal type (Bennett et al., 2017; Chen et al., 2019; Liang et al., 2016). These soils contain mycorrhizal genotypes that are well-suited to colonizing the adult trees growing in them (Segnitz et al., 2020; J. Yang et al., 2018). Functional traits are measurable morphological or physiological attributes affecting plant performance (Violle et al., 2007) that can translate into impacts on community dynamics. Despite the important role of plant survival in PSF (Comita et al., 2010; McCarthy-Neumann & Kobe, 2010a), traits promoting faster growth (e.g., specific leaf area, specific root length, height) have been the focus of most PSF studies (Baxendale et al., 2014; Cortois et al., 2016; Xi et al., 2021). Frequently, defensive traits are accounted for by assuming that species with fast growth rates have low investment in defense, and vice-versa (Cortois et al., 2016; Xi et al., 2021). However, tree seedling survivorship is likely to have greater effects on future community dynamics and composition than growth (Pacala et al., 1996). Thus, while little studied, functional traits that influence tree seedling survivorship in response to PSFs could be a crucial mechanism governing seedling and forest community dynamics. Plant functional traits could influence PSFs and vice-versa (P. Ke et al., 2015; Kuťáková et al., 2018; Xi et al., 2021). Functional traits that influence plant defense against and recovery from attack by soil-borne microbes include phenolics, lignin, and nonstructural carbohydrates (NSC). Phenolics and lignin can serve as chemical (Ichihara & Yamaji, 2009) and physical (Augspurger, 1990) defenses against soil-borne microbes. NSC can be mobilized to repair 31 damaged tissues (Dietze et al., 2014). Additionally, percent root colonization by mycorrhizae can be treated as a trait (Maherali, 2020) conferring defense against pathogens (Bennett et al., 2017). AMF can provide indirect defense against pathogens by competing for space on plant roots (Borowicz, 2001) and EMF can provide direct defense by forming a protective physical sheath on young roots (Laliberté et al., 2015). Also, both AMF and EMF can increase their host plant’s resource acquisition, which can be allocated to defensive and recovery traits (Liang et al., 2015; Sikes, 2010). Phenolics, lignin, and NSC are likely affected by soil source (e.g., conspecific versus heterospecific soil). Phenolics production can be induced by mycorrhizal colonization (Wallis & Galarneau, 2020) and potentially by fungal pathogens (Witzell & Martín, 2008). I expect that phenolics production should subsequently be higher in conspecific soils, where there should be higher colonization by mycorrhizal fungi and infection by effectively-specialized soil-borne pathogens (Benítez et al., 2013; Hersh et al., 2012). It is unclear whether lignin production is influenced by conspecific soils. However, lignin production could be driven by soil nutrient availability, which can be impacted by microbes (J. Li et al., 2020; Luo et al., 2022). NSC should be lower in conspecific soils, due to greater resource allocation to symbionts (Schiestl-Aalto et al., 2019) and recovery against pathogen infection (Martínez-Vilalta, 2014; Saffell et al., 2014). Both PSFs and functional traits can shift across environmental gradients like light availability (McCarthy-Neumann & Ibáñez, 2013; Smith-Ramesh & Reynolds, 2017). However, most studies have not integrated abiotic factors when evaluating both traits and PSFs (Cortois et al., 2016). Shifts in light can change microbial composition and abundance (Koorem et al., 2017; Y. Liu & He, 2019), which may alter seedlings’ ability to defend against or recover from disease. AMF are more abundant in higher light (Bereau et al., 2000; Koorem et al., 2017; Shi et al., 32 2014); however, in low light they can act parasitically and thereby decrease seedling survival (Konvalinková & Jansa, 2016). Higher mortality from pathogens typically occurs in low light (McCarthy-Neumann & Ibáñez, 2013; McCarthy-Neumann & Kobe, 2010a), where wetter and cooler conditions enhance microbe reproduction and dispersal (Hersh et al., 2012; Y. Liu & He, 2019; Reinhart et al., 2010). Light availability can also modify functional trait level, including reduced production of phenolics (Ichihara & Yamaji, 2009) and lignin (Falcioni et al., 2018; Rogers et al., 2005). Additionally, carbon limitation in shade and lower stored nonstructural carbohydrates (NSC) may constrain recovery from disease (Kobe, 1997; Kobe et al., 2010). My overall conceptual framework (Figure 3.1) is that soil source and light availability influence trait levels, which in turn influence tree seedling survival. Thus, plant traits have an important role in mediating PSFs. We hypothesized that: 1) Negative PSFs are widespread across tree species and are more prevalent under low than high light. Furthermore, these differences in PSFs are only present when soil-borne microbes are present. This result would indicate that soil-borne microbes drive negative PSFs in low light availability directly through increased pathogen abundance and/or a shift from positive to negative in the plant-mycorrhizal fungi relationship, and/or indirectly through decreased levels of defensive traits. 2) Mycorrhizal colonization is greater in conspecific soils and in higher light. This result would indicate that mycorrhizal colonization is promoted by effectively-specialized microbes in conspecific soils and greater resource availability (e.g., NSC) in high light. 3) The defensive functional traits phenolics and lignin are induced to higher levels in soils cultured by conspecific adults and in high light availability. This result would indicate that defensive functional trait production is driven by the presence of effectively-specialized 33 parasitic microbes expected in conspecific soils and by greater carbon income expected in higher light. 4) The recovery trait NSC is lower in soils cultured by conspecific adults and in low light availability. This result would indicate that NSC is drawn down in the presence of effectively-specialized parasitic microbes expected in conspecific soils, in addition to lower carbon income relative to use expected in lower light. 5) Finally, I hypothesized that seedling survival increases as mycorrhizal colonization, phenolics, lignin, and NSC also increase. This result would indicate that PSFs can, in part, be mediated by the degree of mycorrhizal colonization and changes in functional trait values responding to variation among soil types and in light availability. Figure 3.1 Conceptual diagram demonstrating the relationships between light availability, functional traits (phenolics, lignin, and nonstructural carbohydrates [NSC]), colonization by mycorrhizal fungi, and tree seedling survival. Green, solid lines indicate a positive relationship. Red, dashed lines indicate a negative relationship. Lines that directly influence tree seedling survival are thicker. Stars (*) next to the lines linking ‘Mycorrhizae’ with ‘NSC’ and ‘Seedling Survival’ indicate that this relationship is usually positive but can shift to neutral or negative. 34 MATERIALS AND METHODS I conducted a factorial blocked design field experiment, consisting of four tree species, seven soil sources (sterilized conspecific, live conspecific, and five heterospecific), and a gradient of forest understory light levels (low, medium, and high), for a total of 3,024 seedlings. I monitored seedling survival twice per week over one growing season, and I randomly selected subsets of seedlings to measure mycorrhizal colonization and phenolics, lignin, and NSC measurements at three weeks. I used Cox proportional hazards survival models to evaluate survival and linear mixed effects models to test how light availability and soil source influence defense and recovery traits. Study location The research site is a 100 ha mixed hardwood forest stand in mid-Michigan, at Alma College’s Ecological Field Station (43°23'32.0"N 84°53'41.5"W). Alma College granted permission to undertake this research and collect plant and soil materials; a formal field permit was not required. This forest has not been logged since 1897 and lies in an ecological tension zone between northern coniferous and southern deciduous forests. The dominant species in this forest is sugar maple (Acer saccharum), a shade-tolerant canopy tree species. Other common trees in the forest include red maple (A. rubrum) and big-toothed aspen (Populus grandidentata). Species selection We identified adult trees for soil collection and established field plots in a 3 ha mapped section of the forest (Table 3.1). I initially chose six tree species native to the research site: red maple (A. rubrum), sugar maple (A. saccharum), big-toothed aspen (P. grandidentata), black cherry (Prunus serotina), white oak (Quercus alba), and northern red oak (Q. rubra). A. rubrum and P. grandidentata seedlings experienced high (> 80%) mortality within two weeks of 35 planting, suggesting poor seed source or propagation methods. Thus, while still included as soil sources, they were not included in analyses of seedling survival, mycorrhizal colonization, or functional traits. Table 3.1 Local adult abundance, shade tolerance, seed weight, and primary mycorrhizal association for each of our study species. 1Local adult abundance was calculated as stems/ha at Alma College’s Ecological Preserve; only adults ≥ 5cm dbh were included in this count. 2Shade tolerance is presented as intolerant, intermediate, or tolerant and as mean ± std. dev., on a standardized scale from 1 (least tolerant) to 5 (most tolerant), calculated by Niinemets and Valladares (Niinemets & Valladares, 2006a). 3Seed weight data was collected from Burns and Honkala (Burns & Honkala, 1990a). 4AMF = arbuscular mycorrhizal fungi and EMF = ectomycorrhizal fungi. Species Acer rubrum Acer saccharum Populus grandidentata Prunus serotina Quercus alba Quercus rubra Local adult abundance1 131 285 Shade tolerance2 (3.44 ± 0.23) (4.76 ± 0.11) Seed weight (mg)3 19.7 64.9 Mycorrhizal association4 AMF AMF 82.33 4.33 12.67 71.67 (1.21 ± 0.27) 0.2 AMF & EMF (2.46 ± 0.34) (2.85 ± 0.17) (2.75 ± 0.18) 94.3 6,677 4,127 AMF AMF & EMF AMF & EMF Soil sources and planting I collected intact soil cores from May to July 2016 and April to May 2017 (Table A3.1). To minimize potential for multispecies culturing of soil, I took soil cores under trees that were at least two crown diameters away from adults of other species. Using a custom-made mechanized soil core sampler (Giddings Machine Co; Windsor, CO, USA), I removed intact soil cores (9 cm diameter and 30 cm deep for planted A. saccharum and P. serotina seedlings, or 46 cm long for Q. alba and Q. rubra seedlings) within 1 m from the bole of six mature randomly selected adults for each of the six study species (36 trees total). I maintained soil cores from each adult as separate replicates (Reinhart & Rinella, 2016; Rinella & Reinhart, 2018). 36 Intact soil cores with plastic liners were converted into pots by drilling two 7.5 cm diameter holes into the sides and adhering a 0.5 µm nylon mesh covering over side holes and the bottom opening. Such pots are an established method for studying common mycorrhizal networks in forests (Bingham & Simard, 2012; McGuire, 2007; Teste et al., 2006). The mesh prevents roots, fungal hyphae, oomycetes and pathogenic fungi from passing in or out, with minimal effect on water and nutrient flows (Allison et al., 2013). I did not use multi-stage greenhouse culturing (Bever et al., 2010), because in-situ natural culturing already had occurred for these long-lived trees and should more closely characterize PSFs occurring in the field. After resting, pots were transplanted into eighteen 8.4 ´ 6.6 m common-garden field plots that fell within three general light groupings (low, medium, and high). Existing vegetation and leaves in each plot were removed to reduce light interception. I then took precise measurements of light availability by analyzing canopy photos with HemiView software (Delta-T Devices, Ltd., Burwell, England; Figure A3.5). Soil samples for the sterilized conspecific soil treatment were exposed to gamma irradiation (30-70 kGy; Sterigenics International, Schaumburg, IL, USA) in July 2017 and allowed to rest for at least one month to minimize post-sterilization nutrient spikes. Gamma irradiation is highly effective at killing soil microorganisms and typically has minimal effects on soil chemical and physical properties (McNamara et al., 2003). Nevertheless, I tested the sterilized versus live soils using plant root simulator (PRSTM) probes (Western Ag Innovations Inc., Saskatchewan, Canada) and found no effect of sterilization on soil nutrient availability (Tables A3.3 & A3.4; Figure A3.3). I planted 108 seedling pots per species ´ soil source, evenly distributed across the 18 field plots. A single surface-sterilized seed with a newly-emerged radicle was planted into each 37 pot. Seeds for Q. alba were purchased from Sheffields Seed Co (Locke, NY, USA) and seeds for all other species were collected from mid-Michigan forests. Variation among seed source populations in survival, mycorrhizal colonization and functional traits was likely minimal (McCarthy-Neumann & Ibáñez, 2012). In June 2018, one week prior to planting, I added 1 cm of a 1:1 mixture of peat moss and fresh or sterilized soil to increase transplant success and provide fresh inoculum. In a previous trial run, I found that seedlings planted with peat moss and fresh soil had reduced transplant shock; personal observation). To minimize disease from non-experimental soil sources, seeds were surface sterilized with 0.6% NaOCl solution prior to stratification and prior to germination. To avoid cross- contamination, all tools and surfaces that were exposed to soil were soaked in 10% bleach or surface sprayed with 70% EtOH and then rinsed with deionized water. To minimize browsing and digging-up of seedlings by vertebrates, I erected galvanized hardware cloth (6 ´ 6 cm openings) to 1.8 m height around each plot. I also glued hardware cloth with 0.25 cm ´ 0.25 cm openings to the top of each pot. Seedlings likely did not experience significant shading due to the addition of the hardware cloth and often grew above the cloth within 2 weeks of planting. Survival and functional traits I censused seedling survival twice per week for 16 weeks. Mortality at the first two censuses after planting were attributed to transplant shock or poor seed source; these seedlings were not used in subsequent analyses, and pots were re-planted with the same seedling species. Three weeks after planting, I harvested six seedlings per treatment combination to measure mycorrhizal colonization, phenolics, lignin, and NSC. I chose this harvest date since, in a previous greenhouse experiment, mortality curves for tree seedlings subjected to soil-borne pathogens often increased at week three and peaked between four to six weeks after germination 38 (McCarthy-Neumann & Ibáñez, 2012). For measurements, I used established protocols: AMF and EMF colonization (McGonigle et al., 1990; Vierheilig et al., 1998), phenolics (Ainsworth & Gillespie, 2007; P. Waterman & Mole, 1994), lignin (ANKOM Technologies, Macedon, NY, USA), and NSC (Landhäusser, Chow, Dickman, et al., 2018). Due to the small size of three- week-old seedlings and the destructive nature of each measurement, half of the harvested seedlings were allocated to measurement of NSC (stem and root), and half of the seedlings were allocated to measurement of phenolics (hypocotyl), lignin (stem), and mycorrhizal colonization (roots). Statistical analysis To evaluate hypotheses 1, I analyzed seedling survival over 16 weeks with Cox proportional hazards regression (Cox & Oakes, 2017). I ran species-specific models, using soil source and light availability as fixed effects, and plot and adult tree as random effects. The best fitting models for seedling survival did not include any interactions. I compared survival in live conspecific versus heterospecific soils (Gómez-Aparicio et al., 2017; Xi et al., 2021). Greater survival in live conspecific than heterospecific soils indicated positive PSFs. I compared survival in sterilized versus live conspecific soils. Higher survival in sterilized than live conspecific soils indicated that microbes influenced PSFs. To evaluate hypotheses 2-4, we analyzed measured amounts of mycorrhizal colonization, phenolics, lignin, and NSC with linear mixed effects models. I ran species-specific models for each trait, using soil source and light availability as fixed effects, and plot and adult tree as random effects. I used a priori contrasts to compare levels of measured traits in live conspecific versus heterospecific soils and to compare levels of measured traits in sterilized versus live conspecific soils. 39 To evaluate hypothesis 5, I analyzed seedling survival over 16 weeks with Cox proportional hazards regression (Cox & Oakes, 2017). I ran species-specific models, with mycorrhizal colonization, phenolics, lignin, and NSC as fixed effects. I imputed colonization and trait data from seedlings harvested at three weeks, for each combination of seedling species, plot, soil source, and light level. I accounted for possible collinearity between traits by calculating variance inflation factors (VIF) for each model and removing variables with VIF > 5. For A. saccharum, A. rubrum, and Q. rubra, NSC was removed from the final models, and for Q. alba, lignin was removed from the final models. NSC was highly correlated with lignin for all study species and with phenolics for Q. rubra (Figure A3.8). For all models, light availability was first evaluated as a continuous variable, using Indirect Site Factor (ISF) quantified with canopy photos. ISF represents the proportion of diffuse (indirect) solar radiation reaching a given location, relative to an open site and was calculated using HemiView software (Delta-T Devices, Ltd., Burwell, England). For post-hoc comparisons and figures, I divided seedlings according to light group, splitting the range of light availability into three bins (low = 0.032-0.075 ISF, medium = 0.075-0.118 ISF, and high = 0.118-0.161 ISF). These light thresholds were determined by dividing the range of light availability across the field plots into three bins. Heterospecific soils were modeled as both pooled and unpooled/specific soils; when evaluating post-hoc comparisons, we used pooled heterospecific soils. All analyses were performed with R version 3.5.1 (R Core Team, 2020). I used the “coxph” function in the survival package (Therneau & Grambsch, 2000) to fit Cox proportional hazards regression models. I tested the significance of main effects using a likelihood ratio test with the “Anova” function. I tested for multicollinearity variance inflation factors using the “vif” function in the car package (Fox & Weisberg, 2019). Post-hoc Tukey pairwise comparisons of 40 significant main effects and Bonferonni corrections for multiple comparisons were made using the “emmeans” function in the multcomp package (Hothorn et al., 2008; Lenth, 2020). I used the missForest package (Stekhoven & Buehlmann, 2012) to impute trait data for seedlings monitored for survival. RESULTS Negative PSFs were not widespread among tree species, nor were they more prevalent in low light availability No species experienced negative PSFs (defined as lower survival in conspecific versus heterospecific soils). However, A. saccharum experienced positive PSFs with higher survival in conspecific than pooled heterospecific soil (LRc2 = 8.60, p < 0.01; Figure 3.2A, Table 3.2A). Seedling survival was lower in live than sterilized conspecific soil for both A. saccharum (LRc2 = 61.78, p < 0.01) and P. serotina (LRc2 = 1.52, p < 0.01), suggesting an effect of soil-borne microbes. Although there was a positive effect of light on survival for P. serotina (LRc2 = 4.09, p = 0.04) and Q. rubra (LRc2 = 9.02, p < 0.01; Fig 2B, Table 3.2B) there was no significant interaction between soil source and light availability for any of the models with pooled heterospecific soils. When heterospecific soils were not pooled, there was a significant interaction between light and soil source, but only for P. serotina (LRc2 = 6.860, p < 0.01). Thus, the expectation that negative PSFs are widespread among species and are more prevalent in low light availability was not supported. 41 A) Figure 3.2 Kaplan-Meier plots evaluating the effects of A) soil source (conspecific, pooled heterospecific, and sterilized conspecific) on seedling survival, and B) light availability on seedling survival. For visualization, light availability was binned into 3 levels: Low = 0.032 - 0.075 ISF, Med = 0.075-0.118 ISF, and High = 0.118 - 0.161 ISF. Shaded regions indicate 95% confidence intervals about the mean. 42 Figure 3.2 (cont’d) B) 43 Table 3.2 Number of surviving seedlings at the end of the growing season. Data is presented for each A) species ´ soil source and B) species ´ light level as both an absolute number and percentage. Because there was not a significant interaction between soil source and light availability on seedling survival, they are presented separately, corresponding with Figures 3.2A and B. Soil sources include sterilized conspecific, live conspecific, and pooled heterospecific soils. Light availability was binned into 3 levels: Low = 0.032 - 0.075 ISF, Med = 0.075 - 0.118 ISF, and High = 0.118 - 0.161 ISF. Species Acer saccharum Prunus serotina Quercus alba Quercus rubra Sterilized conspecific 1 (1%) 5 (8.9%) 36 (50.7%) 33 (49.3%) A) Soil source Live conspecific 0 (0%) 3 (3.3%) 23 (50%) 27 (54%) Hetero- specific 0 (0%) 12 (2.8%) 199 (66.3%) 152 (65%) B) Light availability Low Med High 0 (0%) 8 (3.9%) 83 (60.1%) 66 (52%) 1 (0.2%) 7 (2.3%) 150 (64.7%) 118 (62.8%) 0 (0%) 5 (7.7%) 25 (51.1%) 28 (77.8%) Mycorrhizal colonization and seedling functional traits varied across both soil source and light availability AMF colonization was 11% higher in conspecific than pooled heterospecific soil only for A. saccharum (t2344 = 1.84, marginally significant at p = 0.07; Figure 3.3A). For the other study species, AMF colonization was higher in pooled heterospecific than conspecific soils: 12% for P. serotina (t2344 = 3.88, p < 0.01), 16% for Q. alba (t2344 = 2.38, p = 0.02), and 12% for Q. rubra (t2344 = 2.05, p = 0.04). As predicted, AMF colonization increased with light for P. serotina (slope = 108% / ISF, F1,2344 = 35, p < 0.01) and Q. rubra (slope = 53% / ISF, F1,2344 = 6.42, p < 0.01), but not for A. saccharum. Contrary to our expectations, AMF colonization decreased with light for Q. alba, which is primarily associated with EMF (slope = -49% / ISF, F1,2344 = 6.08, p = 0.01). 44 Figure 3.3 Effect of soil source and light availability on percent mycorrhizal colonization. By A) AMF and B) EMF (only Q. alba and Q. rubra are colonized by EMF). Shaded regions indicate 95% confidence intervals about the mean. Solid lines have a slope significantly different from zero (p < 0.05). EMF colonization was higher in conspecific than pooled heterospecific soil by 16% for Q. alba across all light levels (t1063 = 2.72, p = 0.01; Fig 3.3B) and by 22% for Q. rubra in high, but not low light (t1063 = 4.74, p < 0.01). EMF colonization increased with light availability for Q. rubra (slope = 264.2% / ISF, F1,1063 = 63.02, p < 0.01), especially in conspecific soil (slope = 413% / ISF). Phenolics (nmol Gallic acid equivalents per mg dry extract) were higher in live than sterilized conspecific soils for A. saccharum (227%, t793 = 7.85, p < 0.001), P. serotina (173%, 45 t793 = 6.77, p < 0.001), and Q. alba (51.7%, t793 = 43.73, p < 0.001). As expected, phenolics were higher in conspecific than pooled heterospecific soil for A. saccharum (23%, t2344 = 10.56, p < 0.01; Figure 3.4A) and Q. alba (4%, t2344 = 4.44, p < 0.01). Conversely, phenolics were 69% higher in pooled heterospecific soil for P. serotina (t2344 = 6.96, p < 0.01). For Q. rubra, phenolics were 18% higher in conspecific soil at high light (t2344 = 5.89, p < 0.01) and 29% higher in pooled heterospecific soil at low light (t2344 = 12.77, p < 0.01). Phenolics increased with light availability for all four study species (A. saccharum: slope = 5.70 nmol / ISF, F1,2344 = 64.7, p < 0.01; P. serotina: slope = 4.15 nmol / ISF, F1,2344 = 34.48, p < 0.001; Q. alba: slope = 10.51 nmol / ISF, F1,2344 = 187.07, p < 0.01; Q. rubra: slope = 12.73 nmol / ISF, F1,2344 = 249.42, p < 0.01). For Q. rubra, this trend appeared to be driven by conspecific soil (slope = 23.26 nmol / ISF). Percent dry mass lignin was higher in conspecific than pooled heterospecific soil by 11% for Q. alba (t1,2344 = 8.60, p < 0.01) and 5.8% for Q. rubra (t2344 = 5.61, p < 0.01), across all light levels (Figure 3.4B). For both A. saccharum and P. serotina, lignin did not vary between conspecific and pooled heterospecific soil. Lignin increased with light availability for A. saccharum (slope = 38% / ISF, F1,2344 = 82.34, p < 0.01) and P. serotina (slope = 57% / ISF, F1,2344 = 184.51, p < 0.01). There was no effect of light on lignin for Q. alba. Contrary to our predictions, for Q. rubra, lignin decreased with light availability (slope = -22% / ISF, F1,2344 = 21.42, p < 0.01); this trend appeared to be driven by conspecific soil (slope = -40% / ISF). Indicating a potential effect of soil biota, lignin (percent dry mass) was higher in live than sterilized conspecific soils for all four study species: A. saccharum (12%, t793 = 6.40, p < 0.001), P. serotina (13%, t793 = 4.18, p < 0.001), Q. alba (12%, t793 = 10.56, p < 0.001), and Q. rubra (2.4%, t793 = 2.46, p = 0.014). 46 Figure 3.4 Effect of soil source and light availability on functional traits. Traits include: A) phenolics (nmol Gallic acid equivalents per mg dry mass), B) percent dry mass lignin, and C) percent dry mass NSC. Some lines are truncated, because not enough seedlings survived in that light level. Shaded regions indicate 95% confidence intervals about the mean. Solid lines have a slope significantly different from zero (p < 0.05). 47 Percent dry mass NSC was higher in pooled heterospecific than conspecific soil across all light levels for Q. alba (15%, t2344 = 9.96, p < 0.01; Figure 3.4C). For P. serotina and Q. rubra, NSC was higher in conspecific soil at low light (P. serotina: 1.9%, t2344 = 3.41, p < 0.01; Q. rubra: 13%, t2344 = 9.34, p < 0.01), but did not vary with soil source at high light. For all four study species, NSC increased with light availability (A. saccharum: slope = 21% / ISF, F1,2344 = 22.05, p < 0.01; P. serotina: slope = 75% / ISF, F1,2344 = 287.75, p < 0.01; Q. alba: slope = 19% / ISF, F1,2344 = 16.39, p < 0.01; Q. rubra: slope = 47% / ISF, F1,2344 = 87.42, p < 0.01). This trend appeared to be driven by pooled heterospecific soil for Q. alba (slope = 30), and conspecific soil for Q. rubra (slope = 76% / ISF). NSC was higher in sterilized than live conspecific soils for Q. alba (6.7%, t793 = 4.17, p < 0.001). Mycorrhizal colonization and functional traits had limited effects on seedling survival From the Cox survival models, I interpreted hazard ratios (HR), an integration of the hazard experienced by seedlings across the study duration. HR < 1 indicates decreased hazard relative to the baseline (i.e., increased survival); HR > 1 indicates increased hazard (i.e., decreased survival). Traits predicted survival for two species: phenolics had a positive effect on survival for A. saccharum (HR = 0.73, LRc2 = 4.20, p = 0.04; Figure 3.5) and AMF colonization had a negative effect for Q. alba (HR = 1.04, LRc2 = 4.18, p = 0.04). 48 Figure 3.5 Hazard ratios (HR) demonstrating the effect of mycorrhizal colonization and functional traits on seedling survival over the growing season. HR > 1 indicates an increase in mortality and HR < 1 indicates a decrease in mortality as the trait increases. Statistically- significant effects (p < 0.05) are colored blue. Species x trait combinations that are blank were removed from the final models due to high collinearity (VIF > 5). 49 DISCUSSION While I found limited evidence of survival-based PSFs in this field study, defense and recovery traits varied in response to soil source and light availability in seedlings as young as three weeks old. I also found limited associations between defense and recovery traits and seedling survivorship. These results support that, while seedling defense and recovery traits are very responsive to soil source and light availability, PSFs may be less prevalent in field conditions where multiple environmental factors influence seedling survivorship. PSFs were not widespread among species, nor were they more prevalent in low light availability Only one of four study species experienced PSFs between conspecific and heterospecific soils, and there were very few interactions between light availability and soil source on seedling survival. I expected to find stronger negative PSFs in low light conditions (McCarthy-Neumann & Ibáñez, 2013), due to both greater limitation of light availability and higher prevalence of soil- borne pathogens. Both A. saccharum and P. serotina experienced lower survival in live than sterilized conspecific soils, indicating that soil-borne microbes have negative effects on seedling survival. For P. serotina, although soil-borne microbes cultured in conspecific soils may have a negative effect on survival, the net effect of PSFs (assessed as survivorship in conspecific versus pooled heterospecific soils) appeared to be neutral (Esch & Kobe, 2021; McCarthy-Neumann & Kobe, 2019; Packer & Clay, 2003). A. saccharum seedlings experienced net positive PSFs, having greater survival in conspecific than heterospecific soils. However, they had even greater survival in sterilized than live conspecific soils, consistent with McCarthy-Neumann and Ibáñez (2013) and suggesting net 50 negative effects of soil-borne microbes. There are at least three mutually compatible explanations for these results: 1) mutualistic microbes may provide greater benefit in conspecific than heterospecific soils; 2) there may be a greater negative effect of harmful microbes in heterospecific soils; or, 3) there may be unmeasured, more favorable abiotic effects in soils modified by A. saccharum adults in comparison to heterospecific soils (McCarthy-Neumann & Ibáñez, 2012). Comparisons of PSFs in this study are presented as differences in rates (calculated as LRc2), rather than differences in total number of surviving seedlings at the end of the growing season (Figure 3.2, Table 3.2). Both A. saccharum and P. serotina had zero (or near-zero) survival by the end of the growing season. However, investigating the environmental conditions that influence mortality rates for these seedlings is still meaningful for understanding forest communities. When adult A. saccharum and P. serotina trees produce thousands of seeds in a single growing season (Burns & Honkala, 1990a), small differences in survival rates can scale up to meaningful impacts on community composition over longer time periods. Mycorrhizal colonization and functional traits varied across both soil source and light availability These results demonstrate that tree seedlings, even as young as three weeks old, express intraspecific variation in mycorrhizal colonization and functional trait values, in response to conspecific versus heterospecific soil source and light availability. AMF colonization was higher in heterospecific than conspecific soil for most of the measured species, with the largest difference being for P. serotina at high light availability. This was in contrast to a study evaluating PSF in temperate tree species across North America, which found that AMF colonization was equal or greater in conspecific relative to heterospecific soils (Bennett et al., 51 2017). My result may be because AMF are more generalized in host associations than EMF for our study species (S. Smith & Read, 2008). AMF colonization was highest in P. grandidentata soil for P. serotina and Q. alba seedlings, and in A. saccharum soil for P. serotina and Q. rubra, suggesting that AMF from P. grandidentata and A. saccharum soils can readily colonize multiple seedling species. Surprisingly, AMF colonization increased with light availability only for P. serotina. Across species, P. serotina also had the highest total AMF colonization. I speculate that, as a shade intolerant species, P. serotina seedlings may regulate mycorrhizal colonization (MacLean et al., 2017; Mangan et al., 2010) by investing more resources into colonization at high light, where carbon is less limiting (Grman et al., 2012). Consistent with expectations, EMF colonization was higher in conspecific than heterospecific soil for Q. alba and Q. rubra, especially in higher light availability (Trocha et al., 2016; Turner et al., 2009), perhaps reflecting the higher specialization of EMF than AMF (S. Smith & Read, 2008). For Q. alba, EMF colonization also was high in soils cultured by P. grandidentata and Q. rubra, suggesting association with multiple EMF species. Phenolics increased with light availability for all study species, but there were no consistent effects of live soil source across all seedling species. Phenolics were higher in conspecific than heterospecific soil for A. saccharum across light levels, and for Q. rubra at high light. Additionally, phenolics were higher in live than sterilized conspecific soils, suggesting that phenolics increase in the presence of soil-borne microbes. Phenolics production can be induced in response to soil-borne microbes (Pozo & Azcón-Aguilar, 2007; Whipps, 2004), which should be more prevalent in conspecific soil. Although I did not quantify pathogen abundance, I did find that EMF colonization and phenolics were correlated for Q. alba and Q. rubra. Since EMF colonization was higher in conspecific than heterospecific soil for both species, my results 52 suggest that EMF colonization or the presence of host-specific pathogens induced production of phenolics, especially in conspecific soil. For EMF-associated species, lignin was higher in conspecific than heterospecific soil, but did not vary with light. Seedling production of lignin may have already reached the upper limit in response to light availability. Seedlings may also achieve greater trait production under less stressful growing conditions, such as in conspecific soils with mutualistic EMF (Valladares et al., 2007). By improving nutrient availability, EMF can indirectly affect the allocation of seedling resources, potentially impacting lignin synthesis. NSC increased with light availability for all study species, regardless of soil biota present. This result was consistent with previous studies (Dillaway et al., 2007; Piper et al., 2009; Zhang et al., 2013), including Q. alba (Dillaway et al., 2007). Contrary to expectations, for Q. alba, NSC was higher in heterospecific soil. For P. serotina and Q. rubra, NSC was higher in conspecific soil but only at low light. While one might speculate that seedlings may allocate more NSC to mycorrhizal mutualists or recovery from pathogens in conspecific soil or high light, I did not find strong correlations with NSC for either AMF or EMF colonization. Mycorrhizal colonization and functional traits had limited effects on seedling survival Species differed in which traits, if any, influenced survival. Phenolics, which provide direct chemical defense against soil-borne microbes, could increase survival for A. saccharum seedlings and may be the mechanism behind their positive PSFs, supported by greater production of phenolics in conspecific soils (Pozo & Azcón-Aguilar, 2007; Whipps, 2004). However, A. saccharum seedling survival was higher in sterilized than conspecific soil, suggesting that the positive effects of phenolics on survival did not overcome the negative effects of microbes. Furthermore, higher survival in sterilized soil confirms that microbes drove the observed positive 53 PSFs. For P. serotina, phenolics were much higher in pooled heterospecific soil and were positively correlated with AMF colonization. P. serotina seedlings may be more readily colonized by AMF, regardless of soil source, and thus produce more phenolics in response; this may explain why P. serotina are frequently found to have high mortality in conspecific soils (Esch & Kobe, 2021; Packer & Clay, 2000). For typically EMF-associating Q. alba seedlings, survival decreased as AMF colonization increased. A potential explanation is that AMF can act parasitically in some environmental conditions (Ibáñez & McCarthy-Neumann, 2016; Konvalinková & Jansa, 2016) while EMF provide better direct protection against pathogens. While it is not well-understood if AMF colonization cause negative PSFs for EMF-associating tree species (Chilvers et al., 1987; Duponnois et al., 2003), previous studies (Bennett et al., 2017) found no effect of soil cultured by AMF-associating species on seedlings of EMF-associating species. Furthermore, the pot- based study design may have precluded the benefits of an EMF common mycorrhizal network (Simard & Durall, 2004), heightening the negative influence of AMF colonization. Interestingly, Q. rubra seedling survival was not influenced by AMF colonization, suggesting that Q. rubra may be less reliant upon common mycorrhizal networks or less susceptible to parasitic effects of AMF. I expected to see traits emerge as stronger drivers of seedling survival, given large intraspecific trait variation in response to soil source and light availability. Although I found that functional traits were influenced by both soil source and light availability, I found limited instances of trait influences on seedling survival, which was likely due to abnormally high mortality experienced by seedlings during the field season. The lack of strong effects could have been driven by stressful field conditions that obscure the importance of functional traits. In 54 contrast to previous greenhouse studies, I did not find PSFs for A. rubrum, P. serotina, or Q. rubra. PSFs quantified in the highly controlled greenhouse conditions often overestimate field measured PSFs (Forero et al., 2019). In this study, I took precautions against potential competition and above-ground herbivory from rodents and deer. However, seedlings experienced high amounts of mortality, killing almost all of the A. saccharum and P. serotina seedlings, and killing half of the Q. alba and Q. rubra seedlings. This may be in part due to the great variability in rainfall and maximum temperature experienced by seedlings across the growing season. Heavy rainfall washed away smaller seedlings and would stand in pots if preceded by dryer, warmer periods, which caused soil in the field pots to pull away from the sides of the container and harden. Evaluation of weather data (National Oceanic and Atmospheric Administration) revealed a higher amount of rainfall throughout the field season and a large rainfall event (> 7”) in July. Variation in weather could have overridden the effects of traits (Putten et al., 2016) or light availability (Catford et al., 2022). PSF experiments carried out in the greenhouse may not detect such environmental effects (Beals et al., 2020) or may overestimate the strength of PSFs (Heinze et al., 2020). There are several additional caveats to consider. I expect that there was some contamination of pots via airborne microbes, splash from rain, and falling leaf litter, which may have reduced effect sizes. However, we assume that the reported significant effects are the result of treatments, because invasions are random and microbial priority effects should be dominant, especially in whole-soil cores (P. J. Ke et al., 2021). In addition, the relationship between functional traits and seedling survival is correlative rather than causative since I did not manipulate levels of seedling functional traits. It is also difficult to disentangle some of these trait-survival relationships. For example, while I expect phenolics and lignin to be higher in 55 conspecific soils, and for increases in these traits to lead to higher seedling survival, I also expect higher mortality in conspecific soils where effectively-specialized pathogens are more abundant. However, these results still elucidate the role of soil-borne microbes on tree seedling trait levels. I was unable to separate the impacts of soil-borne mutualists and pathogens on seedling trait values and subsequent survival. Also, I cannot distinguish between direct AMF colonization effects of pathogen reduction through displacement or indirect effects inducing production of phenolics, both of which can enhance seedling survival. Additionally, I was unable to tease apart the effects of colonization type (AMF versus EMF) and seed size, since the EMF-associated species used in this study were large-seeded, and vice-versa (Table 3.1). Furthermore, this study is limited to four species occurring in a single forest; future studies examining the generalizability of these results should consider additional species and habitats. Conclusion Linking plant traits and environmental conditions to PSFs may help us better understand the role of PSFs in community dynamics (Baxendale et al., 2014; Bennett & Klironomos, 2018; P. Ke et al., 2015). However, most studies have focused on herbaceous plants, which are predominantly colonized by AMF. A focus on tree seedling traits under different environmental conditions, especially in natural field conditions, offers both broader ecological understanding as well as potential applications for forest management. For example, selecting sites with soil and light conditions that promote higher production of defensive and recovery compounds could increase likelihood of seedling restoration success (e.g., A. saccharum in conspecific soil). Similarly, it may be beneficial to plant EMF-associating seedlings in soils cultured by other EMF-associating species, to increase potential for positive EMF colonization effects and limit potential negative AMF colonization effects. While environmental conditions could dilute trait 56 effects on seedling survival, in the absence of extreme conditions (as supported by related greenhouse studies), a sharper focus on traits promoting survival rather than growth traits will provide a more mechanistic understanding of forest regeneration dynamics. A modified version of this chapter has been published in PLOS One. The original publication is available at https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0293906. 57 CHAPTER 4 Tree seedling responses to plant-soil feedbacks depend on mycorrhizal type of the adults culturing the soil ABSTRACT The matching or mismatching of mycorrhizal type between canopy trees and recruiting seedlings may interact with the juvenile’s functional traits and thus affect their responses to plant-soil feedbacks (PSFs). Understanding these interactions will provide a more mechanistic understanding of forest community dynamics. To assess the role of mycorrhizal type matching on juvenile trees’ trait response and PSFs, I carried out a greenhouse experiment where I grew seedlings of five temperate tree species under soils cultured by conspecific versus heterospecific adults. Seedlings were also grown under three levels of light availability to assess potential shifts in the PSF effects (i.e., positive at high light, and negative at low light). After 12 weeks, I quantified seedling survival, colonization by arbuscular- and ecto-mycorrhizal fungi (AMF and EMF) and measured their defense and recovery traits (phenolics, lignin, and nonstructural carbohydrates [NSC]). I found that negative PSFs (lower survival in conspecific versus heterospecific soils) experienced by seedlings associating with AMF almost always occurred when they were compared with heterospecific adults associating with EMF. Conversely, positive PSF experienced by EM seedlings occurred when compared to soils cultured by AM adults. Although PSFs occurred regardless of light level, the magnitude of effect – both negative for AM and positive for EM seedlings – was greatest at low light. Soil microbes from conspecific-cultured soils reduced survival for AM species but had no effect on survival for EM species. Percent 58 mycorrhizal colonization and functional trait values were higher in high light and, except for NSC, were often higher in conspecific soils. Furthermore, PSFs for AM seedlings became less negative as percent AMF colonization and defense/recovery traits increased, and PSFs for EM seedlings became less positive as percent AMF colonization and lignin increased. These results suggest that increased colonization by mycorrhizal fungi and increased amounts of phenolics, lignin, NSC effectively neutralize both negative and positive PSFs, providing new insights into how mismatching of mycorrhizal type interacts with defense and recovery traits to influence PSFs, and thus forest community dynamics. INTRODUCTION Identifying the mechanisms that maintain tree species richness is a central question in plant community ecology. The seeding-establishment phase is a critical stage for the maintenance of future community-wide species diversity, as this phase is a major demographic bottleneck for populations (Gurevitch et al., 2020). Seedling recruitment is influenced by plant- soil feedbacks (PSFs), a continuous feedback loop wherein adult trees modify the soil in which they are growing, which in turn shapes seedling community assembly (Bever et al., 1997). PSFs can be positive (better seedling performance in conspecific than heterospecific soils), negative (better seedling performance in heterospecific than conspecific soils), or neutral (no difference in seedling survival in conspecific versus heterospecific soils). The direction and strength of PSFs may be partially explained by the mycorrhizal fungal type of the host plant species, with negative feedbacks more often experienced by plant species associated with arbuscular mycorrhizal fungi (AMF) and positive feedbacks more often experienced by species associated with ectomycorrhizal fungi (EMF) (Bennett et al., 2017). In the present study, I expand this 59 framework by also testing the matching or mismatching of mycorrhizal type between juvenile and adult trees culturing the soil, in addition to how defense and recovery traits interact to influence PSFs. The primary biotic agents of PSFs are soil-borne microbes, including pathogens and mycorrhizal fungi (Bever et al., 2010; Jiang et al., 2020). Soil-borne pathogens (fungi, oomycetes, and bacteria), increase seedling mortality, sometimes killing entire seedling cohorts (Mangan et al., 2010; Terborgh, 2012). While mycorrhizae typically act as mutualists, providing water and nutrients in exchange for sugars, their relationship with seedlings can shift to parasitic in low light environments (Konvalinková & Jansa, 2016; McCarthy-Neumann & Ibáñez, 2013), when photosynthates are more limited. Mycorrhizae can also confer protection against antagonists, like pathogens, but the degree of protection depends upon mycorrhizal type (Bennett et al., 2017). AMF can provide indirect defense against pathogens by competing for space on plant roots (Borowicz, 2001), whereas EMF can provide direct defense by forming a protective sheath on young roots (Laliberté et al., 2015). The net effects of PSFs may range from positive to negative, depending on interactions between soil-pathogens and mycorrhizal fungi (Laliberté et al., 2015; Reinhart & Callaway, 2006). The type of mycorrhizal symbiont (AMF or EMF) hosted by tree seedling species can have large impacts on PSFs. (Here, I refer to species that typically associate with AMF and EMF as “AM species” and “EM species”, respectively). AM trees typically experience negative or neutral PSFs (i.e., inhibition of seedlings around conspecific adults); conversely, EM trees experience positive or neutral PSFs (i.e., facilitation around conspecific adults) (Bennett et al., 2017; Kadowaki et al., 2018). AM trees have a higher abundance of plant pathogens in the soil beneath their crowns, relative to soils beneath EM trees (Eagar et al., 2022, 2023), which may be 60 explained by the relatively-low amount of protection that AMF confer against these pathogens. In addition, Chen et al. (2019) found that AM seedlings accumulate soil-borne pathogens faster under AM than EM adults, providing a potential explanation for why seedlings of AM tree species experience higher root pathogen damage when grown in conspecific soil, whereas EM tree species have elevated mycorrhizal colonization in conspecific soil (Bennett et al., 2017). However, when there is mismatching of mycorrhizal types (e.g., AM seedlings growing beneath EM trees and vice-versa), both AM and EM seedlings appear to experience positive or neutral PSFs (Kadowaki et al., 2018). Furthermore, seedling sensitivity to negative PSFs is often restricted to low light availability and increases with seedling shade intolerance (McCarthy-Neumann & Ibáñez, 2013; McCarthy-Neumann & Kobe, 2010b, 2010a). Soil moisture conditions associated with low light environments can result in greater soil pathogen density and increase fungal pathogen colonization (Hersh et al., 2012; Y. Liu & He, 2019; Reinhart et al., 2010). Colonization by mycorrhizal fungi also reduces seedling performance in low light conditions (Ibáñez & McCarthy-Neumann, 2014, 2016). In high light, AMF root colonization is greater and can compete with pathogens for carbon from plant roots (Borowicz, 2001), directly reducing pathogen infection. Mycorrhizal fungi can also indirectly ameliorate pathogen effects (Liang et al., 2015), by providing water and nutrients to the host plant (Borowicz, 2001; Graham, 2001) and inducing host plant defense traits (Pozo & Azcón-Aguilar, 2007; Zamioudis & Pieterse, 2012) that protect against pathogens (Azcón-Aguilar et al., 2002; Violle et al., 2012). Conversely, in low light, seedling mortality may be greater when seedlings encounter both mycorrhizal fungi and pathogens together (König et al., 2016), which may be due to the combined carbon costs of maintaining the mutualism and defense/recovery from pathogen attack. 61 Seedling functional traits – that is, measurable morphological or physical attributes that impact seedling performance (Pérez-Harguindeguy et al., 2013) – are likely influenced by soil source, mycorrhizal type, and light availability. Functional traits, such as those that confer defense against and recovery from pathogens, likely increase seedling survival in the first growing season and thus have a large influence on future community dynamics. Growth-related functional traits (e.g., specific leaf area, specific root length, height) are linked to PSFs (Baxendale et al., 2014; Cortois et al., 2016; Xi et al., 2021), but defensive, survival-related traits are rarely examined directly. When defensive traits are included, they are often inferred from an inverse relationship with growth traits (Cortois et al., 2016; Xi et al., 2021). Traits that confer greater defense and recovery against soil-borne microbes include phenolics, lignin, and nonstructural carbohydrates (NSC). Phenolics and lignin act as chemical (Ichihara & Yamaji, 2009) and physical (Augspurger, 1990) defenses, whereas NSC can be mobilized to repair damaged tissues (Dietze et al., 2014). In previous greenhouse (Chapter 2) and field (Chapter 3) studies, I found that mycorrhizal fungi were associated with higher amounts of seedling phenolics and NSC. In those studies, I also found that increasing seedling AMF colonization, phenolics, and lignin were associated with higher survival in the presence of soil- borne pathogens. In this study, I planted seedlings of five temperate tree species under soils cultured by adults of each of those species and in three levels of light availability in the greenhouse. My global hypothesis is that, for tree seedlings, the influence of defense and recovery traits on PSFs (calculated as the difference in survival between conspecific and heterospecific soils) is driven by mismatches between seedling mycorrhizal type and the mycorrhizal type of the adult tree culturing the soil. I predicted that: 62 1) AM seedlings experience negative PSFs (lower survival in conspecific versus heterospecific soils) and EM seedlings experience positive PSFs (higher survival in conspecific versus heterospecific soils) across light levels. 2) AM seedlings experience greater negative PSFs in low than high light. EM seedlings experience greater positive PSFs in high than low light. 3) Soil-borne microbes from conspecific soils have negative effects on survival for AM species and positive effects on survival for EM species, which could partly explain the PSFs found in hypotheses 1-2. 4) All study species have the greatest mycorrhizal colonization at high light availability and when grown in soils cultured by the same mycorrhizal type. 5) Phenolics, lignin, and NSC increase as light availability increases. Phenolics and lignin are higher, and NSC is lower, in conspecific soils. 6) Seedling PSFs become less negative for AM species and more positive for EM species as mycorrhizal colonization and defense/recovery traits increase. These relationships are enhanced in low light availability. This study provides insights into how matching or mismatching of mycorrhizal type between the juvenile growing in and adult tree culturing the soil can shift the strength and direction of PSFs. Furthermore, I show that both defense/recovery traits and light availability can interact to mediate PSFs. 63 MATERIALS AND METHODS I conducted a factorial blocked greenhouse experiment, consisting of five tree species, seven soil sources (sterilized conspecific, non-sterilized conspecific, and five heterospecific), across three light levels (low, medium, and high), replicated over 30 seedlings (n = 30), for a total of 3,150 seedlings. I monitored seedling survival twice per week over twelve weeks. When seedlings were three and twelve weeks old, I measured percent mycorrhizal colonization (AMF and EMF) and amounts of defense and recovery functional traits (phenolics, lignin, and NSC). Species selection The experiment was conducted at the Michigan State University Tree Research Center in Lansing, MI, USA (42.7 ºN, 84.5 ºW) in spring 2018. Soils were collected in a 100 ha mixed hardwood forest stand in mid-Michigan, at Alma College’s Ecological Field Station (43°23'32.0"N 84°53'41.5"W) in summer 2017. This forest has not been logged since 1897 and lies in an ecological tension zone between northern coniferous and southern deciduous forests. The dominant species in this forest is sugar maple (Acer saccharum), a shade-tolerant canopy tree species. Other common trees in the forest include red maple (A. rubrum) and big-toothed aspen (Populus grandidentata). I chose six tree species native to the research site: red maple (A. rubrum), sugar maple (A. saccharum), big-toothed aspen (P. grandidentata), black cherry (Prunus serotina), white oak (Quercus alba), and northern red oak (Q. rubra). The six study species vary in local adult abundance, shade tolerance, seed size, and mycorrhizal association type (Table 4.1). Due to difficulty acquiring seeds, I did not grow P. grandidentata seedlings. However, P. grandidentata soils were still included as a treatment. 64 Soil sources and sterilization To minimize potential for multispecies culturing of soil, I chose adult trees for soil collection that were at least two crown diameters away from adults of other study species. In August 2017, I collected soil (top 15 cm) from within 1 m of each stem of the focal trees. I prepared soil by dicing roots and sifting through a 1 cm mesh sieve, retaining all fine roots, and maintained soil from each adult as separate replicates (e.g., Rinella and Reinhart 2018). All pots were filled with a 1:1 mixture of prepared field soil and Fafard #2 commercial soil mixture; previous trials using 100% field soil resulted in high seedling mortality in the first three weeks (personal observation). Table 4.1 Local adult abundance, shade tolerance, seed weight, and primary mycorrhizal association for each of the study species. 1Local adult abundance was calculated as number of individuals ≥ 5cm dbh/ha at Alma College’s Ecological Preserve. 2Shade tolerance is presented as intolerant, intermediate, or tolerant and as mean ± std. dev., on a standardized scale from 1 (least tolerant) to 5 (most tolerant), calculated by Niinemets & Valladares (2006). 3Seed weight data was collected from Burns and Honkala (1990). 4AMF = arbuscular mycorrhizal fungi and EMF = ectomycorrhizal fungi. Species Acer rubrum Acer saccharum Prunus serotina Populus grandidentata Quercus alba Quercus rubra Table from Chapter 3. Local adult abundance1 131 285 4.33 Shade tolerance2 (3.44 ± 0.23) (4.76 ± 0.11) (2.46 ± 0.34) 82.33 (1.21 ± 0.27) 12.67 71.67 (2.85 ± 0.17) (2.75 ± 0.18) Seed weight (mg)3 19.7 64.9 94.3 0.2 6,677 4,127 Mycorrhizal association4 AMF AMF AMF AMF, EMF AMF, EMF AMF, EMF 65 To test for plant-soil feedbacks (PSFs), I compared seedling survival in non-sterilized (live) soil collected beneath conspecific versus heterospecific adult trees. I did not use multi- stage greenhouse culturing (Bever et al., 2010), because in-situ natural culturing already occurred for these long-lived trees and should more closely characterize PSFs occurring in the field. To test for biotic components of these PSFs, I compared seedling performance in sterilized versus non-sterilized (live) soils collected beneath conspecific adults. Soil was sterilized by gamma irradiation (30-70 kGy; Sterigenics International, Schaumburg, IL, USA) and allowed to rest for at least one month to minimize post-fertilization nutrient spikes. Gamma irradiation is highly effective at killing soil microorganisms and usually has minimal effects on soil chemical and physical properties (McNamara et al., 2003). There was almost no seedling colonization by AMF or EMF in sterilized soils (mean = 0.02%, df = 629, t = 84.8, p < 0.01), confirming that my sterilization methods were effective. Light availability I grew seedlings at three light levels (~2%, 15%, and 30% full sun), which represent a typical light range experienced in Michigan forests (Schreeg et al., 2005). I created light treatments by covering greenhouse benches with an inner layer of black shade cloth and an outer layer of reflective knitted poly-aluminum shade cloth (BFG Supply, Burton, OH, USA). I confirmed light levels using PAR (photosynthetically active radiation) measurements at each bench with a LI-COR 205A quantum sensor (LI-COR, Lincoln, NE, USA) on a uniformly overcast day. Seedling planting and measurement Pots were set up on nine different benches in the greenhouse, where all combinations of species and soil source were represented, with three benches per light treatment. I planted 30 66 seedlings per species ´ soil source ´ light treatment, for a total of 3,150 seedlings. A single seed with a newly-emerged radicle was planted into each (655 cm3) deepot (Stuewe and Sons, Tangent, Oregon, USA). To minimize disease from non-experimental soil sources, seeds were surface sterilized with 0.6% NaOCl solution prior to stratification and germination. To avoid cross-contamination, all tools and surfaces that were exposed to soil were soaked in 10% NaOCl solution or surface sprayed with 70% EtOH. I censused seedling survival twice per week for 12 weeks. Mortality at the first two censuses after planting were attributed to transplant shock; these seedlings were not used in subsequent analysis and pots were re-planted. After three weeks of growth, I harvested six seedlings per treatment combination to measure mycorrhizal colonization and defense/recovery traits using established protocols. Half of the seedlings were allocated to measurement of mycorrhizal colonization (roots), phenolics (hypocotyl), and lignin (stem). The other half of the seedlings were allocated to measurement of NSC (stem and root). To quantify mycorrhizal colonization, prior to drying seedlings, 5-10 root fractions per individual (1 cm sections of wet root), were retained, weighed, and stained with 5% Schaeffer black ink and vinegar solution (Vierheilig et al., 1998). Percent root colonization by AMF was quantified by inspecting 100 grid intersections for AMF structures (i.e., vesicles, arbuscules, coils, and hyphae) every 1 mm at 200x magnification (McGonigle et al., 1990). AMF fungal structures were distinguished from other fungi that can inhabit the root interior (e.g., dark septate fungi) by comparing slides to established reference images. Percent root colonization by EMF was quantified by counting the number of intact root tips with and without Hartig nets at 100x magnification every 2 mm along the root until 100 root tips were scored. 67 To quantify phenolics, I collected hypocotyl samples, cut into <1 mm pieces. I extracted phenolics in 5 mL methanol in the dark for 16 hours at room temperature. The methanol extracts were filtered and adjusted to 5 mL, and then I quantified total phenolics using a microplate- adapted colorimetric total phenolics assay with Folin-Ciocalteu reagent (Ainsworth & Gillespie, 2007; P. G. Waterman & Mole, 1994). To quantify lignin, root and stem samples were lyophilized and coarsely ground at 1 mm using a Wiley Mill. We ran 0.5 g root and stem samples through a series of extractions using an ANKOM Fiber Analyzer (ANKOM Technologies, Macedon, NY, USA). I used a neutral detergent fiber extraction to wash off soluble cell contents (e.g., carbohydrates, lipids, pectin, starch, and soluble proteins). I then used an acid detergent fiber extraction with 1.00 normal sulfuric acid to wash off hemicellulose and bound proteins and an acid detergent lignin extraction with 72% sulfuric acid to wash off cellulose, leaving only lignin and recalcitrant materials. Finally, I ashed the samples to quantify dry mass lignin. To quantify nonstructural carbohydrates (NSC), I analyzed stem samples, using a standardized enzyme method for sugar and starch extraction and quantification (Landhäusser, Chow, Turin Dickman, et al., 2018; Quentin et al., 2015). I dried seedling stems and peach leaf standard reference material (MillporeSigma-NIST1547) at 60 °C overnight to remove moisture. I then weighed out 30 mg of each for analysis and separated sugars and starches with hot ethanol extraction. I used a-amylase and amyloglucosidase to convert starch to glucose. I quantified sugars using phenol-sulfuric acid colorimetric assay and starches using a glucose-hexokinase colorimetric assay (MillporeSigma-GAK20). I calculated total NSC as the sum of soluble sugar and starch concentrations derived from the assays. 68 Statistical analysis To analyze seedling survival over the 12 week growing period, I used an individual based counting process in a Cox survival model (Burnham & Anderson, 2002; McCarthy-Neumann & Ibáñez, 2012). Data for each seedling 𝑖 and each time 𝑡, 𝑁!", was coded as 0 until the seedling was found dead, 𝑁!" = 1. I used a count process to model the number of events (mortality, 𝑁!") until the experiment ended at nine weeks. I modeled the likelihood as: and the process as: 𝑁!" ~ 𝑃𝑜𝑖𝑠𝑠𝑜𝑛(λ!") λ𝑖𝑡 = ℎ"𝑒($"), where parameters were estimated as a function of the hazard (ℎ), which is the intrinsic rate of mortality due to individual age or time within the experiment, and of risk (𝜇), which is the extrinsic rate of mortality due to light availability and soil source. Risk (µ) was modeled as an interaction between species and light, plus an interaction between species and soil source, and the random effects of bench). Simulations (3 chains) were run until convergence of the parameters was ensured (25,000 iterations) and then run for another 50,000 iterations, from which the posterior distribution of parameter values and predicted survival were estimated. Predicted survival values were used to assess whether there were differences in how species responded to soil sources and light levels. I then used predicted survival values and their associated uncertainty to test if there were differences in how species responded to low versus high light and in different soil sources. I calculated PSFs as the difference in survival between conspecific and heterospecific soil sources at each light level and for each species. I calculated the biotic effect of the soil (i.e., soil-borne microbes) as the difference in survival between live 69 conspecific and sterilized conspecific soils. Differences that did not include zero in their 95% credible intervals were considered statistically significant (Kruschke, 2014). I used linear mixed effects models to investigate how light availability and soil source influenced measured mycorrhizal colonization and defense/recovery traits. Each trait (i.e., phenolics, lignin, and NSC) was evaluated in a separate model. I analyzed all traits with species, soil source, and light availability as potentially interacting fixed effects, and greenhouse bench and adult tree as random effects. I estimated marginal means for post-hoc analyses of these models. To evaluate the effects of mycorrhizal colonization and defense/recovery traits on PSFs, I also used linear models. Since PSFs are a comparison of seedling performance between conspecific and heterospecific soils, the conspecific value (both in seedling survival, as well as traits) is a constant value per species. Thus, the trait data used in these models was from seedlings grown in soils cultured by heterospecific adults. I analyzed all models at the species level, with mean PSF for each treatment (soil source ´ light level) evaluated as a response of the mean amount of each type of mycorrhizal colonization or defense/recovery trait. I excluded sterilized conspecific soil controls in these analyses, since AMF and EMF colonization, in addition to amounts of phenolics for AM species, in the sterilized conspecific soils were effectively zero. Seedlings were destructively harvested for measurement of mycorrhizal colonization and traits; therefore, I could only compare means of these values to the mean PSFs experienced (survival in conspecific versus heterospecific soils at the end of the 12-week period). All analyses were performed with R version 3.5.1 (R Core Team, 2020). I used the “rjags” package (Plummer et al., 2023) to fit survival models and to run predicted survival and contrast simulations. I used the lme4 package (D. Bates et al., 2015) to evaluate linear models. I 70 tested the significance of main effects using a likelihood ratio test with the “Anova” function. I tested for multicollinearity variance inflation factors using the “vif” function in the car package (Fox & Weisberg, 2019). Post-hoc Tukey pairwise comparisons of significant main effects and comparisons of estimated marginal means were made using the “emmeans” function in the multcomp package (Hothorn et al., 2008; Lenth, 2020). RESULTS AM seedlings experienced negative PSFs and EM seedlings experienced positive PSFs Overall, survival was relatively low for the AM seedling when grow in live conspecific soil and soils cultured by AM adults, while survival was relatively high for EM seedlings, especially when grown in conspecific soil and soils cultured by EM adults. AM seedlings experienced negative PSFs and EM seedlings experienced positive PSFs, and PSFs occurred when there was mismatching of mycorrhizal type (H1) PSFs differed between AM and EM host species (Figure 4.1). Across light availability, all AM seedlings (A. rubrum, A. saccharum, and P. serotina) experienced negative PSFs (lower survival in conspecific than in heterospecific soils). In contrast, EM seedlings (Q. alba and Q. rubra) experienced positive PSFs (higher survival in conspecific than in heterospecific soils). Although not all comparisons were statistically significant. Negative PSFs experienced by AM seedlings almost always occurred when heterospecific soils were cultured by EM adults, while statistically significant positive PSFs in EM seedlings took place when comparing with heterospecific AM soils (Figure 4.1). Overall, Q. alba and Q. rubra seedlings had the highest survival, followed by A. saccharum, A. rubrum, and P. serotina (Table A4.2). 71 Figure 4.1 Differences in predicted seedling survival in conspecific versus heterospecific live soil when grown at low light availability. Data are means with 95% credible intervals; credible intervals that do not overlap with the zero line are statistically significant (indicated with stars, *). Differences in survival above the zero line indicate a positive PSF (higher survival in conspecific than heterospecific soils); differences in survival below the zero line indicate negative PSFs (lower survival in conspecific than heterospecific soils). As an example, A. rubrum experienced lower survival in soils cultured by A. rubrum adults than in soils cultured by Q. alba or Q. rubra adults (i.e., negative PSF). 72 AM seedlings experienced greater negative PSFs in low light (H2) Negative PSFs (i.e., differences in survival between conspecific and heterospecific soils) experienced by AM species were often of greater magnitude (i.e., more negative) in low than high light availability (Figure 4.2). In contrast, positive PSFs experienced by EM species were often more positive in low than high light availability. A. rubrum seedlings experienced up to 32% mortality due to negative PSFs in low light versus 22% in high light. A. saccharum seedlings experienced up to 28% mortality due to negative PSF in low light compared to 8% in high light. P. serotina seedlings experienced up to 30% mortality due to negative PSFs in low light compared to 16% in high light. Q. alba seedlings experienced up to 21% greater survival due to positive PSFs in low light versus 10% in high light. Similarly, Q. rubra seedlings experienced up to 28% greater survival due to positive PSFs in low light compared to 4% in high light. Figure 4.2 PSFs at low versus high light availability. Data are means for each species ´ soil combination (means ± 95% credible intervals). Sterilized conspecific soils are not included. The solid line indicates a one-to-one relationship between PSFs at low light and PSFs at high light. Points above the line demonstrate that seedlings experience more positive PSFs at low than high light; points below the line demonstrate more positive PSFs at high light. 73 Soil-borne microbes in conspecific soils had negative effects on AM seedling survival (H3) Soil-borne microbes in conspecific soil reduced survival for all three AM species (A. rubrum = -30%, A. saccharum = -23% and P. serotina = -40% survival in live versus sterilized conspecific soils). In contrast, there was no difference in survival between live versus sterilized conspecific soils for the two EM species (Figure 4.3). Figure 4.3 Differences in predicted seedling survival in non-sterilized versus sterilized conspecific soils and when grown at average light availability. Data are means with 95% credible intervals; credible intervals that do not overlap with the zero line are statistically significant. Differences in survival below the zero line indicate a negative effect of soil biota (lower survival in non-sterilized than sterilized conspecific soils). Here, A. rubrum, A. saccharum, and P. serotina experience lower survival in non-sterilized than sterilized conspecific soils, indicating a negative effect of soil-borne microbes. 74 AMF colonization was highest in high light availability and in soils cultured by adults of the same mycorrhizal type (H4) Percent AMF colonization was highest when seedlings were grown in high light availability (Table A4.2). For A. saccharum and P. serotina, percent AMF colonization increased from 45% to 54% (t = 6.9, df = 91, p < 0.01) and from 45% to 54% (t = 7.9, df = 89, p < 0.01), respectively, in low versus high light. For Q. alba, percent AMF colonization increased from low to average light availability (13%; t = -3.1, df = 91, p < 0.01), and colonization in medium and high light did not significantly differ (43% at medium and 42% at high light; p > 0.05). Likewise, for Q. rubra, percent AMF colonization increased from low to average light availability (13%; t = 5.8, df = 91, p < 0.01), but colonization in medium and high light did not differ (44% at medium and 47% at high light; p > 0.05). Also, in agreement with hypothesis 5, seedling percent AMF colonization was highest in soils cultured by conspecific adults and was higher in soils cultured by AM heterospecific adults than EM heterospecific adults (Table 4.2). For A. rubrum, percent AMF colonization was highest in conspecific soils (42%) and lowest in heterospecific soils cultured by EM adults (32%; F2,382 = 26, p < 0.01). For A. saccharum, AMF colonization also was highest in conspecific soils (71%) and lowest in heterospecific soils colonized by EM adults (62%; F2,370 = 18.5, p < 0.01). Additionally, for P. serotina, percent AMF colonization was highest in conspecific soils (70%) and lowest in heterospecific soils cultured by EM adults (60%; F2,380 = 30.6, p < 0.01). Conversely, for Q. alba, percent AMF colonization was highest in conspecific soils (44%) and heterospecific soils cultured by EM adults (41%; F2,380 = 10.4, p < 0.01). For Q. rubra, there was no effect of soil source on percent AMF colonization. Seedling AMF colonization increased with light availability for both AM 75 species (from 39 to 45%; and EM species (from 29 to 33%). Overall, AMF seedlings had higher AMF colonization than EMF seedlings (Table A4.2). There was no effect of soil source on EMF colonization. Overall, Q. alba and Q. rubra had the highest amount of EMF colonization (Table A4.2). Defense and recovery traits generally increased with light availability, but soil source effects varied and often depended on seedling mycorrhizal type In agreement with hypothesis 6, seedling phenolics, lignin, and NSC generally increased with light availability for both AM and EM species (Table A4.2). Seedling phenolics (nmol Gallic acid equivalents per mg dry extract) increased in high versus low light for all five study species: P. serotina (100% increase; F2,72 = 16.9, p < 0.01), Q. alba (20% increase; F2,72 = 69.1, p < 0.01), and Q. rubra (20% increase; F2,72 = 29.2, p < 0.01). For A. rubrum and A. saccharum, phenolic values in low light were essentially zero, but increased to 0.07 nmol (F2,72 = 5.2, p < 0.01) and 0.06 nmol (F2,74 = 9.7, p < 0.01), respectively, at high light availability. Lignin (percent dry mass) increased in high versus low light: A. rubrum (53% increase; F2,102 = 38.2, p < 0.01), A. saccharum (20% increase; F2,105 = 17.1, p < 0.01), P. serotina (60% increase; F2,102 = 46.9, p < 0.01), Q. alba (7% increase; F2,102 = 9.8, p < 0.01), Q. rubra (6% increase; F2,102 = 6.5, p < 0.01). Similarly, NSC (percent dry mass) also increased in high versus low light for A. rubrum (47% increase; F2,61 = 49.7, p < 0.01), A. saccharum (13% increase; F2,61 = 5.4; p < 0.01), P. serotina (51% increase; F2,61 = 55.1, p < 0.01), Q. alba (14% increase; F2,61 = 18, p < 0.01), and Q. rubra (16% increase; F2,61 = 14.3, p < 0.01). The effect of soil source on seedling phenolics (nmol Gallic acid equivalents) varied. For A. rubrum, phenolics were highest in live (0.19 nmol) and sterilized (0.06 nmol) conspecific soils and were essentially zero in heterospecific soils (F5,273 = 44.2, p < 0.01; Table A4.2). For A. 76 saccharum, phenolics were highest in live conspecific soils (0.19 nmol) and were essentially zero in sterilized conspecific and heterospecific soils (F5,284 = 2.6, p = 0.02). For Q. rubra, effects of soil source on seedling phenolics depended on light availability: phenolics were lowest in live conspecific soils at low light (2.5 nmol) and were highest in live conspecific soils at high light (4.3 nmol; F10,505 = 6.2, p < 0.01). For P. serotina and Q. alba, there was no effect of soil source on phenolics. Lignin was highest in conspecific soils for the EM seedlings. For Q. alba, percent dry mass lignin was highest in live conspecific soils (23%), followed by heterospecific-EM (21%), heterospecific-AM (21%), and sterilized conspecific (21%) soils (F5,320 = 4.7, p < 0.01; Table 2). For Q. rubra, lignin was highest in conspecific soils, regardless of if they were live (25%) or sterilized (25%), followed closely by heterospecific-AM (24%) and heterospecific-EM (24%) soils (F5,320 = 3.7, p < 0.01). There was no effect of soil source on lignin for any of the three AM species. NSC was lowest in conspecific soils for both EM species, with the greatest differences occurring in low light availability. For Q. alba, at low light, percent dry mass NSC was 18% in conspecific soils and 19% in heterospecific soils (F10,510 = 2, p = 0.03; Table 2). For Q. rubra, at low light, NSC was 11% in conspecific soils and 15% in heterospecific soils (F10,510 = 3, p < 0.01). There was no effect of soil source on NSC for any of the three AM species. Overall, Q. alba and Q. rubra had higher amounts of phenolics, lignin, and NSC than A. saccharum, A. rubrum, and P. serotina (Table A4.2). 77 Table 4.2 Differences in amounts of seedling mycorrhizal colonization and defense (phenolics and lignin) and recovery (NSC) traits (% difference) in live conspecific soils compared to sterilized conspecific, heterospecific AM, and heterospecific EM soils (only significant % differences are provided). A positive % difference indicates a higher amount, and a negative % difference indicates a lower amount of the traits, in conspecific versus compared soils. Where there was a significant difference between soil sources, but this effect varied with light availability, comparisons at different light levels are also provided. Trait AMF Conspecific vs. Soil Source St. Con. Het. AM Het. EM EMF Phenolics St. Con. St. Con. ACRU †† ACSA †† PRSE †† QUAL †† QURU †† Low: 19.4% Low: 16.8% High: 35.5% Low: 22.0% Avg: 39.6% High: 35.1% Low: 28.1% High: 11.5% Avg: 15.2% High: 10.3% Avg: 28.1% High: 19.7% High: 38.9% †† †† Low: 45.5% Het. AM Het. EM +%† +%† +%† +%† High: +%† High: 24.6% High: 15.4% Low: -31.3% High: 24.4% Low: -26.5% High: 13.5% Lignin St. Con. Het. AM Het. EM Het. AM Het. EM NSC Low: 11.1% Low: 10.1% Low: -23.5% Low: -25.1% † Indicates that phenolics values in one of the comparisons was essentially zero. “+%” indicates that phenolics were higher in the conspecific soil. †† Indicates that the comparison for mycorrhizal colonization is against sterilized conspecific soils, which were essentially zero. 10.5% 10.1% 9.6% Avg: -17.4% Avg: -17.7% 78 Increasing AMF and defense/recovery traits were associated with more neutral PSFs As amounts of AMF colonization and defense/recovery traits in heterospecific soils increased, PSFs became less negative for the AM seedling species, but, in conflict with H7, became less positive for the EM seedlings (Figure 4.4). For A. rubrum, PSFs became less negative as AMF colonization (F1,80 = 9.66, p = 0.003) in heterospecific soils increased. For A. saccharum, PSFs became less negative as AMF colonization (F1,80 = 9.43, p = 0.02), phenolics (F1,80 = 8.53, p = 0.005), lignin (marginally-significant at F1,80 = 3.35, p = 0.07), and NSC (marginally-significant at F1,80 = 3.27, p = 0.07) increased in heterospecific soils. For P. serotina, PSFs became less negative as AMF (F1,80 = 13.83, p < 0.001) in heterospecific soils increased. For Q. alba, PSFs became less positive as AMF colonization (F1,80 = 4.80, p = 0.03) in heterospecific soils increased. For Q. rubra, PSFs became less positive as AMF colonization (F1,80 = 7.61, p = 0.007) and lignin (marginally-significant at F1,80 = 3.65, p = 0.07) in heterospecific soils increased. It is important to note that the above trend was found for increasing amounts of mycorrhizal colonization and defense/recovery traits when seedlings were grown in heterospecific soils and does not account for amounts of traits in conspecific soils. As an example, increasing percent colonization by AMF in heterospecific soils was associated with less negative PSFs (less mortality in conspecific versus heterospecific soils) for the AM species and less positive PSFs for the EM species. In general, AM seedlings had higher percent root colonization by AMF in conspecific than heterospecific soils. As amounts of colonization in heterospecific soils increased to levels comparable to the conspecific soils, differences in survival between these two soil sources (i.e., PSFs) decreased. 79 Figure 4.4 Effects of AMF colonization and defense/recovery traits (phenolics, lignin, and NSC) on seedling PSFs. Each point is a mean mycorrhizal colonization or trait value and PSF (survival in conspecific versus heterospecific soil) for a given species ´ soil ´ light combination. Trait values for conspecific soils are used for reference; they are larger in size and lie along the y = 0 axis. Species are distinguished with line color, and light level is distinguished with point shape. Regression line significance is indicated with linetype (solid = p < 0.05, dashed = p < 0.1). As an example, as percent colonization by AMF in heterospecific soils increases, PSFs for A. rubrum, A. saccharum, and P. serotina become less negative and PSFs for Q. alba and Q. rubra become less positive. 80 DISCUSSION Disentangling the mechanisms underlying differences in AM and EM seedling PSFs is critical to predicting forest community dynamics. My results show that, not only does the mismatching of mycorrhizal type drive negative PSFs for AM seedlings and positive PSFs for EM seedlings, but higher amounts of seedling defense and recovery traits may effectively neutralize these PSFs, mediated by light availability. Mismatching of mycorrhizal type drives negative PSFs for AM seedlings and positive PSFs for EM seedlings In this study, the strength and direction of PSFs could be explained by mismatching of mycorrhizal type between the juvenile and adult tree. AM species experienced negative PSFs in soils cultured by AM adults and neutral PSFs in soils cultured by EM adults. In contrast, EM species experienced positive PSFs in soils cultured by EM adults and neutral PSFs in soils cultured by AM adults. Differences in AM and EM seedling responses may be due to differences in protection against pathogens conferred by AMF and EMF. Soils cultured by AM adults have a higher abundance of pathogens (Eagar et al., 2022, 2023), but seedlings growing in those soils also have higher levels of pathogen infection (Chen et al. 2019). My results are consistent with field observations in Michigan forests, where there is an increasing abundance of maple (Acer) and black cherry (Prunus) seedlings in oak (Quercus) understories. Canopy oak replacement by maples may be driven by differences in species ability to take advantage of canopy gaps (Allen et al., 2018), wherein more shade tolerant maple and cherry seedlings can successfully recruit beneath oak canopies. Differences in seedling recruitment patterns may also be driven by mismatching of mycorrhizal type, as found in this study. I found that maple and black cherry experience negative PSFs when grown beneath 81 crowns of those same species (i.e., negative PSFs when growing beneath AM adults). However, when growing beneath oak (EM species) crowns, they experience neutral PSFs. Thus, by growing beneath oak canopies, maple and black cherry seedlings may be able to escape pathogens and/or negative effects of AMF that drive lower survival in those soils. Escape from negative soil-borne microbes likely acts in conjunction with differences in species shade tolerance (Allen et al., 2018), enabling AM species to gain a competitive advantage in EM- dominated forests. It is important to note that, if I had grouped AM- and EM-heterospecific adults in my analyses, or only included heterospecific adults matching the seedlings’ mycorrhizal type, I would have found more neutral PSFs overall, thereby limiting my interpretation of both the strength and direction of the actual PSFs. AM seedlings experienced neutral to negative PSFs, and EM seedlings experienced neutral to positive PSFs (Figure 4.1). Previous studies have also reported that AM species generally experience negative PSFs and EM species experience positive PSFs (Bennett et al., 2017; van der Putten et al., 2013). However, to disentangle the mechanisms underlying PSFs, heterospecific soils must also be evaluated by mycorrhizal type (Kadowaki et al., 2018). PSFs became more neutral as light availability increased PSFs became more neutral as light availability increased (Figure 4.2). For AM seedlings, PSFs were more negative at low light availability. Negative PSFs in low light could be driven by soil-borne pathogens, which are often more abundant in cooler, wetter conditions associated with shade (Hersh et al., 2012; Y. Liu & He, 2019; Reinhart et al., 2010). Moreover, AMF symbioses could also shift from mutualistic to parasitic at low light (Ibáñez & McCarthy-Neumann, 2016), thus reducing seedling photosynthates and exacerbating negative effects of pathogens on 82 seedling survival. In a previous experiment, I found that the presence of AMF can decrease Acer seedling survival, even when soil-borne pathogens are not present (Chapter 2), signifying that AMF can have detrimental impacts on seedling survival comparable to those of pathogens, when resources are limited. Conversely, PSFs for EM seedlings in this study were more positive at low light availability and became less positive as light increased. To my knowledge, there are no instances in which EMF act parasitically (but see Ågren et al., 2019). I propose that EMF were less beneficial for seedling survival at high light availability, where resources were more readily available and pathogens were likely less abundant. At low light availability, the physical protection against pathogens conferred by EMF may result in higher seedling survival, whereas, at high light availability, benefits of EMF may manifest in higher growth. Together, these results indicate that PSFs can enhance performance of AM seedlings in higher light conditions, such as canopy gaps, but favor EM species under closed canopies. Whereas AM seedlings typically experience negative PSFs when grown under low light conditions, increasing light availability also increases survival, thereby decreasing differences between conspecific and heterospecific soils. For EM species, however, survival responses are higher under low light availability, allowing them to persist in shaded conditions, even when soil-borne pathogens are more prevalent. Differences in survival in low light availability (i.e., shade tolerance) only occur when soil-borne microbes are present (Chapter 2), which could enhance differences in niche across light gradients. Soil-borne microbes appear to drive PSFs for AM seedlings Soil-borne microbes appear to drive the PSFs observed in this study, but only for AM seedlings (Figure 4.3). I found that AM seedlings experienced lower survival in live than 83 sterilized conspecific soils, indicating that the net effect of soil-borne microbes in conspecific soils was detrimental to seedling health. Soils cultured by AM adults may encourage higher amounts of pathogens than those cultured by EM adults (Eagar et al., 2022, 2023), resulting in higher seedling mortality. For all three AM seedling species, percent AMF colonization was highest in conspecific soils and heterospecific soils cultured by AM adults (Table 4.2). Percent colonization by AMF was higher in conspecific soils. For AM species, this result is consistent with previous studies that also found that AMF and EMF colonization is higher in soils cultured by adult trees of the same mycorrhizal type (Bennett et al., 2017; Chen et al., 2019; Liang et al., 2016). However, unexpectedly for Q. alba, an EM species, percent AMF colonization was also highest, not in the heterospecific soils cultured by AM adults, but in conspecific soils and heterospecific soils cultured by EM adults. Soils contain mycorrhizal genotypes that are well-suited to colonizing the adult tree species growing in them (Chen et al., 2019; Segnitz et al., 2020; J. Yang et al., 2018), which could explain the higher amounts of AMF colonization in conspecific soils. Higher amounts of AMF colonization could also be explained by differences in mycorrhizal life history strategies. The greenhouse pots would have excluded any potential common mycorrhizal networks, linking adult and juvenile trees in the forest (Bücking et al., 2016; Simard & Durall, 2004). Contrary to expectation, I did not find similar patterns for EMF colonization; there was no effect of soil source on EMF colonization for the two EM species. I expected to see higher percent EMF colonization, especially in conspecific soils, due to the assumed higher host- specificity of EMF versus AMF (Chen et al., 2019; S. Smith & Read, 2008). Overall, these results suggest that AMF may be more specialized than expected, even when associating with EM trees species. Furthermore, soils in the greenhouse pots likely selected for more ruderal 84 species of mycorrhizae, which are often dominated by AMF (Gao et al., 2023; García de León et al., 2016). Together, these results support the idea that mycorrhizal communities associating with oak species shift during different stages of forest succession. AMF colonization may predominate early stages of succession, but EMF colonization (especially via common mycorrhizal networks) is more prevalent in older, established stands (Egerton-Warburton & Allen, 2001). In contrast to the AM seedlings, EM seedling survival did not differ between live and sterilized conspecific soils, despite EM seedlings experiencing lower survival when growing near conspecific adults Jevon et al. (2022). I also did not find negative effects of any heterospecific soil sources on EM seedling survival. Lacking growth data, I cannot exclude the possibility that EMF may have beneficial effects on EM seedling growth but have less noticeable effects on survival. I was unable to fully test the biotic mechanisms behind PSFs in this study because I sterilized only conspecific and not heterospecific soils. Thus, I can only draw conclusions about the role of soil-borne microbes in the conspecific soils and speculate about the role of microbes in heterospecifics soils. This is an important distinction, because PSFs (defined here as survival in conspecific versus heterospecific soils) can be driven by either soil source. Positive PSFs may be due to greater abundance of mutualists in conspecific soils or lower abundance of pathogenic microbes in heterospecific soils. For instance, Eagar et al. (2022, 2023) found higher abundance of pathogenic fungi in soils beneath AM adults. Quercus species growing under AM trees may also be infected by these pathogens, resulting in positive PSFs. Thus, the driving factor in this example would be due to the lack of pathogens in conspecific versus heterospecific soils, not necessarily a higher abundance of mutualists. Alternatively, PSFs could also be driven not just 85 by microbes, but also by abiotic factors (McCarthy-Neumann & Ibáñez, 2012). For instance, adult trees may modify nutrient availability or soil moisture, or excrete allelochemicals in the soil beneath their crown, thereby limiting seedling establishment. As differences in percent mycorrhizal colonization and seedling defense/recovery traits between conspecific versus heterospecific soils decreased, PSFs become more neutral Multiple studies have reported that plant traits could be important predictors of PSFs, though, the focus has been on traits promoting faster growth (Baxendale et al., 2014; Cortois et al., 2016; Xi et al., 2021). In this study, I focused on survival-based PSFs and defense/recovery traits that were expected to drive survival differences between conspecific and heterospecific soils that are likely due to soil-borne pathogens (Song & Corlett, 2022). Here, I found that, as differences in mycorrhizal colonization and defense/recovery traits between conspecific and heterospecific soils decreased, differences in survival between these soil sources (PSFs) also decreased (i.e., became more neutral). For AM species, PSFs became less negative, and for EM species, PSFs became less positive as mycorrhizal colonization and amounts of defense/recovery traits increased. However, mycorrhizal colonization and defense/recovery traits also increased with light availability, making it difficult to disentangle the roles of these factors on PSFs. A. saccharum illustrates the importance of traits in neutralizing PSFs. As percent colonization by AMF increased, PSFs became more neutral for all species (i.e., less negative for AM species and more positive for EM species). For the AM species, as amount of colonization in heterospecific soils increased, approaching the levels of colonization in soils cultured by conspecific adults, PSFs became less negative (i.e., seedling survival in heterospecific soils decreased to levels like those in conspecific soils). This result suggests that negative effects of 86 higher amounts of AMF colonization drive the differences in survival in conspecific versus heterospecific soils (i.e., PSFs). Furthermore, as measured amounts of phenolics in heterospecific soils increased to levels present in conspecific soils, PSFs became more neutral. Phenolics production, which acts as a chemical defense against pathogens (Ichihara and Yamaji 2009), can be induced to higher amounts when soil-borne microbes, especially pathogens, are present (Chapter 2). Soil-borne pathogens are often more abundant in conspecific soils, where they are effectively-specialized for seedlings of that species; this aligns with results from this study, in which phenolics were produced in higher amounts in conspecific soils. However, although higher amounts of phenolics are associated with higher seedling survival (Chapters 2 and 3), they may not be enough to completely overcome high mortality caused by pathogens. In heterospecific soils, where there are fewer effectively-specialized pathogens, A. saccharum seedlings both produce fewer phenolics and have higher survival. As amounts of lignin in heterospecific soils increased to levels similar to those in conspecific soils (i.e., the difference in lignin in conspecific versus heterospecific soils decreased), differences in survival between conspecific and heterospecific soils (i.e., PSFs) became less positive. This result suggests that lignin confers defense against soil-borne pathogens and possibly mediates PSFs for EM species. Light availability likely mediates the relationships between amounts of lignin and NSC in heterospecific soils and PSFs, at least for AM species. Together, these results indicate that mycorrhizal colonization and defense/recovery traits can influence the strength and direction of PSFs. As differences in mycorrhizal colonization and defense/recovery traits values between soil sources decrease, so do differences in survival. 87 However, both mycorrhizal colonization and amounts of defense/recovery traits also increase with light availability, making it difficult to disentangle these two factors. Caveats The relationship between defense and recovery traits and seedling survival is correlative and thus may not reflect a causal mechanism. I also was unable to separate the impacts of soil- borne mutualists and pathogens on seedling functional traits and subsequent survival. Similarly, I cannot distinguish between direct AMF effects of pathogen reduction through displacement versus indirect effects inducing production of phenolics or NSC, both of which can enhance seedling survival. I sampled very young seedlings (3 weeks old), which resulted in some systemic measurement error when samples were smaller than typical protocols called for, especially for phenolics. The absolute amounts of phenolics in these seedlings should be interpreted with caution, but we still provide interpretation of relative amounts of phenolics between species and treatments. Also, I was unable to disentangle the effects of seed size and mycorrhizal type, since both EM seedlings were large-seeded Quercus species. I suggest that future studies investigate cooccurring AM and EM species with overlapping seed sizes and potentially separate out the role of mycorrhizae and soil-borne pathogens utilizing a technique, such as the wet-sieving method used in Chapter 2. When evaluating the effects of defense and recovery traits on seedling survival in this study, it is important to note the strong relationship between light availability and those traits. Seedlings growing in higher light availability also had higher amounts of phenolics, lignin, and NSC, likely due to the higher amounts of photosynthates available. Soil sources and associated microbial communities also appear to influence defense and recovery trait production, but not as strongly as light. In this study, I was unable to disentangle the effects of light and traits on 88 seedling survival. However, I can infer that higher amounts of defense and recovery traits likely neutralize PSFs, and these differences are largely driven by light availability. Implications for forest community dynamics Understanding the mechanisms underlying PSFs is key to predicting forest community dynamics. Mismatching of mycorrhizal type between the juvenile and adult tree may shift PSFs and subsequent community dynamics. If AM species experience more negative PSFs (i.e., lower survival beneath conspecific than heterospecific adults), they are more likely to be replaced by heterospecific trees, thereby increasing diversity within that forest. For example, if AM species experience more positive PSFs (i.e., higher survival beneath conspecific than heterospecific adults), they are more likely to be replaced by conspecific trees, thereby decreasing diversity within that forest. The same could be said for EM species, as well. These results on mycorrhizal mismatch between seedlings and cultured soils provide a possible additional explanation behind the widespread transition from oak (Quercus) to shade- tolerant mesophytic tree species, especially maple (Acer), in many eastern United States forests (Abrams, 1996; Knott et al., 2019; Nowacki & Abrams, 2008). Mesophytic trees appear to be making conditions more favorable for their own regeneration (e.g., shadier, cooler, wetter, and with less flammable woody debris) and less favorable for oak generation over time (Alexander et al., 2021; Kreve et al., 2011; Nowacki & Abrams, 2008), which has contributed to limited regeneration and seedling establishment for major oak species (Dey et al., 2008). My results suggest that mesophytic species that are associated with AMF, such as Acer, will have greater survival near Quercus adults (which are associated with EMF) due to the negative PSFs these species experience when establishing under conspecific crowns relative to establishing near 89 Quercus adults. Maple seedlings established in soils cultured by oak adults can escape the higher abundance of pathogens typically present beneath conspecifics. Furthermore, maple seedlings growing in shaded oak understories can escape potentially parasitic relationships with AMF. This escape from harmful microbes, in addition to high shade tolerance, enables maple seedlings to outcompete oak seedlings and shift forest species composition. Likewise, oak establishment is reduced as the abundance of canopy adults of these AMF-associated mesophytic species increases, resulting in fewer areas for oaks to disperse that is associated with EMF where these oak seedings experience enhanced survival (i.e. positive PSF). These findings are consistent with a demographic study in a Michigan hardwood forest where newly established oak trees (>3.2 cm DBH) were less likely to establish near maple and black cherry canopy adults, but those mesophytic species were more likely to recruit near oak canopy adults (Allen et al., 2018). Although these findings were thought to be due to the mesophytic species’ ability to take better advantage of the light levels under oak canopies, my work suggests that their response may be at least partially due to PSFs. Changes in defense and recovery trait levels associated with higher light availability and conspecific soil sources can act to neutralize PSFs. While AM seedlings typically experience negative PSFs, higher amounts of defense and recovery traits can result in more neutral PSFs. Likewise, for EM seedlings, higher amounts of traits result in less positive PSFs. In this study, I demonstrated that seedlings produce greater amounts of phenolics, lignin, and NSC when grown under higher light availability. Additionally, some traits can be induced and produced in higher amounts when conspecific soil-borne microbes are present (also see Chapter 2). While there are several studies investigating the role of growth-related traits on PSFs, it is still unclear what other traits directly related to defense and recovery from soil-borne microbes can also mediate PSFs. 90 I suggest that future studies investigate other potential seedling functional traits and whether they are influenced by matching of mycorrhizal type. Also, studies should investigate the environmental conditions under which these relationships may shift. In addition to light availability, PSFs may shift with climate change, including increased drought, warming (Hassan et al., 2022), and wildfires (Warneke et al., 2023). Moreover, these shifts may depend on mycorrhizal type (Bennett & Klironomos, 2018). This study demonstrates that it is important to consider not only the mycorrhizal type of the seedlings, but also the mycorrhizal type of the trees that cultured the soil in which the seedlings occur (i.e., matching or mismatching mycorrhizal type). 91 CHAPTER 5 Mycorrhizal type and light availability explain differences in biomass response to plant-soil feedback ABSTRACT Plant-soil feedbacks (PSFs) are key drivers of seedling recruitment patterns in forests and thus influence community dynamics. Recent studies have found that PSFs may be mediated by seedling mycorrhizal type (associated with arbuscular mycorrhizal fungi [AMF] or ectomycorrhizal fungi [EMF]. Furthermore, the strength and direction of PSFs may vary when measuring biomass or survival and under different light conditions. To investigate the influence of mycorrhizal type and light availability on PSFs, I conducted parallel greenhouse and field experiments, growing temperate tree species in soils collected beneath adults of those species and under three light levels ranging from shaded understory to light gap. I measured seedling survival, biomass, and colonization by mycorrhizal fungi after one growing season in the greenhouse and two growing seasons in the field. In this study, I found that PSFbiomass is mediated by both mycorrhizal type and light availability. Whereas AM species tend to have negative PSFbiomass that becomes more positive as light increases, EM species experience positive PSFbiomass regardless of light level. Also, there is a negative relationship between PSFbiomass at high light availability and PSFsurvival at low light availability, with differences in seedling mycorrhizal type driving this relationship. Understanding the mechanisms underlying PSFs is key for understanding forest regeneration dynamics. While previous research has demonstrated that strength and direction of PSFs varies with mycorrhizal type (AM vs. EM), I show that variation within mycorrhizal types may be 92 attributed to differences in light availability, especially for AM seedlings. Moreover, tree seedlings express a negative relationship between PSFbiomass at high light availability and PSFsurvival at low light availability, with relative importance of biomass and survival also mediated by mycorrhizal type. Together these results provide a more mechanistic understanding of the biotic and abiotic factors underlying PSFs and subsequent seedling regeneration dynamics. INTRODUCTION Plant-soil feedbacks (PSFs) are key drivers of seedling recruitment patterns in forests and subsequent community dynamics (Crawford et al., 2019; Putten et al., 2016). In forest communities, PSFs are a continuous feedback loop in which adult trees modify properties of the soil beneath their crown, thereby influencing the ability of seedlings to grow and survive in that soil (Bever et al., 1997). The strength and direction of PSFs experienced by seedlings can regulate forest community dynamics by acting as stabilizing or destabilizing mechanisms underpinning species coexistence (Chesson, 2000). PSFs are usually calculated as a comparison of plant performance in conspecific versus heterospecific soils (Bever et al., 1997; Kulmatiski et al., 2008). Potential drivers of PSFs include soil-borne microbes, like mycorrhizae and pathogens (Bever et al., 2010; Jiang et al., 2020). Light availability could also influence microbial abundance and subsequent PSFs (Chapter 3, Chapter 4). PSFs can influence both the survival and growth of tree seedlings. Survival likely has large impacts that peak within the first few months after germination and then declines with age (McCarthy-Neumann & Ibanez 2012), whereas the influence on growth accumulates with age (Dudenhöffer et al., 2018). In addition, PSF responses differ when measured in the greenhouse versus in the field (Forero et al., 2019; 93 Kulmatiski & Kardol, 2008). Here, we compare seedling survival and growth responses to PSFs at both low and high light in both greenhouse and field experiments spanning 16 weeks to 2 years in duration. PSFs are often driven by soil-borne microbes, such as pathogens and mycorrhizal fungi (Bever et al., 2010; Jiang et al., 2020), and interactions with microbes can vary with light availability (McCarthy-Neumann & Ibáñez, 2013; Chapter 2). Soil-borne pathogens, including fungi, oomycetes, and bacteria, can kill entire seedling cohorts (Mangan et al. 2010; Terborgh 2012). Furthermore, pathogens are often more abundant in low light availability, where they can proliferate in cool and damp conditions (Y. Liu & He, 2019). Another group of microbes, mycorrhizal fungi, act as mutualists, exchanging water and nutrients for sugars (S. Smith & Read, 2008). Seedlings typically have higher root colonization by mycorrhizal fungi in high light availability (Bereau et al., 2000; Koorem et al., 2017; Shi et al., 2014). However, at low light availability, where photosynthate production is more limited, they may act parasitically (Konvalinková & Jansa, 2016; McCarthy-Neumann & Ibáñez, 2013). Interactions between soil-borne pathogens and different groups of mycorrhizal fungi can shift the strength and direction of PSFs. At low light availability, the cost of maintaining the mycorrhizal symbiosis may exacerbate the negative effects of pathogens. However, mycorrhizal fungi can also confer protection against pathogens, but the degree of protection depends upon mycorrhizal type (Bennett et al., 2017). Arbuscular mycorrhizal fungi (AMF), which form arbuscules within plant root cells, provide indirect defense against pathogens by competing for space on plant roots (Borowicz, 2001). In contrast, ectomycorrhizal fungi (EMF), which form a Hartig net around plant fine roots, can provide direct defense where the net acts as a protective sheath (Laliberté et al., 2015). 94 Moreover, matching or mismatching of mycorrhizal type (AMF or EMF) between the seedling growing in and the adult tree culturing the soil may influence PSFs. Hereafter, we refer to species that typically associate with AMF and EMF as “AM species” and “EM species”, respectively. Whereas AM trees typically experience negative PSFs (i.e., inhibition of seedlings around conspecific adults), EM trees more often experience positive PSFs (i.e., facilitation around conspecific adults) (Bennett et al., 2017; Kadowaki et al., 2018). In addition, AM trees have a higher abundance of plant pathogens in their soil (Eagar et al., 2022, 2023) and AM seedlings accumulate soil-borne pathogens faster when growing under AM adults (Chen et al., 2019). However, when there is mismatching of mycorrhizal type (e.g., AM seedlings growing beneath EM trees, and vice-versa), both AM and EM seedlings experience positive or neutral PSFs (Kadowaki et al. 2018; Chapter 4). Together, these trends provide a potential explanation why AM seedlings experience more negative PSFs. PSFs are typically quantified as biomass (PSFbiomass) or survival (PSFsurvival). The strength and direction of PSFbiomass and PSFsurvival may depend upon several factors, including seedling mycorrhizal type and light availability. PSFsurvival at low light may be especially important for AM species, which experience more negative PSFs (Bennett et al., 2017; Chen et al., 2019; Kadowaki et al., 2018). In contrast, PSFbiomass may be more important for EM seedlings, which may have higher survival in low light but exhibit larger differences in growth in high light availability. Furthermore, negative relationships between PSFsurvival at low light availability and PSFbiomass at high light availability might provide a mechanistic basis for a growth-defense tradeoff often posited for tree species (Kobe et al., 1995). More negative PSFsurvival (reduced survival in conspecific versus heterospecific soils) is common in low light availability (Chapter 95 3, Chapter 4), where survival against soil-borne pathogens and resilience against parasitic mycorrhizal fungi drive seedling performance. Shade tolerant species are typically less vulnerable to mortality by soil-borne microbes than shade intolerant species (Alvarez-Clare & Kitajima, 2007; McCarthy-Neumann & Kobe, 2010a; Wood, Kobe, et al., 2023) at least partly because shade tolerant species allocate more carbon to traits that confer defense against pathogens (Wood, Kobe, et al., 2023). Similarly, slow-growing species (which are likely to invest carbon in defense, rather than growth) experience more positive PSFs, whereas fast- growing species experience more negative PSFs (Baxendale et al., 2014). My overall conceptual framework is that PSFbiomass is driven by seedling mycorrhizal type and light availability. Additionally, species exhibit a negative relationship between PSFsurvival at low light and PSFbiomass at high light, with AM species experiencing negative PSFsurvival at low light, but having higher PSFbiomass at high light, and vice-versa for EM species. I hypothesized that: 1) AM seedlings experience negative PSFbiomass (lower biomass on conspecific versus EM heterospecific soils) and EM seedlings experience positive PSFbiomass (higher biomass in conspecific versus AM heterospecific soils). 2) AM seedlings experience greater negative PSFbiomass in low than high light. EM seedlings experience greater positive PSFbiomass in high than low light. 3) Soil-borne microbes from conspecific soils have negative effects on biomass for AM species and positive effects on biomass for EM species, which could partly explain patterns of PSFbiomass found in hypotheses 1 and 2. 4) Across species, there is a negative relationship between PSFsurvival at low light and PSFbiomass at high light. 96 This study provides insight into how species mycorrhizal type can mediate PSF responses to light availability. Additionally, I present novel work, comparing biomass and survival across both a greenhouse and field study. Disentangling the impact of both seedling survival and growth will advance understanding of how PSFs can impact seedling recruitment and forest community composition. MATERIALS AND METHODS I conducted parallel factorial blocked field (see Chapter 3) and greenhouse (see Chapter 4) experiments to investigate the above hypotheses. In the field, I grew two temperate tree species in seven soil sources (conspecific live, conspecific sterilized, and five heterospecific), at three light levels, for two growing seasons. In the greenhouse, I grew five species in the same seven soil sources, at three light levels, for one growing season. Throughout each experiment, I monitored seedling survival. At the end of each experiment, I measured biomass and colonization by mycorrhizal fungi (AMF and EMF). Field experiment The field experiment (see Chapter 3) was conducted in 100 ha mixed hardwood forest stand in mid-Michigan, at Alma College’s Ecological Field Station (43°23'32.0"N 84°53'41.5"W). The forest has not been logged since 1897 and lies in a transition zone between northern coniferous and southern deciduous forests. The dominant species in this forest is sugar maple (Acer saccharum), a shade-tolerant canopy tree species. Other common species in the forest include red maple (A. rubrum) and big-toothed aspen (Populus grandidentata). I chose six species: red maple (A. rubrum), sugar maple (A. saccharum), big-toothed aspen (P. grandidentata), black cherry (Prunus serotina), white oak (Quercus alba), and 97 northern red oak (Q. rubra). The six study species vary in local adult abundance, shade tolerance, seed size, and mycorrhizal association type (Table 5.1). Due to difficulty acquiring seeds and/or poor germination, I did not grow A. rubrum seedlings in the field experiment. However, I still included A. rubrum soil as a treatment. Table 5.1 Local adult abundance, shade tolerance, seed weight, and primary mycorrhizal association for each of the study species. 1Local adult abundance was calculated as number of individuals ≥ 5cm dbh/ha at Alma College’s Ecological Preserve. 2Shade tolerance is presented as intolerant, intermediate, or tolerant and as mean ± std. dev., on a standardized scale from 1 (least tolerant) to 5 (most tolerant), calculated by Niinemets & Valladares (2006). 3Seed weight data was collected from Burns and Honkala (1990). 4AMF = arbuscular mycorrhizal fungi and EMF = ectomycorrhizal fungi. Species Acer rubrum Acer saccharum Prunus serotina Populus grandidentata Quercus alba Quercus rubra Table from Chapter 3. Local adult abundance1 131 285 4.33 Shade tolerance2 (3.44 ± 0.23) (4.76 ± 0.11) (2.46 ± 0.34) Seed weight (mg)3 19.7 64.9 94.3 Mycorrhizal association4 AMF AMF AMF 82.33 (1.21 ± 0.27) 0.2 AMF, EMF 12.67 71.67 (2.85 ± 0.17) (2.75 ± 0.18) 6,677 4,127 AMF, EMF AMF, EMF I planted two tree species (Q. alba and Q. rubra) in seven soil sources (sterilized conspecific, live conspecific, and five heterospecific, including A. rubrum, A. saccharum, P. grandidentata, and P. serotina), and a gradient of light levels (ranging from deep shade to light gap), for a total of 1,512 seedlings. I originally planted four species (A. saccharum, P. serotina, Q. alba, and Q. rubra); however, due to low survival of A. saccharum and P. serotina seedlings in the first growing season, I was unable to evaluate biomass or traits for these species. I collected intact soils cores from May to June 2016 and April to May 2017. To minimize potential for multispecies culturing of soil, I took soil cores under adult trees. I took soils from within 1 m of six mature randomly-selected adults for each of the six study species (36 trees 98 total), ensuring that each tree was at least two crown diameters away from adults of other species. I used a custom-made, mechanized soil core sampler (Giddings Machine Co; Windsor, CO, USA) to remove 9 cm ´ 46 cm long intact soil cores. I maintained adults as separate replicates for statistical analysis (Reinhart & Rinella, 2016; Rinella & Reinhart, 2018). Intact soil cores with plastic liners were converted into pots by drilling two 7.5 cm diameter holes into the sides. I also adhered a 0.5 µm nylon mesh over two side holes and the bottom opening for each pot. Small nylon mesh pots are an established method for studying mycorrhizal networks in forests (McGuire 2007; Bingham and Simard 2012; Teste et al. 2017; Chapter 3). The small pores in the mesh prevent roots, fungal hyphae, oomycetes, and pathogenic fungi from passing in or out, but have minimal effects on water and nutrient flows (Allison et al., 2013). Pots were transplanted into eighteen 8.4 m ´ 6.6 m common-garden field plots along a gradient of light availability (0.032 to 0.161 indirect site factor [ISF]), which I grouped into three general light levels (low, medium, high). Existing vegetation and leaves in each plot were removed to minimize potential light interception of pots. I then took precise measurements of light availability by analyzing canopy photos with HemiView software (Delta-T services, Ltd., Burwell, England; Figure A3.5). Sterilized soil treatments were created by exposing a subset of conspecific soils in the plastic liner pots to gamma irradiation (30-70 kGY; Sterigenics International, Schaumburg, IL, USA) in July 2017. I allowed the pots to rest at least once month after irradiation, to minimize post-sterilization nutrient spikes. Gamma irradiation is highly effective at killing soil microorganisms and has minimal effects on both soil chemical and physical properties (McNamara et al., 2003). I also tested the sterilized versus live soil for difference in nutrient 99 availability, using plant root simulator (PRSTM) probes (Western Ag Innovations Inc., Saskatchewan, Canada). I found no effect of sterilization on soil nutrient availability (Tables A3.4 and A3.5; Figure A3.3). In the field, I planted 108 seedlings per species ´ soil source, evenly distributed across the eighteen field plots. I planted a single surface-sterilized seed, with a newly-emerged radicle, in each custom-made pot. I purchased seeds for Q. alba from Sheffields Seed Co (Locke, NY, USA). I collected seeds for all other study species from mid-Michigan forests. I expect that variation among seed source populations in survival, mycorrhizal colonization, and defense/recovery traits (e.g., phenolics, lignin and NSC) was minimal (McCarthy-Neumann & Ibáñez, 2012). Additionally, in June 2018, one week prior to planting seeds, I added 1 cm of a 1:1 mixture of peat moss and fresh or sterilized soil. This soil amendment was added to increase transplant success and provide fresh inoculum to the seedlings. In a previous trial run, I found that seedlings planted with peat moss and fresh soil experienced less transplant shock (personal observation). To minimize disease from non-experimental sources, seeds were surface-sterilized with a 0.6% NaOCl solution prior to stratification and again prior to germination. Also, to avoid cross- contamination, all tools and surfaces that were exposed to soil were soaked in 10% bleach or surface sprayed with 70% EtOH and then rinsed with deionized water. To minimize browsing and excavation of seedlings by vertebrates, I constructed enclosures around each field plot from galvanized hardware cloth (6 cm ´ 6 cm) to 1.8 m. To the top of each pot, I glued 0.25 cm ´ 0.25 cm hardware cloth. Seedlings likely did not experience significant shading from the hardware cloth, and most seedlings grew above the cloth within 2 weeks of planting. 100 After planting, I censused seedling survival twice per week for 16 weeks (one growing season). I then re-censused seedlings at the end of the second growing season and quantified biomass for all surviving seedlings at the end of the experiment. Seedling mortality in the field was exceptionally high, especially for A. saccharum and P. serotina (see Chapter 3), with almost 100% of seedlings for both species dying by the end of the growing season. Mortality for Q. alba and Q. rubra was lower (about 50% of seedlings), but was still relatively high for these species. High mortality was likely due to a dry period preceding a large rainfall event (7” in one week). Due to the seedling pots' design, drying caused soil to pull away from the edges, which then flooded during heavy rainfall. Greenhouse experiment I conducted a parallel greenhouse experiment (see Chapter 4) at the Michigan State University Tree Research Center in Lansing, MI, USA (42.7 ºN, 84.5 ºW). I planted five tree species (A. rubrum, A. saccharum, P. serotina, Q. alba, and Q. rubra). Due to difficulty acquiring seeds, I did not grow P. grandidentata seedlings. However, I still included P. grandidentata soil as treatments. I grew seedlings in seven soil sources (sterilized, conspecific, live conspecific, and five heterospecific). I collected soils from the Alma College Ecological Field Station in August 2017 (top 15 cm from within 1 m of the adult tree bole), using the same species and adult trees as the field experiment. I prepared the soils by dicing roots and sifting them through a 1 cm mesh sieve, retaining all fine roots and maintaining soil from each adult as separate replicates (Reinhart & Rinella, 2016; Rinella & Reinhart, 2018). All pots were filled with a 1:1 mixture of prepared field soil and Fafard #2 commercial soil mixture. Previous trials using 100% field soil resulted in high seedling mortality in the first three weeks (personal observation). I did not use multi-stage 101 greenhouse culturing (Bever et al., 2010), since in-situ natural culturing had already occurred and likely more accurately represented PSFs experienced in the forest. Like in the field experiment, I sterilized a subset of conspecific soils by gamma irradiation (30-70 kGy; Sterigenics International, Schaumburg, IL, USA) and allowed the soil to rest for one month before planting, to minimize post-sterilization spikes in nutrient availability. There was almost zero seedling colonization by mycorrhizal fungi in sterilized soils (mean = 0.02%, df = 629, t = 84.8, p < 0.01), confirming that my sterilization methods were effective. I grew seedlings at three light levels (2%, 15%, and 30% sun), representing the typical light range experienced by Michigan forests (Schreeg et al., 2005). I created light treatments in the greenhouse by covering benches with an inner layer of black shade cloth and an outer layer of reflective knitted poly-aluminum shade cloth (BFG Supply, Burton, OH, USA). I confirmed light levels using PAR (photosynthetically active radiation) methods at each bench with a LI- COR 205A quantum sensor (LI-COR, Lincoln, NE, USA) on a uniformly-overcast day. I set up pots on nine different benches in the greenhouse, where all combinations of species and soil source were represented, with three benches per light treatment. I planted 30 seedlings per species ´ soil sources ´ light treatment, for a total of 3,150 seedlings. I planted a single seed with a newly-emerged radicle into each pot. To minimize disease from non- experimental soil sources, seeds were surface sterilized with 0.6% NaOCl solution prior to stratification and germination. To avoid cross-contamination, all tools and surfaces that were exposed to soil were soaked in 10% NaOCl solution or surface sprayed with 70% EtOH and then rinsed with deionized water. I monitored seedling survival twice per week for the equivalent of one growing season (16 weeks). I also measured seedling biomass at the end of the experiment. 102 Statistical analysis I used linear mixed effects models to investigate how soil source and light availability influence seedling biomass. Models were run for each species, with soil source and light availability as fixed effects, and greenhouse bench and adult tree as random effects. PSFbiomass was calculated using the response-ratio of biomass in conspecific versus sterilized conspecific or heterospecific soils. I then log10 transformed the response-ratios to get a relative measure of PSFs, independent of differences in biomass across species. PSFbiomass was calculated using bootstrapping (S. E. Bates et al., 2020). To evaluate low light survival versus high light growth, I compared PSFsurvival at low light availability to PSFbiomass at high light availability (sample sizes for each seedling species ´ soil source ´ light level in the field and greenhouse experiments are available in Table A5.1 and Table A5.2). PSFsurvival values for the field and greenhouse were previously calculated in Chapters 2 and 3, respectively. For the field experiment, I analyzed seedling survival over 16 weeks with frequentist Cox proportional hazards regression (Cox & Oakes, 2017). I ran species- specific models, using soil source and light availability as fixed effects, and plot and adult tree as random effects. The best fitting models for seedling survival did not include any interactions. For the greenhouse experiment, I calculated PSFsurvival by using an individual based counting process in a Bayesian Cox survival model (see Chapter 4; Burnham and Anderson 2002; McCarthy-Neumann and Ibáñez 2012). I used predicted survival values to calculate differences in seedling survival between soil treatments and light levels. All analyses were performed with R version 3.5.1 (R Core Team, 2020). For frequentist survival analysis, I used the “coxph” function in the survival package (Therneau & Grambsch, 2000) to fit Cox proportional hazards regressions models. For Bayesian survival analysis, I used 103 the “rjags” package to fit models and run predicted survival and contrast simulations (Plummer et al., 2023). I used the lme4 package (D. Bates et al., 2015) to evaluate linear models. I tested the significance of main effects using a likelihood ratio test with the “Anova” function. I tested for multicollinearity variance inflation factors using the “vif” function in the car package (Fox & Weisberg, 2019). I used the “emmeans” function in the multcomp package to evaluate post-hoc Tukey pairwise comparisons of significant main effects, estimated marginal means, and odds- ratios (Hothorn et al., 2008; Lenth, 2020). RESULTS Field seedlings PSFbiomass was negative for Q. alba and positive for Q. rubra, regardless of soil source and light level (H1) In the field, PSFbiomass was negative for Q. alba with lower biomass in conspecific relative to all heterospecific cultured soils at all light levels (Figure 5.1, Figure A5.1, and Figure A5.2). Conversely, Q. rubra experienced positive PSFbiomass with greater biomass in conspecific relative to all heterospecific soils at all light levels. Biomass values for each species ´ soil source ´ light level in the field are presented in Table A5.3. 104 Figure 5.1 Log response ratio ± standard error seedling biomass in conspecific versus heterospecific soils at low light availability, in the field. Values < 1 indicate negative PSF and values > 0 indicate positive PSF. Values that are statistically different from 0 (p < 0.05) are indicated with a star *. Figures for medium and low light are available in the supporting information (Figures A5.3 and A5.4). As an example, Q. alba had lower biomass when growing in soils cultured by Q. alba adults than in soils cultured by adults of any other species (i.e., negative PSFbiomass). Greenhouse seedlings PSFbiomass varied with seedling mycorrhizal type and light availability (H1 & H2) AM species experienced negative PSFbiomass in low light availability. As light increased, PSFbiomass became neutral, or even positive. EM species experienced positive PSFbiomass in low and high light, and some neutral PSFbiomass in medium light. For AM seedlings (A. saccharum and P. serotina), PSFbiomass varied with light availability (Figure 5.2A). At low light, seedlings experienced negative PSFbiomass, with reduced biomass in conspecific relative to all heterospecific cultured soils. A. saccharum also experienced negative PSFbiomass, but only in comparison to P. grandidentata soils. At medium light availability, A. rubrum PSFbiomass was neutral in all heterospecific soils, except for a negative PSFbiomass in A. saccharum soils. Additionally, A. saccharum seedlings experienced positive PSFbiomass in comparison to soils cultured by EM adults (higher survival in conspecific soils compared to soils 105 cultured by Q. alba and Q. rubra). P. serotina experienced positive PSFbiomass in all heterospecific soils. At high light, A. rubrum and P. serotina experienced neutral PSFbiomass when compared to soils cultured by heterospecific AM adults and positive PSFbiomass in soils cultured by heterospecific EM adults. A. saccharum experienced positive PSFbiomass in all heterospecific soils. EM seedlings (Q. alba and Q. rubra) generally experienced positive PSFbiomass, with higher biomass in conspecific relative to all heterospecific soils at low and high light (Figure 5.2B). At medium light, Q. alba experienced neutral PSFbiomass compared to A. rubrum, Q. alba, and Q. rubra soils. Additionally, Q. rubra experienced negative PSFbiomass compared to Q. alba soil. Furthermore, variation in the strength of PSFbiomass varied much more in low light availability than at either medium or high light. Biomass values for each species ´ soil source ´ light level in the greenhouse are presented in Table A5.4. Effects of conspecific microbes on seedling biomass varied with light availability (H3) Soil-borne microbes had a positive effect on Q. alba biomass in the field at low light availability (Figure 5.3A). At both medium and high light availability, microbes had a negative effect on Q. alba biomass and a positive effect on Q. rubra biomass, but sample sizes were too small to determine if these effects were statistically significant (Table A5.2). In the greenhouse, soil-borne microbes had a negative effect on P. serotina biomass and a positive effect on Q. rubra biomass at low light availability (Figure 5.3B). However, effects of microbes shifted to neutral at high light availability for P. serotina. Also, at low and high, but not medium, light availability, soil-borne microbes had a positive effect on biomass Q. rubra. 106 Figure 5.2 Log response ratio ± standard error of seedling biomass in conspecific versus heterospecific soils for A) AM species and B) EM species, in the greenhouse. Values > 0 indicate positive PSF. Values that are statistically different from 0 (p < 0.05) are indicated with a star *. As an example, at low light availability, A. rubrum seedlings had lower biomass in soils cultured by A. rubrum adults than in soils cultured by adults of any other species (i.e., negative PSFbiomass). At high light availability, A. rubrum experienced higher biomass in soils cultured by A. rubrum adults than in soils cultured by P. grandidentata, Q. alba, or Q. rubra adults (i.e., positive PSFbiomass). 107 Figure 5.3 Log response ratio ± standard error of seedling biomass in live conspecific versus sterilized conspecific soils (i.e., biotic effects) at all three light levels, in the A) field and B) greenhouse. Values > 0 indicate positive PSF. Values that are statistically different from 0 (p < 0.05) are indicated with a star *. As an example, in the field, Q. alba seedlings had higher biomass in live than sterilized conspecific soils (i.e., positive PSFbiomass) and there was no difference in biomass for Q. rubra seedlings growing in live versus sterilized soils (i.e., neutral PSFbiomass). 108 There was a negative relationship between low light PSF-survival versus high light PSF- growth (H4) There was a negative relationship between PSFsurvival in low light availability and PSFbiomass in high light availability (Figure 5.4; R2 = 0.21, F6,1 = 23, p = 0.02). There was no significant relationship between PSFsurvival in low light availability and PSFbiomass in high light availability in the field experiment (p > 0.05; Figure A5.3). Figure 5.4 Relationship between PSFsurvival at low light availability and PSFbiomass at high light availability in the greenhouse experiment. Each point is a mean PSFsurvival and PSFbiomass value for a given species heterospecific soil treatment combination for the given light level, where PSF is calculated at the relative performance in conspecific versus heterospecific soil. A positive PSF indicates higher performance in conspecific than heterospecific soils, and vice-versa. The dashed vertical line at x = 0 is the break between negative and positive PSFsurvival. Here, all AM seedlings appear to the left and all EM species appear to the right of the dashed low-light PSFsurvival line. Differences appear to be driven by both the mycorrhizal type of the seedlings and the adults culturing the soil. 109 DISCUSSION In this study, I found that PSFbiomass is mediated by both mycorrhizal type and light availability. Whereas AM species tend to have negative PSFbiomass that becomes more positive as light increases, EM species experience positive PSFbiomass regardless of light level. Also, I found a negative relationship between PSFbiomass at high light availability and PSFsurvival at low light availability, with differences in seedling mycorrhizal type driving this relationship. PSFbiomass responses to light availability differ between AM and EM species Environmental gradients, like light availability, can modify plant-soil interactions (Beals et al., 2020). Despite this, there are few studies investigating how light availability can influence PSFs and results are variable among those that have (McCarthy-Neumann and Ibáñez 2013; Smith and Reynolds 2015; Chapter 3; Chapter 4). A possible explanation for variation in PSFs across light availability is mycorrhizal type. In this study, I found that AM seedlings experience negative PSFs in low light, but PSFs shifted to positive as light availability increased. Previous research has demonstrated that AM species typically experience negative or neutral PSFs and EM species experience positive or neutral PSFs (Bennett et al., 2017; Kadowaki et al., 2018). AM species may be more susceptible to soil- borne pathogens that are more prevalent in low light conditions (Hersh et al., 2012; Y. Liu & He, 2019; Reinhart et al., 2010). In addition, AMF may act parasitically in low light (Ibáñez & McCarthy-Neumann, 2016; Konvalinková & Jansa, 2016), potentially exacerbating the negative effects of pathogens on seedling performance. In a previous study, I found that AMF can have negative effects on AM seedling survival, even when pathogens are not present, suggesting that AMF can have detrimental effects on AM species performance similar to pathogens (Chapter 2). 110 In contrast, EM seedlings in this study, when grown in the greenhouse always experienced positive or neutral PSFs, regardless of light availability (Figure 5.1). EMF may confer greater protection from pathogens by forming a physical sheath around seedling root tips (Laliberté et al., 2015). Furthermore, EMF do not appear to act parasitically at low light availability, even when resources are limited (but see Ågren et al., 2019). Additionally, the matching or mismatching of mycorrhizal type between the juvenile and adult tree culturing the soil may also influence PSFs. For example, PSFs may shift in both strength and direction when AM seedlings are grown in EM, rather than AM soils, and vice- versa (Kadowaki et al. 2018). In a previous study (see Chapter 4), I found that negative PSFsurvival for AM species almost always occurred when heterospecific soils were cultured by EM adults; positive PSFsurvival for EM seedlings occurred when heterospecific soils were cultured by AM adults. However, PSFbiomass in this study did not differ with mycorrhizal type, suggesting that mycorrhizal type mismatching between the juvenile and adult culturing the soil may be more important for survival than for biomass. Effects of soil-borne microbes in conspecific soils were limited There did not appear to be strong effects of soil-borne microbes in conspecific soils across species. In the greenhouse, soil-borne microbes had negative effects on biomass for P. serotina only at low light and positive effects on biomass for Q. rubra at low and high light (Figure 3). Similarly, in the field experiment, soil-borne microbes had a positive effect on Q. rubra at all light levels. AM species, like P. serotina, typically experience more negative PSFs due to higher abundance of specialized soil-borne pathogens in conspecific soils (Chen et al., 2019). Additionally, EM species, like Q. rubra, are better-defended against pathogens (Bennett et al., 2017) and likely derive more benefits from this symbiosis. 111 In the field experiment, soil-borne microbes in conspecific soils had a positive effect on Q. alba biomass in low light but a negative effect on biomass in medium and high light (Figure 5.3). This may explain the negative PSFbiomass (lower biomass in conspecific than heterospecific soils) experienced by Q. alba seedlings in the field (Figure 5.1). At low light, negative PSFbiomass is likely driven by positive effects of microbes in heterospecific soils that overwhelm the positive soil microbe effects in conspecific soils. It is important to note that I was unable to determine if PSFs in either study were driven by soil-borne microbes in the conspecific or heterospecific soils sources, since I only compared live versus sterilized conspecific soils. It is important to distinguish which soil source drives biotic effects of PSFs. Positive PSFs may be driven by greater abundance of mutualists in conspecific soils or lower abundance of pathogenic microbes in heterospecific soils. Alternatively, negative PSFs may be driven by a higher relative abundance of pathogens in conspecific soils. Mycorrhizal type and light availability drive differences in PSFbiomass and PSFsurvival Whether biomass or survival is measured may also contribute to differences in PSFs across mycorrhizal type and light levels. Studies that measure survival (PSFsurvival; McCarthy- Neumann and Ibáñez 2013; Chapter 4) seem to find stronger feedbacks (negative and positive) in low light availability (but see Chapter 3). Other studies measuring biomass found more negative feedbacks at high light availability that became neutral or positive as light availability decreased (L. M. Smith & Reynolds, 2015). Conversely, in this study, I found that PSFbiomass generally became more positive as light increased, aligning with findings by Xi et al. (2023). In the greenhouse experiment, there was a negative relationship between high-light PSFbiomass and low-light PSFsurvival. PSFsurvival may be more responsive at low light, especially for 112 species that are more susceptible to mortality from soil-borne pathogens, whereas PSFbiomass may be more important at high light, especially for seedlings that do not experience strong differences in survival when grown in different soil sources. The negative relationship between PSFbiomass at high light and PSFsurvival in low light closely aligns with a trade-off between high light growth and low light survival (Kobe et al., 1995). Shade intolerant species, which allocate more resources to growth and fewer to defense, should have better growth in high light and poor survival in low light. Shade intolerant species are also more susceptible to negative effects of soil-borne microbes in low light availability (Chapter 2). Differences in PSFsurvival at low light and PSFbiomass at high light appear to be driven by seedling mycorrhizal type. In this study, PSFsurvival appears to be important for AM species or species that allocate more resources to fast growth. In agreement with these results, research in grasslands has demonstrated that fast-growing, early successional species tend to experience more negative PSFs (Cortois et al., 2016; Xi et al., 2021). In a previous study, I also found that shade tolerant species, which often have slower-growth strategies and higher investment in defense traits, have lower mortality due to soil-borne microbes (Chapter 2). In contrast, PSFbiomass may be more important for EM species or species that allocate more resources to defense/survival traits, rather than growth. I recommend future studies include both AM and EM species with differences in resource acquisition strategies to disentangle a potential growth- defense tradeoff in PSFs. In addition to mycorrhizal type of the seedling, mycorrhizal type of the adult culturing the soil could influence the low-light PSFsurvival and high-light PSFbiomass relationship, especially for AM species. AM seedling experienced a lot of variation in high-light PSFbiomass, with PSFbiomass being lower in soils cultured by AM than EM adults. These differences appear to drive 113 the differences between low-light PSFsurvival and high-light PSFbiomass. The large amount of variation in AM seedling responses to soil sources could drive differences in seedling recruitment patterns. Differences in PSFbiomass at high light and PSFsurvival at low light may also be explained by seed size. In this study, I was unable to tease apart effects of mycorrhizal type and seed size. The AM species in this study were all small-seeded and were thus likely influenced more strongly by the experimental treatments. In contrast, the EM species in this study were all large-seeded and had a longer period to rely upon seed reserves and cotyledon support, thus leaving less time to respond to the treatments. Previous research has demonstrated that large-seeded species typically experience more positive PSFs (Moles, 2005). Seed-size advantage might be especially important in the first growing season, when seedlings are still reliant upon maternal seed reserves. Additional Caveats In this study, I was unable to directly compare the results of the field and greenhouse studies. All A. saccharum and P. serotina seedlings died by the end of the first growing season in the field (Chapter 3), and I did not harvest any Quercus seedlings at the end of the first growing season. However, I am still able to infer differences between field and greenhouse trends, especially between the Quercus seedlings that survived through both experiments. Furthermore, field experiments, regardless of whether they measure PSFsurvival (Chapter 3) or PSFbiomass, seem to result in weaker PSF responses, compared to the greenhouse. Greenhouse experiments tend to overestimate PSFs, relative to field experiments where more stochastic environment conditions may overwhelm PSF effects (Forero et al., 2019; Kulmatiski & Kardol, 2008). I suggest that future research pair both greenhouse and field experiments to better understand the underlying 114 mechanisms driving PSFs and also a realistic measure of the strength and direction of PSFs in the field. Other abiotic factors, like nutrient availability, can also influence the strength and direction of PSFs. Nutrient availability might alter PSFs by increasing resources available for seedling, thus increasing their overall health and indirectly increasing their resilience against harmful microbes. I found no effect of sterilization on soil nutrient availability, nor did I find differences in soil availability in soils beneath adult trees versus soils collected for the plastic liner pots. I also did not find substantial differences in nutrient availability between adult tree species. Therefore, I do not think that nutrient availability played a strong role in mediating PSFs for either of these experiments. Implications for forest community dynamics Understanding the mechanisms underlying PSFs is key for understanding forest regeneration dynamics. Negative PSFs (lower performance in conspecific versus heterospecific soils) may have positive effects on forest community diversity and stabilize species coexistence by increasing the likelihood that a seedling of a different species will replace an adult tree when it dies (Bever et al., 1997; Chesson, 2000). Conversely, positive PSFs (greater performance in conspecific than heterospecific soils) may decrease community diversity by increasing the likelihood that an adult tree is replaced by a seedling of the same species. Differences in seedling PSFs can drive forest successional dynamics and management (Jiang et al., 2020; Q. Liu & Zhao, 2023). EM species appear to derive benefit from soil-borne microbes cultured in soils by conspecific adults, regardless of light level. In contrast, AM species have reduced performance both near conspecific adults and in low light conditions, likely due to detrimental effects of pathogens or parasitic effects of mycorrhizal fungi. If seedlings are limited 115 by the negative effects of soil-borne microbes cultured beneath the crowns of conspecific adults (i.e., AM species), they may have more recruitment success away from conspecific adults in low light availability. Conversely, if seedlings benefit more from positive effects of soil-borne microbes cultured by conspecific adults (i.e., EM species), they may have limited recruitment success in soils farther away from conspecific adults. Together, these results help develop a deeper understanding of seedling regeneration dynamics in the context of mycorrhizal type and light availability. 116 CHAPTER 6 Conclusion Disentangling the mechanisms underlying plant-soil feedbacks (PSFs) is critical to understanding forest community dynamics. In this dissertation, I examined how seedling mycorrhizal type and defense/recovery traits (e.g., phenolics, lignin and NSC) can interact with light availability to mediate PSFs. In Chapter 2, I separated the effects arbuscular mycorrhizal fungi (AMF) and pathogens on these traits and survival for three species in the genus Acer. In Chapter 3, I investigated variation in seedling defense and recovery traits and their effects on PSFsurvival in the field. In Chapter 4, I used a greenhouse experiment, paralleling the field experiment in Chapter 3, to study how mismatching of mycorrhizal type between juvenile and adult trees can alter these interactions. Finally, in Chapter 5, I evaluated PSFbiomass in both the field (Chapter 3) and greenhouse (Chapter 4) experiments. I also examined the relationship between PSFbiomass in high light and PSFsurvival in low light availability. Seedling shade tolerance is mediated by soil-borne microbes and defense/recovery traits In Chapter 1, I demonstrated that survival of newly germinated tree seedlings in low versus high light, or “shade tolerance,” may be due to interactions between low light and soil- borne microbes and be mediated by defense and recovery traits. Previous studies have demonstrated that seedling defense and recovery traits are influenced by light and microbes and that these traits can influence growth and survival at low light (Falster et al., 2018), but have not linked resources, traits, and survival, as in this study. This study provides a needed first step in developing a mechanistic understanding of how soil-borne microbes impact seedling shade tolerance, explained through defense and recovery 117 traits. Although fast-growing shade intolerant species may be expected to outcompete shade tolerant species in high light (Pacala et al., 1996), shade intolerant species can be limited by the negative interactive effects of soil-borne microbes at low light (Y. Liu & He, 2019; McCarthy- Neumann & Kobe, 2010a), restricting their recruitment niche to areas with higher light and fewer soil-borne microbes. For example, A. negundo seedlings may be limited to forest edges and open fields, due to high mortality in the presence of pathogens and parasitic effects of mycorrhizae in forest understories. Conversely, A. saccharum is able to recruit in deeply shaded environments, likely due to higher carbon allocation to defense and recovery traits, and thus lower susceptibility to negative effects of soil-borne microbes. In this chapter, I demonstrated the importance of interactions between soil-borne microbes and light availability in determining tree seedling survival. Furthermore, I related both intra- and interspecific differences in survival to defense and recovery traits, supporting a more trait-based and mechanistic approach to understanding forest community dynamics. Mismatching of mycorrhizal type and interactions with light availability to mediate PSFs In Chapter 4, I found that mismatching of mycorrhizal type between seedlings and the adults culturing the soil plays a strong role in the direction of PSFs. AM seedlings generally experienced neutral PSFs when grown in heterospecific soils cultured by AM adults. However, when grown in heterospecific soils cultured by EM adults, AM seedlings experienced negative PSFs. Similarly, EM seedlings generally experienced neutral PSFs when grown in soils cultured by EM adults, but positive PSFs when grown in soils cultured by AM adults. Mismatching of mycorrhizal type can drive recruitment patterns in Michigan forests. For example, Acer species, which are associated with AMF, often replace Quercus species, which primarily associate with EMF. Acer species typically experience negative PSFs when grown in soils cultured by AM 118 adults. However, when they grow in soils cultured by adult Quercus, they are able to escape effectively-specialized pathogens and potentially parasitic effects of AMF. In addition, PSFs can enhance performance of AM seedlings in higher light conditions, such as canopy gaps, but favor EM species under closed canopies. Whereas AM seedlings typically experience negative PSFs when grown under low light conditions, increasing light availability also increases survival, thereby decreasing differences between conspecific and heterospecific soils. For EM species, however, survival responses are higher under low light availability, allowing them to persist in shaded conditions, even when soil-borne pathogens are more prevalent. AM species may be able to escape soil-borne pathogens and parasitic effects of AMF when recruiting into areas with high light availability or away from conspecific adults. This lends further support to field observations that Acer species often replace Quercus in older forest stands. Seedling traits vary in response to soil source, light availability, and mycorrhizal type Through this dissertation, and especially in Chapter 3, I provide some of the first evidence that defense and recovery traits can vary in seedlings as young as three weeks old. Furthermore, measured amounts of these traits vary in response to both abiotic (light availability) and biotic (soil-borne microbes) factors. Despite the important role of plant survival in PSF (Comita et al., 2010; McCarthy-Neumann & Kobe, 2010a), traits promoting faster growth (e.g., specific leaf area, specific root length, height) have been the focus of most PSF studies (Baxendale et al., 2014; Cortois et al., 2016; Xi et al., 2021). Frequently, defensive traits are accounted for by assuming that species with fast growth rates have low investment in defense, and vice-versa (Cortois et al., 2016; Xi et al., 2021). However, tree seedling survivorship is 119 likely to have greater effects on future community dynamics and composition than growth (Pacala et al., 1996). While little studied, defense and recovery traits that influence tree seedling survivorship in response to PSFs could be a crucial mechanism governing seedling and forest community dynamics. Seedlings growing in conditions where they produce higher amounts of defense and recovery traits (i.e., high light availability) are better able to recruit into those spaces, even when soil-borne microbes are present. For example, Acer seedlings growing in soils cultured by Acer adults would be expected to experience negative PSFs (i.e., lower survival in that conspecific soil than other heterospecific soils, especially those cultured by EM species). However, under high light availability, Acer seedlings may produce enough defensive and recovery traits to overcome the limitations of soil-borne microbes. Defense and recovery traits may effectively neutralize PSFs Results from Chapter 4 indicate that higher amounts of seedling defense and recovery traits may effectively neutralize these PSFs. Differences in seedling responses to soils cultured by AM and EM adults, and changes in PSFs with light availability, can be explained by defense and recovery traits. For AM seedlings, increasing percent colonization by AMF, phenolics, lignin, and NSC – which were influenced by light level and soil source – were associated with less negative (i.e., more neutral) PSFs. For EM seedlings, increasing colonization by AMF and higher amounts of lignin – which were influenced by light level and soil source – were associated with less positive PSFs. This study demonstrates that it is important to consider not only the mycorrhizal type of the seedlings, but also the mycorrhizal type of the trees that cultured the soil in which the seedlings occur (i.e., matching or mismatching mycorrhizal type). 120 There is a negative relationship between PSFbiomass at high light and PSFsurvival at low light While previous research has demonstrated that strength and direction of PSFs varies with mycorrhizal type (AM vs. EM), we have also shown that variation within mycorrhization types may be attributed to differences in light availability, especially for AM seedlings. Environmental gradients, like light availability, can modify plant-soil interactions (Beals et al., 2020). Despite this, there are few studies investigating how light availability can influence PSFs and results are variable among those that have (McCarthy-Neumann and Ibáñez 2013; Smith and Reynolds 2015; Wood et al. 2023b; Chapter 4). Moreover, tree seedlings may express a negative relationship between PSFbiomass at high light availability and PSFsurvival at low light availability, with relative importance of biomass and survival also mediated by mycorrhizal type. Together, these results provide a more mechanistic understanding of the factors underlying PSFs. The negative relationship between PSFbiomass at high light and PSFsurvival in low light closely aligns with a trade-off between high light growth and low light survival (Kobe et al., 1995). Shade intolerant species, which allocate more resources to growth and fewer to defense, should have better growth in high light and poor survival in low light. Shade intolerant species are also more susceptible to negative effects of soil-borne microbes on survival in low light availability (Chapter 2). Implications for forest community dynamics and management Although PSFs are often studied in the context of individual seedlings, the impacts of these PSFs can have broader impacts on forest community dynamics. A focus on tree seedling traits under different environmental conditions, especially in natural field conditions, offers both broader ecological understanding as well as potential applications for forest management. For example, selecting sites with soil and light conditions that promote higher production of defense 121 and recovery compounds could increase likelihood of seedling restoration success (e.g., A. saccharum in conspecific soil). Similarly, it may be beneficial to plant EMF-associating seedlings in soils cultured by other EMF-associating species, to increase potential for positive EMF colonization effects and limit potential negative AMF colonization effects. While environmental conditions could dilute trait effects on seedling survival, in the absence of extreme conditions (as supported by related greenhouse studies), a sharper focus on traits promoting survival, especially under low light conditions, can provide a more mechanistic understanding of forest regeneration dynamics. Linking mycorrhizal type, defense and recovery traits, and environmental conditions like light availability to seedling survival and biomass may help us better understand the role of PSFs in forest communities (Baxendale et al., 2014; Bennett & Klironomos, 2018; P. Ke et al., 2015). This dissertation provides some first steps in disentangling these factors and helps develop a deeper understanding of seedling regeneration dynamics. 122 BIBLIOGRAPHY Abrams, M. (1996). Distribution, historical development and ecophysiological attributes of oak species in the eastern United States. Annales Des Sciences Forestières, 53(2–3), 487– 512. https://doi.org/10.1051/forest:19960230 Ainsworth, E. A., & Gillespie, K. M. (2007). Estimation of total phenolic content and other oxidation substrates in plant tissues using Folin–Ciocalteu reagent. Nature Protocols, 2(4), 875–877. https://doi.org/10.1038/nprot.2007.102 Alexander, H. D., Siegert, C., Brewer, J. S., Kreye, J., Lashley, M. A., McDaniel, J. K., Paulson, A. K., Renninger, H. J., & Varner, J. M. (2021). Mesophication of Oak Landscapes: Evidence, Knowledge Gaps, and Future Research. BioScience, 71(5), 531–542. https://doi.org/10.1093/biosci/biaa169 Allen, D., Dick, C. W., Strayer, E., Perfecto, I., & Vandermeer, J. (2018). Scale and strength oak-mesophyte interactions in a transitional oak-hickory forest. Canadian Journal of Forest Research. Journal Canadien de La Recherche Forestiere, 48(11), 1366–1372. https://doi.org/10.1139/cjfr-2018-0131 Allison, S. D., Lu, Y., Weihe, C., Goulden, M. L., Martiny, A. C., Treseder, K. K., & Martiny, J. B. H. (2013). Microbial abundance and composition influence litter decomposition response to environmental change. Ecology, 94(3), 714–725. https://doi.org/10.1890/12- 1243.1 Alvarez-Clare, S., & Kitajima, K. (2007). Physical defense traits enhance seedling survival of neotropical tree species. Functional Ecology, 21(6), 1044–1054. https://doi.org/10.1111/j.1365-2435.2007.01320.x Al-Yahya’ei, M. N., Oehl, F., Vallino, M., Lumini, E., Redecker, D., Wiemken, A., & Bonfante, P. (2011). Unique arbuscular mycorrhizal fungal communities uncovered in date palm plantations and surrounding desert habitats of Southern Arabia. Mycorrhiza, 21(3), 195– 209. https://doi.org/10.1007/s00572-010-0323-5 Augspurger, C. K. (1984a). Light requirements of Neotropical tree seedlings: A comparative study of growth and survival. The Journal of Ecology, 72(3), 777. https://doi.org/10.2307/2259531 Augspurger, C. K. (1984b). Seedling survival of tropical tree species: Interactions of dispersal distance, light-gaps, and pathogens. Ecology, 65(6), 1705–1712. https://doi.org/10.2307/1937766 Augspurger, C. K. (1990). Spatial patterns of damping-off disease during seedling recruitment in tropical forests. In J. J. Burdon, S. R. Leather, & British Society for Plant Pathology (Eds.), Pests, Pathogens, and Plant Communities (pp. 131–144). Blackwell Scientific. 123 Augspurger, C. K., & Kelly, C. K. (1984). Pathogen mortality of tropical tree seedlings: Experimental studies of the effects of dispersal distance, seedling density, and light conditions. Oecologia, 61(2), 211–217. https://doi.org/10.1007/BF00396763 Azcón-Aguilar, C., Jaizme-Vega, M. C., & Calvet, C. (2002). The contribution of arbuscular mycorrhizal fungi to the control of soil-borne plant pathogens. In S. Gianinazzi, H. Schüepp, J. M. Barea, & K. Haselwandter (Eds.), Mycorrhizal Technology in Agriculture (pp. 187–197). Birkhäuser Basel. https://doi.org/10.1007/978-3-0348-8117-3_15 Ballaré, C. L. (2014). Light regulation of plant defense. Annual Review of Plant Biology, 65(1), 335–363. https://doi.org/10.1146/annurev-arplant-050213-040145 Ballaré, C. L., & Austin, A. T. (2019). Recalculating growth and defense strategies under competition: Key roles of photoreceptors and jasmonates. Journal of Experimental Botany, 70(13), 3425–3434. https://doi.org/10.1093/jxb/erz237 Bates, D., Machler, M., Bolker, B., & Walker, S. (2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67(1), 1–48. Bates, S. E., Wandrag, E. M., & Duncan, R. P. (2020). Calculating the uncertainty associated with log response ratios in plant–soil feedback studies. Plant Ecology, 221(9), 829–836. https://doi.org/10.1007/s11258-019-00981-6 Baxendale, C., Orwin, K. H., Poly, F., Pommier, T., & Bardgett, R. D. (2014). Are plant–soil feedback responses explained by plant traits? New Phytologist, 204(2), 408–423. https://doi.org/10.1111/nph.12915 Beals, K. K., Moore, J. A. M., Kivlin, S. N., Bayliss, S. L. J., Lumibao, C. Y., Moorhead, L. C., Patel, M., Summers, J. L., Ware, I. M., Bailey, J. K., & Schweitzer, J. A. (2020). Predicting plant-soil feedback in the field: Meta-analysis reveals that competition and environmental stress differentially influence PSF. Frontiers in Ecology and Evolution, 8. https://www.frontiersin.org/article/10.3389/fevo.2020.00191 Benítez, M.-S., Hersh, M. H., Vilgalys, R., & Clark, J. S. (2013). Pathogen regulation of plant diversity via effective specialization. Trends in Ecology & Evolution, 28(12), 705–711. https://doi.org/10.1016/j.tree.2013.09.005 Bennett, J. A., & Klironomos, J. (2018). Climate, but not trait, effects on plant-soil feedback depend on mycorrhizal type in temperate forests. Ecosphere, 9(3), e02132. https://doi.org/10.1002/ecs2.2132 Bennett, J. A., Maherali, H., Reinhart, K. O., Lekberg, Y., Hart, M. M., & Klironomos, J. (2017). Plant-soil feedbacks and mycorrhizal type influence temperate forest population dynamics. Science, 355(6321), 181–184. https://doi.org/10.1126/science.aai8212 Bereau, M., Barigah, T. S., Louisanna, E., & Garbaye, J. (2000). Effects of endomycorrhizal development and light regimes on the growth of Dicorynia guianensis Amshoff 124 seedlings. Annals of Forest Science, 57(7), 725–733. https://doi.org/10.1051/forest:2000153 Bever, J. D., Dickie, I. A., Facelli, E., Facelli, J. M., Klironomos, J., Moora, M., Rillig, M. C., Stock, W. D., Tibbett, M., & Zobel, M. (2010). Rooting theories of plant community ecology in microbial interactions. Trends in Ecology & Evolution, 25(8), 468–478. https://doi.org/10.1016/j.tree.2010.05.004 Bever, J. D., Westover, K. M., & Antonovics, J. (1997). Incorporating the soil community into plant population dynamics: The utility of the feedback approach. The Journal of Ecology, 85(5), 561. https://doi.org/10.2307/2960528 Bingham, M. A., & Simard, S. (2012). Ectomycorrhizal networks of Pseudotsuga menziesii var. Glauca trees facilitate establishment of conspecific seedlings under drought. Ecosystems, 15(2), 188–199. https://doi.org/10.1007/s10021-011-9502-2 Borowicz, V. A. (2001). Do arbuscular mycorrhizal fungi alter plant-pathogen relations? Ecology, 82(11), 3057–3068. https://doi.org/10.2307/2679834 Bücking, H., Mensah, J. A., & Fellbaum, C. R. (2016). Common mycorrhizal networks and their effect on the bargaining power of the fungal partner in the arbuscular mycorrhizal symbiosis. Communicative & Integrative Biology, 9(1), e1107684. https://doi.org/10.1080/19420889.2015.1107684 Burnham, K. P., & Anderson, D. R. (2002). Model selection and multimodel inference (2nd ed.). Springer-Verlag. Burns, R. M., & Honkala, B. H. (1990a). Silvics of North America: 2. Hardwoods (Vol. 2). U.S. Department of Agriculture, Forest Service. Burns, R. M., & Honkala, B. H. (1990b). Silvics of North America (Volume 2: Hardwoods). In Agriculture Handbook 654 (Vol. 2). Forest Service, United States Department of Agriculture. Callaway, R. M., Bedmar, E. J., Reinhart, K. O., Silvan, C. G., & Klironomos, J. (2011). Effects of soil biota from different ranges on “Robinia” invasion: Acquiring mutualists and escaping pathogens. Ecology, 92(5), 1027–1035. Catford, J. A., Wilson, J. R. U., Pyšek, P., Hulme, P. E., & Duncan, R. P. (2022). Addressing context dependence in ecology. Trends in Ecology & Evolution, 37(2), 158–170. https://doi.org/10.1016/j.tree.2021.09.007 Chen, L., Swenson, N. G., Ji, N., Mi, X., Ren, H., Guo, L., & Ma, K. (2019). Differential soil fungus accumulation and density dependence of trees in a subtropical forest. Science, 366(6461), 124–128. https://doi.org/10.1126/science.aau1361 Chesson, P. (2000). Mechanisms of maintenance of species diversity. Annual Review of Ecology and Systematics, 31, 343–366. 125 Chilvers, G. A., Lapeyrie, F. F., & Horan, D. P. (1987). Ectomycorrhizal vs endomycorrhizal fungi within the same root system. The New Phytologist, 107(2), 441–448. https://doi.org/10.1111/j.1469-8137.1987.tb00195.x Coley, P. D., & Barone, J. A. (1996). Herbivory and plant defenses in tropical forests. Annual Review of Ecology and Systematics, 27(1), 305–335. https://doi.org/10.1146/annurev.ecolsys.27.1.305 Coley, P. D., Bryant, J. P., & Chapin, F. S. (1985). Resource availability and plant antiherbivore defense. Science, 230(4728), 895–899. https://doi.org/10.1126/science.230.4728.895 Comita, L. S., Muller-Landau, H. C., Aguilar, S., & Hubbell, S. P. (2010). Asymmetric density dependence shapes species abundances in a tropical tree community. Science, 329(5989), 330–332. https://doi.org/10.1126/science.1190772 Cortois, R., Schröder‐Georgi, T., Weigelt, A., Putten, W. H., & De Deyn, G. B. (2016). Plant– soil feedbacks: Role of plant functional group and plant traits. Journal of Ecology, 104(6), 1608–1617. https://doi.org/10.1111/1365-2745.12643 Cox, D. R., & Oakes, D. (2017). Analysis of Survival Data. Chapman and Hall/CRC. https://doi.org/10.1201/9781315137438 Crawford, K. M., Bauer, J. T., Comita, L. S., Eppinga, M. B., Johnson, D. J., Mangan, S. A., Queenborough, S. A., Strand, A. E., Suding, K. N., Umbanhowar, J., & Bever, J. D. (2019). When and where plant‐soil feedback may promote plant coexistence: A meta‐ analysis. Ecology Letters, 22(8), 1274–1284. https://doi.org/10.1111/ele.13278 Dietze, M. C., Sala, A., Carbone, M. S., Czimczik, C. I., Mantooth, J. A., Richardson, A. D., & Vargas, R. (2014). Nonstructural carbon in woody plants. Annual Review of Plant Biology, 65(1), 667–687. https://doi.org/10.1146/annurev-arplant-050213-040054 Dillaway, D. N., Stringer, J. W., & Rieske, L. K. (2007). Light availability influences root carbohydrates, and potentially vigor, in white oak advance regeneration. Forest Ecology and Management, 250(3), 227–233. https://doi.org/10.1016/j.foreco.2007.05.019 Dudenhöffer, J.-H., Ebeling, A., Klein, A.-M., & Wagg, C. (2018). Beyond biomass: Soil feedbacks are transient over plant life stages and alter fitness. Journal of Ecology, 106(1), 230–241. https://doi.org/10.1111/1365-2745.12870 Duponnois, R., Diédhiou, S., Chotte, J. L., & Sy, M. O. (2003). Relative importance of the endomycorrhizal and (or) ectomycorrhizal associations in Allocasuarina and Casuarina genera. Canadian Journal of Microbiology, 49(4), 281–287. https://doi.org/10.1139/w03- 038 Eagar, A. C., Mushinski, R. M., Horning, A. L., Smemo, K. A., Phillips, R. P., & Blackwood, C. B. (2022). Arbuscular Mycorrhizal Tree Communities Have Greater Soil Fungal Diversity and Relative Abundances of Saprotrophs and Pathogens than Ectomycorrhizal 126 Tree Communities. Applied and Environmental Microbiology, 88(1), e01782-21. https://doi.org/10.1128/AEM.01782-21 Eagar, A. C., Smemo, K. A., Phillips, R. P., & Blackwood, C. B. (2023). Context-dependence of fungal community responses to dominant tree mycorrhizal types in Northern hardwood forests. Soil Biology and Biochemistry, 178, 108971. https://doi.org/10.1016/j.soilbio.2023.108971 Egerton-Warburton, L., & Allen, M. F. (2001). Endo- and ectomycorrhizas in Quercus agrifolia Nee. (Fagaceae): Patterns of root colonization and effects on seedling growth. Mycorrhiza, 11(6), 283–290. https://doi.org/10.1007/s005720100134 Esch, C. M., & Kobe, R. K. (2021). Short-lived legacies of Prunus serotina plant–soil feedbacks. Oecologia, 196(2), 529–538. https://doi.org/10.1007/s00442-021-04948-1 Falcioni, R., Moriwaki, T., de Oliveira, D. M., Andreotti, G. C., de Souza, L. A., dos Santos, W. D., Bonato, C. M., & Antunes, W. C. (2018). Increased gibberellins and light levels promotes cell wall thickness and enhance lignin deposition in xylem fibers. Frontiers in Plant Science, 9. https://www.frontiersin.org/articles/10.3389/fpls.2018.01391 Falster, D. S., Duursma, R. A., & FitzJohn, R. G. (2018). How functional traits influence plant growth and shade tolerance across the life cycle. Proceedings of the National Academy of Sciences, 115(29), E6789–E6798. https://doi.org/10.1073/pnas.1714044115 Forero, L. E., Grenzer, J., Heinze, J., Schittko, C., & Kulmatiski, A. (2019). Greenhouse- and field-measured plant-soil feedbacks are not correlated. Frontiers in Environmental Science, 7, 184. https://doi.org/10.3389/fenvs.2019.00184 Fox, J., & Weisberg, S. (2019). An R companion to applied regression. SAGE. Gao, C., Courty, P., Varoquaux, N., Cole, B., Montoya, L., Xu, L., Purdom, E., Vogel, J., Hutmacher, R. B., Dahlberg, J. A., Coleman‐Derr, D., Lemaux, P. G., & Taylor, J. W. (2023). Successional adaptive strategies revealed by correlating arbuscular mycorrhizal fungal abundance with host plant gene expression. Molecular Ecology, 32(10), 2674– 2687. https://doi.org/10.1111/mec.16343 García de León, D., Moora, M., Öpik, M., Jairus, T., Neuenkamp, L., Vasar, M., Bueno, C. G., Gerz, M., Davison, J., & Zobel, M. (2016). Dispersal of arbuscular mycorrhizal fungi and plants during succession. Acta Oecologica, 77, 128–135. https://doi.org/10.1016/j.actao.2016.10.006 Gehring, C. A. (2003). Growth responses to arbuscular mycorrhizal fungi by rain forest seedlings vary with light intensity and tree species. Plant Ecology, 167(1), 127–139. https://doi.org/10.1023/A:1023989610773 Gleason, S. M., & Ares, A. (2004). Photosynthesis, carbohydrate storage and survival of a native and an introduced tree species in relation to light and defoliation. Tree Physiology, 24(10), 1087–1097. https://doi.org/10.1093/treephys/24.10.1087 127 Gómez-Aparicio, L., Domínguez-Begines, J., Kardol, P., Ávila, J. M., Ibáñez, B., & García, L. V. (2017). Plant-soil feedbacks in declining forests: Implications for species coexistence. Ecology, 98(7), 1908–1921. https://doi.org/10.1002/ecy.1864 Graham, J. H. (2001). What do root pathogens see in mycorrhizas? New Phytologist, 149(3), 357–359. https://doi.org/10.1046/j.1469-8137.2001.00077.x Grman, E., Robinson, T. M. P., & Klausmeier, C. A. (2012). Ecological specialization and trade affect the outcome of negotiations in mutualism. The American Naturalist, 179(5), 567– 581. https://doi.org/10.1086/665006 Gurevitch, J. S., Scheiner, M., & Fox, G. A. (2020). The Ecology of Plants (3rd ed.). Sinauer Associates. Hassan, K., Golam Dastogeer, K. M., Carrillo, Y., & Nielsen, U. N. (2022). Climate change- driven shifts in plant–soil feedbacks: A meta-analysis. Ecological Processes, 11(1), 64. https://doi.org/10.1186/s13717-022-00410-z Heinze, J., Wacker, A., & Kulmatiski, A. (2020). Plant–soil feedback effects altered by aboveground herbivory explain plant species abundance in the landscape. Ecology, 00(00), e03023. https://doi.org/10.1002/ecy.3023 Hersh, M. H., Vilgalys, R., & Clark, J. S. (2012). Evaluating the impacts of multiple generalist fungal pathogens on temperate tree seedling survival. Ecology, 93(3), 511–520. https://doi.org/10.1890/11-0598.1 Hothorn, T., Bretz, F., & Westfall, P. (2008). Simultaneous inference in general parametric models. Biometric Journal, 50(3), 346–363. https://doi.org/10.1002/bimj.200810425 Ibáñez, I., & McCarthy-Neumann, S. (2014). Integrated assessment of the direct and indirect effects of resource gradients on tree species recruitment. Ecology, 95(2), 364–375. https://doi.org/10.1890/13-0685.1 Ibáñez, I., & McCarthy-Neumann, S. (2016). Effects of mycorrhizal fungi on tree seedling growth: Quantifying the parasitism–mutualism transition along a light gradient. Canadian Journal of Forest Research, 46(1), 48–57. https://doi.org/10.1139/cjfr-2015-0327 Ichihara, Y., & Yamaji, K. (2009). Effect of light conditions on the resistance of current-year Fagus crenata seedlings against fungal pathogens causing damping-off in a natural Beech forest: Fungus isolation and histological and chemical resistance. Journal of Chemical Ecology, 35(9), 1077–1085. https://doi.org/10.1007/s10886-009-9687-4 Imaji, A., & Seiwa, K. (2010). Carbon allocation to defense, storage, and growth in seedlings of two temperate broad-leaved tree species. Oecologia, 162(2), 273–281. https://doi.org/10.1007/s00442-009-1453-3 Jevon, F. V., De La Cruz, D., LaManna, J. A., Lang, A. K., Orwig, D. A., Record, S., Kouba, P. V., Ayres, M. P., & Matthes, J. H. (2022). Experimental and observational evidence of 128 negative conspecific density dependence in temperate ectomycorrhizal trees. Ecology, n/a(n/a), e3808. https://doi.org/10.1002/ecy.3808 Jiang, J., Abbott, K. C., Baudena, M., Eppinga, M. B., Umbanhowar, J. A., & Bever, J. D. (2020). Pathogens and mutualists as joint drivers of host species coexistence and turnover: Implications for plant competition and succession. The American Naturalist, 195(4), 591–602. https://doi.org/10.1086/707355 Kadowaki, K., Yamamoto, S., Sato, H., Tanabe, A. S., Hidaka, A., & Toju, H. (2018). Mycorrhizal fungi mediate the direction and strength of plant–soil feedbacks differently between arbuscular mycorrhizal and ectomycorrhizal communities. Communications Biology, 1(1), 1–11. https://doi.org/10.1038/s42003-018-0201-9 Kajihara, K. T., Egan, C. P., Swift, S. O. I., Wall, C. B., Muir, C. D., & Hynson, N. A. (2022). Core arbuscular mycorrhizal fungi are predicted by their high abundance–occupancy relationship while host-specific taxa are rare and geographically structured. New Phytologist, 234(4), 1464–1476. https://doi.org/10.1111/nph.18058 Ke, P. J., Zee, P. C., & Fukami, T. (2021). Dynamic plant–soil microbe interactions: The neglected effect of soil conditioning time. New Phytologist, 231(4), 1546–1558. https://doi.org/10.1111/nph.17420 Ke, P., Miki, T., & Ding, T. (2015). The soil microbial community predicts the importance of plant traits in plant–soil feedback. New Phytologist, 206(1), 329–341. https://doi.org/10.1111/nph.13215 Kitajima, K. (1994). Relative importance of photosynthetic traits and allocation patterns as correlates of seedling shade tolerance of 13 tropical trees. Oecologia, 98(3–4), 419–428. https://doi.org/10.1007/BF00324232 Klironomos, J. N. (2002). Feedback with soil biota contributes to plant rarity and invasiveness in communities. Nature, 417(6884), 67–70. https://doi.org/10.1038/417067a Klironomos, J. N., Bednarczuk, E. M., & Neville, J. (1999). Reproductive significance of feeding on saprobic and arbuscular mycorrhizal fungi by the collembolan, Folsomia candida. Functional Ecology, 13(6), 756–761. https://doi.org/10.1046/j.1365-2435.1999.00379.x Knott, J. A., Desprez, J. M., Oswalt, C. M., & Fei, S. (2019). Shifts in forest composition in the eastern United States. Forest Ecology and Management, 433, 176–183. https://doi.org/10.1016/j.foreco.2018.10.061 Kobe, R. K. (1997). Carbohydrate allocation to storage as a basis of interspecific variation in sapling survivorship and growth. Oikos, 80(2), 226. https://doi.org/10.2307/3546590 Kobe, R. K., Iyer, M., & Walters, M. B. (2010). Optimal partitioning theory revisited: Nonstructural carbohydrates dominate root mass responses to nitrogen. Ecology, 91(1), 166–179. https://doi.org/10.1890/09-0027.1 129 Kobe, R. K., Pacala, S. W., Silander, J. A., & Canham, C. D. (1995). Juvenile tree survivorship as a component of shade tolerance. Ecological Applications, 5(2), 517–532. https://doi.org/10.2307/1942040 König, J., van Kleunen, M., & Dawson, W. (2016). No consistent legacy effects of invasion by giant goldenrod (Solidago gigantea) via soil biota on native plant growth. Journal of Plant Ecology, 9(3), 320–327. https://doi.org/10.1093/jpe/rtv054 Konvalinková, T., & Jansa, J. (2016). Lights off for arbuscular mycorrhiza: On its symbiotic functioning under light deprivation. Frontiers in Plant Science, 7. https://doi.org/10.3389/fpls.2016.00782 Koorem, K., Tulva, I., Davison, J., Jairus, T., Öpik, M., Vasar, M., Zobel, M., & Moora, M. (2017). Arbuscular mycorrhizal fungal communities in forest plant roots are simultaneously shaped by host characteristics and canopy-mediated light availability. Plant and Soil, 410(1–2), 259–271. https://doi.org/10.1007/s11104-016-3004-0 Kruschke, J. (2014). Doing bayesian data analysis (2nd ed.). Elsevier Academic Press. Kulmatiski, A., Beard, K. H., Stevens, J. R., & Cobbold, S. M. (2008). Plant-soil feedbacks: A meta-analytical review. Ecology Letters, 11(9), 980–992. https://doi.org/10.1111/j.1461- 0248.2008.01209.x Kulmatiski, A., & Kardol, P. (2008). Getting plant-soil feedbacks out of the greenhouse: Experimental and conceptual approaches. In U. Lüttge, W. Beyschlag, & J. Murata (Eds.), Progress in Botany (Vol. 69, pp. 449–472). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-540-72954-9_18 Kuťáková, E., Herben, T., & Münzbergová, Z. (2018). Heterospecific plant–soil feedback and its relationship to plant traits, species relatedness, and co-occurrence in natural communities. Oecologia, 187(3), 679–688. https://doi.org/10.1007/s00442-018-4145-z Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. B. (2017). lmerTest package: Tests in linear mixed effects models. Journal of Statistical Software, 82(13), 1–26. https://doi.org/10.18637/jss.v082.i13 Laliberté, E., Lambers, H., Burgess, T. I., & Wright, S. J. (2015). Phosphorus limitation, soil- borne pathogens and the coexistence of plant species in hyperdiverse forests and shrublands. New Phytologist, 206(2), 507–521. https://doi.org/10.1111/nph.13203 Landhäusser, S. M., Chow, P. S., Dickman, L. T., Furze, M. E., Kuhlman, I., Schmid, S., Wiesenbauer, J., Wild, B., Gleixner, G., Hartmann, H., Hoch, G., McDowell, N. G., Richardson, A. D., Richter, A., & Adams, H. D. (2018). Standardized protocols and procedures can precisely and accurately quantify non-structural carbohydrates. Tree Physiology, 38(12), 1764–1778. https://doi.org/10.1093/treephys/tpy118 Landhäusser, S. M., Chow, P. S., Turin Dickman, L., Furze, M. E., Kuhlman, I., Schmid, S., Wiesenbauer, J., Wild, B., Gleixner, G., Hartmann, H., Hoch, G., McDowell, N. G., 130 Richardson, A. D., Richter, A., & Adams, H. D. (2018). Standardized protocols and procedures can precisely and accurately quantify non-structural carbohydrates. Tree Physiology, 38(12), 1764–1778. https://doi.org/10.1093/treephys/tpy118 Lee, M.-H., Jeon, H. S., Kim, S. H., Chung, J. H., Roppolo, D., Lee, H.-J., Cho, H. J., Tobimatsu, Y., Ralph, J., & Park, O. K. (2019). Lignin-based barrier restricts pathogens to the infection site and confers resistance in plants. The EMBO Journal, 38(23), e101948. https://doi.org/10.15252/embj.2019101948 Lenth, R. (2020). emmeans: Estimated marginal means, aka least-squares means (R package version 1.5.1). Li, J., Zhang, X., Luo, J., Lindsey, S., Zhou, F., Xie, H., Li, Y., Zhu, P., Wang, L., Shi, Y., He, H., & Zhang, X. (2020). Differential accumulation of microbial necromass and plant lignin in synthetic versus organic fertilizer-amended soil. Soil Biology and Biochemistry, 149, 107967. https://doi.org/10.1016/j.soilbio.2020.107967 Li, Y.-L., Jin, Z.-X., Luo, G.-Y., Chen, C., Sun, Z.-S., & Wang, X.-Y. (2022). Effects of arbuscular mycorrhizal fungi inoculation on non-structural carbohydrate contents and C:N:P stoichiometry of Heptacodium miconioides under drought stress. Ying Yong Sheng Tai Xue Bao = The Journal of Applied Ecology, 33(4), 963–971. https://doi.org/10.13287/j.1001-9332.202204.014 Liang, M., Liu, X., Etienne, R. S., Huang, F., Wang, Y., & Yu, S. (2015). Arbuscular mycorrhizal fungi counteract the Janzen-Connell effect of soil pathogens. Ecology, 96(2), 562–574. https://doi.org/10.1890/14-0871.1 Liang, M., Liu, X., Gilbert, G. S., Zheng, Y., Luo, S., Huang, F., & Yu, S. (2016). Adult trees cause density-dependent mortality in conspecific seedlings by regulating the frequency of pathogenic soil fungi. Ecology Letters, 19(12), 1448–1456. https://doi.org/10.1111/ele.12694 Liu, Q., & Zhao, W. (2023). Plant–soil microbe feedbacks drive seedling establishment during secondary forest succession: The ‘successional stage hypothesis.’ Journal of Plant Ecology, 16(6), rtad021. https://doi.org/10.1093/jpe/rtad021 Liu, Y., & He, F. (2019). Incorporating the disease triangle framework for testing the effect of soil‐borne pathogens on tree species diversity. Functional Ecology, 33(7), 1211–1222. https://doi.org/10.1111/1365-2435.13345 Luo, R., Kuzyakov, Y., Zhu, B., Qiang, W., Zhang, Y., & Pang, X. (2022). Phosphorus addition decreases plant lignin but increases microbial necromass contribution to soil organic carbon in a subalpine forest. Global Change Biology, 28(13), 4194–4210. https://doi.org/10.1111/gcb.16205 MacLean, A. M., Bravo, A., & Harrison, M. J. (2017). Plant signaling and metabolic pathways enabling arbuscular mycorrhizal symbiosis. The Plant Cell, 29(10), 2319–2335. https://doi.org/10.1105/tpc.17.00555 131 Maherali, H. (2020). Mutualism as a plant functional trait: Linking variation in the mycorrhizal symbiosis to climatic tolerance, geographic range, and population dynamics. International Journal of Plant Sciences, 181(1), 9–19. https://doi.org/10.1086/706187 Mangan, S. A., Herre, E. A., & Bever, J. D. (2010). Specificity between Neotropical tree seedlings and their fungal mutualists leads to plant–soil feedback. Ecology, 91(9), 2594– 2603. https://doi.org/10.1890/09-0396.1 Martínez-Vilalta, J. (2014). Carbon storage in trees: Pathogens have their say. Tree Physiology, 34(3), 215–217. https://doi.org/10.1093/treephys/tpu010 McCarthy-Neumann, S., & Ibáñez, I. (2012). Tree range expansion may be enhanced by escape from negative plant–soil feedbacks. Ecology, 93(12), 2637–2649. https://doi.org/10.1890/11-2281.1 McCarthy-Neumann, S., & Ibáñez, I. (2013). Plant–soil feedback links negative distance dependence and light gradient partitioning during seedling establishment. Ecology, 94(4), 780–786. https://doi.org/10.1890/12-1338.1 McCarthy-Neumann, S., & Kobe, R. K. (2008). Tolerance of soil pathogens co-varies with shade tolerance across species of tropical tree seedlings. Ecology, 89(7), 1883–1892. https://doi.org/10.1890/07-0211.1 McCarthy-Neumann, S., & Kobe, R. K. (2010a). Conspecific and heterospecific plant-soil feedbacks influence survivorship and growth of temperate tree seedlings. Journal of Ecology, 98(2), 408–418. https://doi.org/10.1111/j.1365-2745.2009.01620.x McCarthy-Neumann, S., & Kobe, R. K. (2010b). Conspecific plant-soil feedbacks reduce survivorship and growth of tropical tree seedlings. Journal of Ecology, 98(2), 396–407. https://doi.org/10.1111/j.1365-2745.2009.01619.x McCarthy-Neumann, S., & Kobe, R. K. (2019). Site soil-fertility and light availability influence plant-soil feedback. Frontiers in Ecology and Evolution, 7. https://doi.org/10.3389/fevo.2019.00383 McGonigle, T. P., Miller, M. H., Evans, D. G., Fairchild, G. L., & Swan, J. A. (1990). A new method which gives an objective measure of colonization of roots by vesicular— Arbuscular mycorrhizal fungi. New Phytologist, 115(3), 495–501. https://doi.org/10.1111/j.1469-8137.1990.tb00476.x McGuire, K. L. (2007). Common ectomycorrhizal networks may maintain monodominance in a tropical rain forest. Ecology, 88(3), 567–574. https://doi.org/10.1890/05-1173 McNamara, N. P., Black, H. I. J., Beresford, N. A., & Parekh, N. R. (2003). Effects of acute gamma irradiation on chemical, physical and biological properties of soils. Applied Soil Ecology, 24(2), 117–132. https://doi.org/10.1016/S0929-1393(03)00073-8 132 McPherson, K., & Williams, K. (1998). The role of carbohydrate reserves in the growth, resilience, and persistence of cabbage palm seedlings (Sabal palmetto). Oecologia, 117(4), 460–468. https://doi.org/10.1007/s004420050681 Moles, A. T. (2005). A brief history of seed size. Science, 307(5709), 576–580. https://doi.org/10.1126/science.1104863 Myers, J. A., & Kitajima, K. (2007). Carbohydrate storage enhances seedling shade and stress tolerance in a neotropical forest. Journal of Ecology, 95(2), 383–395. https://doi.org/10.1111/j.1365-2745.2006.01207.x Nicholson, R. L., & Hammerschmidt, R. (1992). Phenolic compounds and their role in disease resistance. Annual Review of Phytopathology, 30, 369–389. Niinemets, Ü., & Valladares, F. (2006a). Tolerance to shade, drought, and waterlogging of temperate northern hemisphere trees and shrubs. Ecological Monographs, 76(4), 521– 547. https://doi.org/10.1890/0012-9615(2006)076[0521:TTSDAW]2.0.CO;2 Niinemets, Ü., & Valladares, F. (2006b). Tolerance to shade, drought, and waterlogging of temperate northern hemisphere trees and shrubs. Ecological Monographs, 76(4), 521– 547. https://doi.org/10.1890/0012-9615(2006)076[0521:TTSDAW]2.0.CO;2 Nowacki, G. J., & Abrams, M. D. (2008). The Demise of Fire and “Mesophication” of Forests in the Eastern United States. BioScience, 58(2), 123–138. https://doi.org/10.1641/B580207 O’Hanlon-Manners, D. L., & Kotanen, P. M. (2004). Evidence that fungal pathogens inhibit recruitment of a shade-intolerant tree, white birch (Betula papyrifera), in understory habitats. Oecologia, 140(4), 650–653. https://doi.org/10.1007/s00442-004-1625-0 Osunkoya, O. O., Ash, J. E., Hopkins, M. S., & Graham, A. W. (1994). Influence of seed size and seedling ecological attributes on shade-tolerance of rain-forest tree species in Northern Queensland. The Journal of Ecology, 82(1), 149. https://doi.org/10.2307/2261394 Pacala, S. W., Canham, C. D., Saponara, J., Silander, J. A., Kobe, R. K., & Ribbens, E. (1996). Forest models defined by field measurements: Estimation, error analysis and dynamics. Ecological Monographs, 66(1), 1–43. https://doi.org/10.2307/2963479 Packer, A., & Clay, K. (2000). Soil pathogens and spatial patterns of seedling mortality in a temperate tree. Nature, 404(6775), 278–281. https://doi.org/10.1038/35005072 Packer, A., & Clay, K. (2003). Soil pathogens and Prunus serotina seedling and sapling growth near conspecific trees. Ecology, 84(1), 108–119. https://doi.org/10.1890/0012- 9658(2003)084[0108:SPAPSS]2.0.CO;2 Pérez-Harguindeguy, N., Díaz, S., Garnier, E., Lavorel, S., Poorter, H., Jaureguiberry, P., Bret- Harte, M. S., Cornwell, W. K., Craine, J. M., Gurvich, D. E., Urcelay, C., Veneklaas, E. J., Reich, P. B., Poorter, L., Wright, I. J., Ray, P., Enrico, L., Pausas, J. G., de Vos, A. C., 133 … Cornelissen, J. H. C. (2013). New handbook for standardised measurement of plant functional traits worldwide. Australian Journal of Botany, 61(3), 167. https://doi.org/10.1071/BT12225 Pierik, R., & Ballaré, C. L. (2021). Control of plant growth and defense by photoreceptors: From mechanisms to opportunities in agriculture. Molecular Plant, 14(1), 61–76. https://doi.org/10.1016/j.molp.2020.11.021 Piper, F. I., Reyes-Díaz, M., Corcuera, L. J., & Lusk, C. H. (2009). Carbohydrate storage, survival, and growth of two evergreen Nothofagus species in two contrasting light environments. Ecological Research, 24(6), 1233–1241. https://doi.org/10.1007/s11284- 009-0606-5 Pizano, C., Mangan, S. A., Graham, J. H., & Kitajima, K. (2014). Habitat-specific positive and negative effects of soil biota on seedling growth in a fragmented tropical montane landscape. Oikos, 123(7), 846–856. https://doi.org/10.1111/oik.01032 Pizano, C., Mangan, S. A., Graham, J. H., & Kitajima, K. (2017). Host-specific effects of soil microbial filtrates prevail over those of arbuscular mycorrhizae in a fragmented landscape. Ecological Applications, 27(6), 1946–1957. https://doi.org/10.1002/eap.1579 Plummer, M. (2019). rjags: Bayesian graphical models using MCMC (Version 4-10) [R]. Plummer, M., Stukalov, A., & Denwood, M. (2023). rjags: Bayesian Graphical Models using MCMC (Version 4-14) [R]. https://mcmc-jags.sourceforge.io Poorter, L., Kitajima, K., Mercado, P., Chubiña, J., Melgar, I., & Prins, H. H. T. (2010). Resprouting as a persistence strategy of tropical forest trees: Relations with carbohydrate storage and shade tolerance. Ecology, 91(9), 2613–2627. https://doi.org/10.1890/09- 0862.1 Pozo, M. J., & Azcón-Aguilar, C. (2007). Unraveling mycorrhiza-induced resistance. Current Opinion in Plant Biology, 10(4), 393–398. https://doi.org/10.1016/j.pbi.2007.05.004 Putten, W. H., Bradford, M. A., Pernilla Brinkman, E., Voorde, T. F. J., & Veen, G. F. (2016). Where, when and how plant–soil feedback matters in a changing world. Functional Ecology, 30(7), 1109–1121. https://doi.org/10.1111/1365-2435.12657 Quentin, A. G., Pinkard, E. A., Ryan, M. G., Tissue, D. T., Baggett, L. S., Adams, H. D., Maillard, P., Marchand, J., Landhäusser, S. M., Lacointe, A., Gibon, Y., Anderegg, W. R. L., Asao, S., Atkin, O. K., Bonhomme, M., Claye, C., Chow, P. S., Clément-Vidal, A., Davies, N. W., … Woodruff, D. R. (2015). Non-structural carbohydrates in woody plants compared among laboratories. Tree Physiology, 35(11), 1146–1165. https://doi.org/10.1093/treephys/tpv073 R Core Team. (2020). R: A language and environment for statistical computing (version 4.0.2). R Foundation for Statistical Computing. 134 Reinhart, K. O., & Callaway, R. M. (2006). Soil biota and invasive plants. New Phytologist, 170(3), 445–457. https://doi.org/10.1111/j.1469-8137.2006.01715.x Reinhart, K. O., & Rinella, M. J. (2016). A common soil handling technique can generate incorrect estimates of soil biota effects on plants. New Phytologist, 210(3), 786–789. https://doi.org/10.1111/nph.13822 Reinhart, K. O., Royo, A. A., Kageyama, S. A., & Clay, K. (2010). Canopy gaps decrease microbial densities and disease risk for a shade-intolerant tree species. Acta Oecologica, 36(6), 530–536. https://doi.org/10.1016/j.actao.2010.07.006 Richards, S. A. (2008). Dealing with overdispersed count data in applied ecology. Journal of Applied Ecology, 45(1), 218–227. https://doi.org/10.1111/j.1365-2664.2007.01377.x Rinella, M. J., & Reinhart, K. O. (2018). Toward more robust plant-soil feedback research. Ecology, 99(3), 550–556. https://doi.org/10.1002/ecy.2146 Rogers, L. A., Dubos, C., Cullis, I. F., Surman, C., Poole, M., Willment, J., Mansfield, S. D., & Campbell, M. M. (2005). Light, the circadian clock, and sugar perception in the control of lignin biosynthesis. Journal of Experimental Botany, 56(416), 1651–1663. https://doi.org/10.1093/jxb/eri162 Saffell, B. J., Meinzer, F. C., Woodruff, D. R., Shaw, D. C., Voelker, S. L., Lachenbruch, B., & Falk, K. (2014). Seasonal carbohydrate dynamics and growth in Douglas-fir trees experiencing chronic, fungal-mediated reduction in functional leaf area. Tree Physiology, 34(3), 218–228. https://doi.org/10.1093/treephys/tpu002 Sattler, S., & Funnell-Harris, D. (2013). Modifying lignin to improve bioenergy feedstocks: Strengthening the barrier against pathogens? Frontiers in Plant Science, 4. https://www.frontiersin.org/articles/10.3389/fpls.2013.00070 Schiestl-Aalto, P., Ryhti, K., Mäkelä, A., Peltoniemi, M., Bäck, J., & Kulmala, L. (2019). Analysis of the NSC storage dynamics in tree organs reveals the allocation to belowground symbionts in the framework of whole tree carbon balance. Frontiers in Forests and Global Change, 2. https://www.frontiersin.org/article/10.3389/ffgc.2019.00017 Schreeg, L. A., Kobe, R. K., & Walters, M. B. (2005). Tree seedling growth, survival, and morphology in response to landscape-level variation in soil resource availability in northern Michigan. Canadian Journal of Forest Research, 35, 263–273. Segnitz, R. M., Russo, S. E., Davies, S. J., & Peay, K. G. (2020). Ectomycorrhizal fungi drive positive phylogenetic plant–soil feedbacks in a regionally dominant tropical plant family. Ecology, 101(8), e03083. https://doi.org/10.1002/ecy.3083 Shi, G., Liu, Y., Johnson, N. C., Olsson, P. A., Mao, L., Cheng, G., Jiang, S., An, L., Du, G., & Feng, H. (2014). Interactive influence of light intensity and soil fertility on root- 135 associated arbuscular mycorrhizal fungi. Plant and Soil, 378(1–2), 173–188. https://doi.org/10.1007/s11104-014-2022-z Shirley, H. L. (1943). Is tolerance the capacity to endure shade? Journal of Forestry, 41(5), 339– 345. Sikes, B. A. (2010). When do arbuscular mycorrhizal fungi protect plant roots from pathogens? Plant Signaling & Behavior, 5(6), 763–765. https://doi.org/10.4161/psb.5.6.11776 Simard, S. W., & Durall, D. M. (2004). Mycorrhizal networks: A review of their extent, function, and importance. Canadian Journal of Botany, 82(8), 1140–1165. Smith, L. M., & Reynolds, H. L. (2015). Plant–soil feedbacks shift from negative to positive with decreasing light in forest understory species. Ecology, 96(9), 2523–2532. https://doi.org/10.1890/14-2150.1 Smith, S., & Read, D. (2008). Mycorrhizal Symbiosis. Academic Press. Smith-Ramesh, L. M., & Reynolds, H. L. (2017). The next frontier of plant–soil feedback research: Unraveling context dependence across biotic and abiotic gradients. Journal of Vegetation Science, 28, 484–494. https://doi.org/10.1111/jvs.12519 Song, X., & Corlett, R. T. (2022). Do natural enemies mediate conspecific negative distance- and density-dependence of trees? A meta-analysis of exclusion experiments. Oikos, 2022(5), e08509. https://doi.org/10.1111/oik.08509 Stekhoven, D. J., & Buehlmann, P. (2012). MissForest—Nonparametric missing value imputation for mixed-type data. Bioinformatics, 28(1), 112–118. https://doi.org/10.1093/bioinformatics/btr597 Terborgh, J. (2012). Enemies maintain hyperdiverse tropical forests. The American Naturalist, 179(3), 303–314. https://doi.org/10.1086/664183 Teste, F. P., Kardol, P., Turner, B. L., Wardle, D. A., Zemunik, G., Renton, M., & Laliberté, E. (2017). Plant-soil feedback and the maintenance of diversity in Mediterranean-climate shrublands. Science, 355(6321), 173–176. https://doi.org/10.1126/science.aai8291 Teste, F. P., Karst, J., Jones, M. D., Simard, S. W., & Durall, D. M. (2006). Methods to control ectomycorrhizal colonization: Effectiveness of chemical and physical barriers. Mycorrhiza, 17(1), 51–65. https://doi.org/10.1007/s00572-006-0083-4 Therneau, T. M., & Grambsch, P. M. (2000). Modeling survival data: Extending the Cox model. Springer. Trocha, L. K., Weiser, E., & Robakowski, P. (2016). Interactive effects of juvenile defoliation, light conditions, and interspecific competition on growth and ectomycorrhizal colonization of Fagus sylvatica and Pinus sylvestris seedlings. Mycorrhiza, 26(1), 47–56. https://doi.org/10.1007/s00572-015-0645-4 136 Turner, G. D., Lewis, J. D., Mates-Muchin, J. T., Schuster, W. F., & Watt, L. (2009). Light availability and soil source influence ectomycorrhizal fungal communities on oak seedlings grown in oak- and hemlock-associated soils. Canadian Journal of Forest Research, 39(7), 1247–1258. https://doi.org/10.1139/X09-051 Vaartaja, O. (1962). The relationship of fungi to survival of shaded tree seedlings. Ecology, 43(3), 547–549. https://doi.org/10.2307/1933389 Valladares, F., Gianoli, E., & Gómez, J. M. (2007). Ecological limits to plant phenotypic plasticity. New Phytologist, 176(4), 749–763. https://doi.org/10.1111/j.1469- 8137.2007.02275.x Valladares, F., Laanisto, L., Niinemets, Ü., & Zavala, M. A. (2016). Shedding light on shade: Ecological perspectives of understorey plant life. Plant Ecology & Diversity, 9(3), 237– 251. https://doi.org/10.1080/17550874.2016.1210262 Valladares, F., & Niinemets, Ü. (2008). Shade tolerance, a key plant feature of complex nature and consequences. Annual Review of Ecology, Evolution, and Systematics, 39(1), 237– 257. https://doi.org/10.1146/annurev.ecolsys.39.110707.173506 van der Putten, W. H., Bardgett, R. D., Bever, J. D., Bezemer, T. M., Casper, B. B., Fukami, T., Kardol, P., Klironomos, J. N., Kulmatiski, A., Schweitzer, J. A., Suding, K. N., Van de Voorde, T. F. J., & Wardle, D. A. (2013). Plant-soil feedbacks: The past, the present and future challenges. Journal of Ecology, 101(2), 265–276. https://doi.org/10.1111/1365- 2745.12054 Vierheilig, H. (2004). Further root colonization by arbuscular mycorrhizal fungi in already mycorrhizal plants is suppressed after a critical level of root colonization. Journal of Plant Physiology, 161(3), 339–341. https://doi.org/10.1078/0176-1617-01097 Vierheilig, H., Coughlan, A. P., Wyss, U., & Piché, Y. (1998). Ink and vinegar, a simple staining technique for arbuscular-mycorrhizal fungi. Applied and Environmental Microbiology, 64(12), 5004–5007. https://doi.org/10.1128/AEM.64.12.5004-5007.1998 Violle, C., Enquist, B. J., McGill, B. J., Jiang, L., Albert, C. H., Hulshof, C., Jung, V., & Messier, J. (2012). The return of the variance: Intraspecific variability in community ecology. Trends in Ecology & Evolution, 27(4), 244–252. https://doi.org/10.1016/j.tree.2011.11.014 Violle, C., Navas, M.-L., Vile, D., Kazakou, E., Fortunel, C., Hummel, I., & Garnier, E. (2007). Let the concept of trait be functional! Oikos, 116(5), 882–892. https://doi.org/10.1111/j.0030-1299.2007.15559.x Wallis, C. M., & Galarneau, E. R.-A. (2020). Phenolic compound induction in plant-microbe and plant-insect interactions: A meta-analysis. Frontiers in Plant Science, 11. https://www.frontiersin.org/article/10.3389/fpls.2020.580753 137 Warneke, C. R., Yelenik, S. G., & Brudvig, L. A. (2023). Fire modifies plant–soil feedbacks. Ecology, 104(5), e3994. https://doi.org/10.1002/ecy.3994 Waterman, P. G., & Mole, S. (1994). Analysis of phenolic plant metabolites. Blackwell Scientific. https://scholar.google.com/scholar_lookup?title=Analysis+of+phenolic+plant+metabolite s&author=Waterman%2C+Peter+G.&publication_year=1994 Waterman, P., & Mole, S. (1994). Analysis of phenolic plant metabolites. Blackwell Scientific. Whipps, J. M. (2004). Prospects and limitations for mycorrhizas in biocontrol of root pathogens. Canadian Journal of Botany, 82(8), 1198–1227. https://doi.org/10.1139/b04-082 Wipf, D., Krajinski, F., Tuinen, D. van, Recorbet, G., & Courty, P.-E. (2019). Trading on the arbuscular mycorrhiza market: From arbuscules to common mycorrhizal networks. New Phytologist, 223(3), 1127–1142. https://doi.org/10.1111/nph.15775 Witzell, J., & Martín, J. A. (2008). Phenolic metabolites in the resistance of northern forest trees to pathogens—Past experiences and future prospects. Canadian Journal of Forest Research, 38(11), 2711–2727. https://doi.org/10.1139/X08-112 Wood, K. E. A., Kobe, R. K., & McCarthy-Neumann, S. (2023). Tree seedling shade tolerance arises from interactions with microbes and is mediated by functional traits. Frontiers in Ecology and Evolution. Wood, K. E. A., Richard K Kobe, McCarthy-Neumann, S., & Inés Ibáñez. (2023). Tree seedling functional traits mediate plant-soil feedback survival responses across a gradient of light availability. Plos One. Xi, N., Adler, P. B., Chen, D., Wu, H., Catford, J. A., Bodegom, P. M., Bahn, M., Crawford, K. M., & Chu, C. (2021). Relationships between plant–soil feedbacks and functional traits. Journal of Ecology, 1365-2745.13731. https://doi.org/10.1111/1365-2745.13731 Xi, N., McCarthy-Neumann, S., Feng, J., Wu, H., Wang, W., & Semchenko, M. (2023). Light availability and plant shade tolerance modify plant-microbial interactions and feedbacks in subtropical trees. New Phytologist, 238, 393–404. https://doi.org/10.1111/nph.18737 Yang, H., Zang, Y., Yuan, Y., Tang, J., & Chen, X. (2012). Selectivity by host plants affects the distribution of arbuscular mycorrhizal fungi: Evidence from ITS rDNA sequence metadata. BMC Evolutionary Biology, 12(1), 50. https://doi.org/10.1186/1471-2148-12- 50 Yang, J., Cao, M., & Swenson, N. G. (2018). Why functional traits do not predict tree demographic rates. Trends in Ecology & Evolution, 33(5), 326–336. https://doi.org/10.1016/j.tree.2018.03.003 138 Zamioudis, C., & Pieterse, C. M. J. (2012). Modulation of host immunity by beneficial microbes. Molecular Plant-Microbe Interactions®, 25(2), 139–150. https://doi.org/10.1094/MPMI- 06-11-0179 Zhang, M., Zhu, J., Li, M., Zhang, G., & Yan, Q. (2013). Different light acclimation strategies of two coexisting tree species seedlings in a temperate secondary forest along five natural light levels. Forest Ecology and Management, 306, 234–242. https://doi.org/10.1016/j.foreco.2013.06.031 Zhu, Y., Li, G., Singh, J., Khan, A., Fazio, G., Saltzgiver, M., & Xia, R. (2021). Laccase directed lignification is one of the major processes associated with the defense response against Pythium ultimum infection in apple roots. Frontiers in Plant Science, 12. https://www.frontiersin.org/articles/10.3389/fpls.2021.629776 139 APPENDIX In Chapter 2, to analyze seedling survival, the hazard was estimated for each time step, ℎ", from a gamma distribution with noninformative parameter values, ℎ" ~ 𝑔𝑎𝑚𝑚𝑎(1,1). This intrinsic mortality rate reflects the temporal variability in mortality that is not accounted for by the risk function 𝜇!". The risk, 𝜇!", was estimated as a function of the covariates included in the analysis, 𝜇!" = 𝑿!"𝑩. 𝑿!" is the matrix of covariates associated with each seedling 𝑖 at each time 𝑡. 𝑩 is the vector of fixed effect coefficients associated with each covariate. These coefficients were estimated from normal distributions with noninformative parameter values, 𝑩 ~ 𝑛𝑜𝑟𝑚𝑎𝑙(0,0.0001). For the final model, covariates included microbe treatment and light level; in initial models, we included greenhouse bench and adult tree as random effects, but their inclusion did not improve fit of the model. 140 Figure A2.1 Parameter estimated (mean posterior values ± 95% credible intervals) for fixed- effect coefficients (light availability x microbe treatment; aka risk) for each species (Acer saccharum, A. rubrum, A. negundo). Figure A2.2 Hazard curves (mean ± 95% credible intervals) over the 9-week study period, for each species (Acer saccharum, A. rubrum, A. negundo). 141 Figure A2.3 Predicted survival (mean ± 95 credible intervals for each species, light level, and microbial filtrate treatment, over the 9-week study period. 142 Table A2.1 Results of fixed effects linear regression model testing the effect of light availability and microbial filtrate size on seedling traits. Species Trait Light Microbe Light x Microbe ACSA AMF Phenolics Lignin NSC (1, 240) = 0.00, c2 p = 0.99 (1, 240) = 2.80, c2 p = 0.09 (1, 240) = 4.24, c2 p = 0.04 (3, 240) = 3305.84, c2 p < 0.001 (3, 240) = 20.93, c2 p < 0.001 (3, 240) = 53.78, c2 p < 0.001 (3, 240) = 30.46, c2 p < 0.001 (3, 240) = 33.35, c2 p < 0.001 (1, 240) = 14.43, c2 p < 0.001 (3, 240) = 11.20, c2 p = 0.01 (3, 240) = 12.97, c2 p = 0.005 ACRU AMF (3, 232) = 2088.96, c2 p < 0.001 Phenolics (1, 232) = 74.21, c2 p < 0.001 (3, 232) = 70.48, c2 p < 0.001 (3, 232) = 20.71, c2 p < 0.001 Lignin NSC (1, 232) = 3.99, c2 p = 0.046 (3, 232) = 13.43, c2 p = 0.004 (3, 232) = 9.44, c2 p = 0.02 (1, 232) = 35.30, c2 p < 0.001 (3, 232) = 88.27, c2 p < 0.001 (3, 232) = 33.96, c2 p < 0.001 ACNE AMF (1, 240) = 0.02, c2 p = 0.90 (3, 240) = 4970.54, c2 p < 0.001 (3, 240) = 68.59, c2 p < 0.001 Phenolics (1, 240) = 40.91, c2 p < 0.001 (3, 240) = 41.76, c2 p < 0.001 Lignin NSC (1, 240) = 0.00, c2 p = 1.00 (3, 240) = 24.01, c2 p < 0.001 (1, 240) = 193.78, c2 p < 0.001 (3, 240) = 39.85, c2 p < 0.001 (3, 240) = 14.50, c2 p = 0.002 (3, 240) = 10.29, c2 p = 0.02 143 Figure A2.4 Correlation matrix for imputated seedling functional traits, pooled across species, light availability, and microbial filtrate treatments. Note that the “None” and “<40 µm” treatments were included for AMF correlations. 144 Table A2.2 Summary table demonstrating A) tree seedling survival in low versus high light in response to soil-borne microbes, B) functional trait responses in high versus low light in response to soil-borne microbes, and C) tree seedling survival responses across all light levels in response to functional traits. Arrows indicate the direction of the effect: ↑ increasing and ↓ decreasing. A) Tree seedling survival in low versus high light in response to soil-borne microbes Sterile Mycorrhizae Pathogens Both1 B) Functional trait responses in high versus low light in response to soil-borne microbes Shade Tolerant (Acer saccharum) Intermediate - ↓ ↓ ↓ Intolerant - ↓ ↓ ↓ Tolerant - - - ↓ Combined1 ↑ ↑ - ↑ Combined1 - - - ↑ Combined1 ↑ ↑ - ↑ Overall ↑ ↑ ↑ ↑ AMF Phenolics Lignin NSC Intermediate (Acer rubrum) Sterilized - - ↑ Sterilized Overall - AMF ↑ ↑ Phenolics - - Lignin NSC ↑ ↑ Shade Intolerant (Acer negundo) 40-250µm ↑ ↑ - ↑ 40-250µm - ↑ - ↑ <20µm - - ↑ <20µm - - ↑ <20µm Sterilized AMF Phenolics Lignin NSC C) Tree seedling survival across all light levels in response to functional traits - - ↑ ↑ - ↑ 40-250µm ↑ ↑ ↑ ↑ Overall ↑ ↑ - ↑ AMF Phenolics Lignin NSC2 Overall - ↑ - ↑ Tolerant - ↑ ↑ ↑ Intermediate - ↑ - ↑ Intolerant ↑ ↑ - ↑ 145 In Chapter 3, we collected intact soil cores during the summers of 2016 and 2017. Before transplanting the soil cores into the common garden field plots, we stored the cores inside the research field station. Cores were stored after mesh had been glued to the bottom and 2 open sides of the enclosing pots. We covered the open top of the pot with a plastic lid specifically designed to fit our pots. This ensured that there was no potential for contamination of the pots before moving them back to the field. Table A3.1 Percent of total soil cores collected in 2016 and 2017. For each seedling species and soil source. We collected intact soil cores during the summers of 2016 and 2017. Before transplanting the soil cores into the common garden field plots, we stored the cores inside the research field station. Cores were stored after mesh had been glued to the bottom and 2 open sides of the enclosing pots. We covered the open top of the pot with a plastic lid specifically designed to fit our pots. This ensured that there was no potential for contamination of the pots before moving them back to the field. Soil source / collection year Acru Acsa Pogr Prse Qual Quru Con. St. Pogr Qual Acsa Quru Prse 2016 2017 2016 2017 2016 2017 2016 2017 2016 2017 82 35 62 62 23 73 94 15 61 43 72 23 87 53 65 38 27 85 57 77 47 33 57 72 12 39 25 76 18 38 77 6 39 18 13 32 54 75 13 43 17 10 82 61 25 88 58 76 82 67 43 28 88 61 75 24 68 46 25 87 57 83 90 18 39 75 12 42 24 18 146 We were concerned that storing soil cores for an extended period of time would have potential negative effects on the microbial community. Specifically, we worried that the soil microbial community would be adversely affected. Preliminary analyses did not indicate any significant effect of soil collection year on seedling trait expression (AMF or EMF colonization, phenolics, lignin, NSC) or survival. Figure A3.1 Prelimninary boxplots showing the effect of soil core year (2016, 2017) on tree seedling traits: A) AMF colonization, B) EMF colonization, C) Phenolics, D) Lignin, E) NSC. We were concerned that storing soil cores for an extended period of time would have potential negative effects on the microbial community. Specifically, we worried that the soil microbial community would be adversely affected. Preliminary analyses did not indicate any significant effect of soil collection year on seedling trait expression (AMF or EMF colonization, phenolics, lignin, NSC) or survival. 147 Nutrient availability We tested if there were differences in soil nutrient supply rates (e.g., NH4 +, NO3 −, PO4 3−, K+, SO4 2−, Ca2+, Mg2+, Al3+, Fe3+, Cu+, Zn+, B3+, Mn2+, Pb4+, and Cd2+) by soil source treatments (between undisturbed soil under adult trees where soil was collected and in the fungal exclusion pots) and sterilization treatment (non-sterilized vs. sterilized conspecific soil in fungal exclusion pots) with plant root simulator (PRS™) probes (Western Ag Innovations Inc., Saskatoon, SK) at 0-7 cm depth. Four replicate PRS probes were installed under the adult trees we collected soil from (6 adults trees for Acer saccharum and Quercus rubra and 3 adult trees for Acer rubrum, Populus granditentata, Prunus serotina and Querucs alba). In addition, PRS probes were installed in 240 fungal exclusion pots planted with A. saccharum seedlings [(6 non-sterilized soil sources x adult trees (6 for Acer saccharum and Quercus rubra and 3 for the other 4 tree species) x 2 light treatments (low and high light field plots) x 4 seedling replicates) + (Acer saccharum sterilized soil x 6 adult trees x 2 light treatments x 4 seedlings replicates)]. PRS probes were installed 3-wks after planting and harvested 3-wks later. “PRS probes are ion exchange resin membranes held in plastic supports that are easily inserted into soil to measure ion supply in situ with minimal disturbance. Anion probes have a positively-charged membrane to simultaneously attract and adsorb all negatively-charged anions... Cation probes have a negatively-charged membrane to simultaneously attract and adsorb all positive-charged cations… Prior to use, ion exchange membranes are saturated with a counter-ion that is easily desorbed, allowing ready absorption of soil ions. Anion probes are saturated with HCO3 - and cation probes are saturated with Na+. When buried, soil ions displace the counter-ions at a rate that depends on their activity and diffusion rate in soil solution. The quantity of soil ions adsorbed during a burial period is a function of all soil properties (physical, 148 chemical, and biological) controlling nutrient availability in soil.” (https://www.westernag.ca/innovations/technology/basics) Due to systemic error in lab processing, sample sizes for some of the treatments were greatly reduced. Subsequent t-tests were conducted with pooled datasets at the nutrient or species level, rather than paired t-tests at the adult tree level. Table A3.2 Results of t-tests comparing nutrient supply rate (micrograms / 10cm2 / burial length) in seedling pots versus beneath adult trees. Bolded values are significant at P < 0.0036, Bonferroni-corrected to a = 0.0036, for original a = 0.05 and n = 14 tested nutrients. Nutrient df Supply rate Al+ B3+ Ca2+ Cu3+ Fe2+ K+ Mg2+ Mn2+ + NH4 NO3 P3- Pb2+ S2+ Zn2+ - 32 18 32 13 24 19 28 29 12 23 21 32 20 24 in pots 15.5 0.61 1849 0.57 33.0 170 346 17.7 4.1 199 4.9 2.6 65 13.7 Supply rate beneath adult trees 14.7 0.68 2142 0.51 17.1 267 371 14.5 5.9 50 8.2 2.1 26 5.1 t p 0.647 0.463 0.655 -0.454 0.106 -1.662 0.668 0.439 0.019 2.509 0.081 -1.838 0.515 -0.660 0.432 0.797 -0.869 0.402 4.280 < 0.001 -3.089 0.006 0.368 0.914 4.998 < 0.001 5.033 < 0.001 149 Soil sterilization had some impacts on nutrient availability (Table A3.3). Sterilized soils had higher amounts of Mn2+, NO3 -, and P3-, which could have potentially contributed to increased survival or biomass that would confound the effects of soil-borne microbes. However, we found limited effects of soil source, regardless of sterilization, on seedling survival and biomass, indicating that other factors (i.e., interactions between pot design and heavy rainfall events) had a larger effect than nutrient availability or soil-borne microbes. Table A3.3 Results of t-tests comparing nutrient supply rates (micrograms / 10cm2 / burial length) in sterilized versus live soil, in seedling pots. Alpha was Bonferroni-corrected to a = 0.0038, for original a = 0.05 and n = 13 tested nutrients. P-values marked with * are marginally significant at original a = 0.05. Nutrient df Al+ B3+ Ca2+ Cu3+ Fe2+ K+ Mg2+ Mn2+ + NH4 NO3 P3- Pb2+ S2+ Zn2+ - 9 4 5 9 16 3 3 2 3 5 10 6 2 Supply rate in sterilized soil 17 Supply rate in live soil 15 t p -1.117 0.294 1733 0.43 31 137 360 40 3.0 381 6.7 2.5 44 21 1844 0.53 32 189 342 13 4.4 155 4.4 2.5 69 12 0.478 0.656 0.885 0.420 0.174 0.866 2.213 0.050 -0.368 0.736 -5.647 0.008 * 1.24 -4.083 0.024 * -0.357 0.018 * 0.071 0.945 2.054 0.089 -1.409 0.287 0.370 Table A3.4 Full dataset results of t-tests comparing nutrient supply rates (micrograms / 10cm2 / burial length) in sterilized versus live soil, in seedling pots. For many nutrients, there were not enough replicate pots for a t-test or sample size was small, so we also provide the results of a t- test using the full dataset. Df 56 Supply rate in sterilized soil 213 Supply rate in non-sterilized soil 200 t p -0.160 0.874 150 We were concerned that the design of the seedling pots might affect soil nutrient availability. However, we found that soils in seedling pots had similar nutrient availability to undisturbed soils beneath adult trees. Figure A3.2 Nutrient supply rate (micrograms / 10cm2 / burial length) in seedling pots versus beneath adult trees (Adj.-R2 = 0.92, P = 0.03). Figure A3.3 Nutrient supply rates (micrograms / 10cm2 / burial length) in sterilized versus live soil, in seedling pots. 151 Adult trees p F Nutrient df 0.383 1.128 5 0.113 2.167 5 0.003 5.678 5 0.010 7.812 4 0.167 1.8 5 0.046 2.289 5 9.559 < 0.001 5 0.577 0.781 5 0.263 1.487 4 0.667 0.647 5 0.372 1.152 5 0.028 5 3.386 0.424 5 1.1045 0.166 1.804 5 Al+ B3+ Ca2+ Cu3+ Fe2+ K+ Mg2+ Mn2+ + NH4 NO3 P3- Pb2+ S2+ Zn2+ - Seedling pots p F df 5 0.458 0.8 5 0.806 0.571 5 0.629 0.682 Table A3.5 Results of ANOVAs comparing nutrient supply rates (micrograms / 10cm2 / burial length) in each soil source (Acru, Acsa, Prse, Pogr, Qual, Quru) in soil beneath adult trees. Adult trees were used as a proxy for seedling pots, since preliminary analyses showed that, for most nutrients, there were no significant differences between nutrient supply rates for soil in seedling -, S2+, and Zn2+, there were significant differences in pots versus beneath adult trees. For NO3 nutrient supply rate for soil in seedling pots versus adult trees, so we also provide results of ANOVAs for seedling pots. Bolded values are significant at a = 0.0036, Bonferroni-corrected for original a = 0.05 and n = 14 tested nutrients. 152 There were few differences in nutrient availability between soil sources (Figure A3.4). Ca2+ was highest in Q. alba soils and Mg2+ was higher in A. saccharum, then P. grandidentata and P. serotina soils. Figure A3.4 Nutrient supply rates (micrograms / 10cm2 / burial length) in each soil source (Acru, Acsa, Prse, Pogr, Qual, Quru) in soil beneath adult trees. For A) Ca2+ and B) Mg2+. 153 Figure A3.5 Light availability in the 18 experimental field plots. Indirect site factor (ISF, the proportion of diffuse solar radiation at a given location, relative to the amount of diffuse solar radiation in the open) in each subplot (n = 5) per common garden plot (n = 18). For analyses in which light availability was included as a categorical variable, low = 0.032-0.075 ISF, medium = 0.075-0.118 ISF, and high = 0.118 – 0.161 ISF. 154 Percent colonization by mycorrhizal fungi To quantify mycorrhizal colonization, prior to drying seedlings, 5-10 root fractions per individual (1cm sections of wet root), were retained, weighed, and stained with 5% Schaeffer black in in vinegar solution (Vierheilig et al., 1998). Percent root colonization by AMF was quantified by inspecting 100 intersections between the microscope eyepiece crosshairs and roots for AMF structures (i.e., vesicles, arbuscules, coils, and hyphae) every 1mm at 200x magnification (McGonigle et al., 1990). AMF fungal structures were distinguished from other fungi that can inhabit the root interior (e.g., dark septate fungi) by comparing slides to established reference images. Percent root colonization by EMF was quantified by counting the number of intact root tips with and without Hartig nets at 100x magnification every 2mm along the root until 100 root tips were scored. Phenolics To quantify phenolics, we collected hypocotyl samples, cut into <1mm pieces. We extracted phenolics in 5mL methanol in the dark for 16 hours at room temperature. The methanol extracts were filtered and adjusted to 5mL, and then we quantified total phenolics using a microplate-adapted colorimetric total phenolics assay with Folin-Ciocalteu reagent (Ainsworth & Gillespie, 2007; P. Waterman & Mole, 1994). Lignin To quantify lignin, root and stem samples were lyophilized and coarsely ground at 1mm using a Wiley Mill. We ran 0.5g root and stem samples through a series of extractions using an ANKOM Fiber Analyzer (ANKOM Technologies, Macedon, NY, USA). We used a Neutral Detergent Fiber extraction to wash off soluble cell contents (e.g., carbohydrates, lipids, pectin, starch, and soluble proteins). We then used an Acid Detergent Fiber extraction with 1.00 normal sulfuric acid to wash off hemicellulose and bound proteins and an Acid Detergent Lignin extraction with 72% sulfuric acid to wash off cellulose, leaving only lignin and recalcitrant materials. Finally, we ashed the samples to quantify dry mass lignin. Nonstructural carbohydrates To quantify nonstructural carbohydrates (NSC), we analyzed stem samples, using a standardized enzyme method for sugar and starch extraction and quantification (Landhäusser, Chow, Turin Dickman, et al., 2018; Quentin et al., 2015). We dried seedling stems and peach leaf standard reference material (MillporeSigma-NIST1547) at 60°C overnight to remove moisture. We then weighed out 30mg of each for analysis and separated sugars and starches with hot ethanol extraction. We used a-amylase and amyloglucosidase to convert starch to glucose. We quantified sugars using phenol-sulfuric acid colorimetric assay and starches using a glucose- hexokinase colorimetric assay (MillporeSigma-GAK20). We calculated total NSC concentrations as the sum of soluble sugar and starch concentrations derived from the assays. 155 Table A3.6 Linear model evaluating the effects of sterilized versus live soil on traits. AMF colonization, EMF colonization, phenolics, lignin, and NSC. For post-hoc comparisons within species, we used joint tests of estimated marginal means. A. saccharum P. serotina Q. alba Q. rubra 0.043 <.001 0.302 p p p p F F F F df - - - - - - 4.109 54.45 1.069 0.012 <.001 0.780 0.031 <.001 0.573 1, 375 1, 375 1, 375 1, 793 1, 793 1, 793 0.365 <.001 0.0127 6.366 98.783 0.078 0.822 34.349 6.242 4.698 275.163 0.317 Response Parameters AMF col. Light Soil Light x Soil EMF col. Light Soil Light x Soil Phenolics Light Soil Light x Soil Lignin Light Soil Light x Soil NSC 207.577 Light 0.998 Soil Light x Soil 9.096 Bonferroni correction for species: alpha level is set to 0.0125. Bolded values are statistically significant at p < 0.0125. Underlined values are marginally significant at p < 0.05. 573.880 1394.556 86.486 649.5881 10.389 28.589 13.670 270.962 3.742 206.174 0.613 0.414 180.293 19.296 1.908 123.267 52.762 0.010 51.792 313.466 35.308 14.497 21.545 4.221 0.959 89.096 1.231 15.844 0.942 0.024 77.208 35.188 0.011 75.411 5.171 0.000 0.064 64.289 21.506 <.001 <.001 0.0538 1, 793 1, 793 1, 793 1, 793 1, 793 1, 793 <.001 <.001 0.040 0.328 <.001 0.268 <.001 <.001 <.001 <.001 0.332 0.876 <.001 0.318 0.003 <.001 <.001 0.917 <.001 <.001 0.168 0.780 <.001 <.001 <.001 <.001 0.919 1,793 1,793 1,793 - - - - - - <.001 0.023 0.995 <.001 0.434 0.520 <.001 0.001 <.001 <.001 <.001 <.001 156 A) B) C) Figure A3.6 Effect of sterilized versus live soil on seedling traits. A) AMF colonization, B) EMF colonization, C) phenolics, D) lignin, E) NSC. 157 Figure A3.6 (Cont’d) D) E) 158 Table A3.7 Linear model evaluating the effects of light availability and soil source (conspecific versus pooled heterospecific) on traits (AMF colonization, EMF colonization, phenolics, lignin, and NSC). Sterilized soil was excluded from the model. For post-hoc comparisons within species, we used joint tests of estimated marginal means. A. saccharum P. serotina Q. alba Q. rubra 0.011 0.100 0.492 p p p p F F F F df - - - - - - 6.083 7.012 1.002 6.420 2.710 0.471 0.014 0.008 0.317 2.421 4.205 0.889 <.001 <.001 <.001 35.056 27.420 17.640 0.1199 0.040 0.346 1, 2344 1, 2344 1, 2344 1, 1063 1, 1063 1, 1063 Response Parameters AMF col. Light Soil Light x Soil EMF col. Light Soil Light x Soil Phenolics Light Soil Light x Soil Lignin Light Soil Light x Soil NSC Light 287.753 Soil 0.018 6.346 Light x Soil Bonferroni correction for species: alpha level is set to 0.0125. Bolded values are statistically significant at p < 0.0125. Underlined values are marginally significant at p < 0.05. 16.388 106.576 4.785 187.073 16.687 0.162 64.695 118.383 9.353 184.507 0.338 1.041 249.422 0.001 170.690 1, 2344 1, 2344 1, 2344 1, 2344 1, 2344 1, 2344 1, 2344 1, 2344 1, 2344 22.048 0.128 0.012 63.024 8.296 19.968 82.338 0.089 0.001 87.418 19.305 32.411 34.475 51.414 2.852 21.421 14.161 14.176 0.482 70.875 0.328 <.001 0.893 0.012 <.001 <.001 0.029 <.001 <.001 0.688 <.001 0.561 0.308 <.001 <.001 0.091 0.488 <.001 0.567 <.001 0.721 0.914 <.001 0.765 0.975 <.001 <.001 0.002 0.252 0.006 0.642 1.313 7.639 0.216 - - - - - - <.001 0.004 <.001 <.001 0.976 <.001 <.001 <.001 <.001 <.001 <.001 <.001 159 Table A3.8 Linear model evaluating the effects of light availability and soil source on traits (AMF colonization, EMF colonization, phenolics, lignin, and NSC). Sterilized soil was excluded from the model. For post-hoc comparisons within species, we used joint tests of estimated marginal means. A. saccharum P. serotina Q. alba Q. rubra 0.002 <.001 0.684 p p p p F F F F df - - - - - - 0.188 <.001 0.850 <.001 <.001 <.001 0.003 <.001 <.001 8.746 16.170 4.416 9.374 14.373 0.620 1.733 12.391 0.399 1, 2312 5, 2312 5, 2312 1, 1047 5, 1047 5, 1047 158.201 68.831 5.388 Response Parameters AMF col. Light Soil Light x Soil EMF col. Light Soil Light x Soil Phenolics Light Soil Light x Soil Lignin Light Soil Light x Soil NSC Light 451.673 Soil 8.146 10.567 Light x Soil Bonferroni correction for species: alpha level is set to 0.0125. Bolded values are statistically significant at p < 0.0125. Underlined values are marginally significant at p < 0.05. 830.835 274.609 2.476 146.587 195.305 6.964 156.314 4.082 1.404 313.426 4.667 4.644 261.664 135.613 92.102 196.841 63.129 1.518 1, 2312 5, 2312 5, 2312 1, 2312 5, 2312 5, 2312 1, 2312 5, 2312 5, 2312 42.638 1.557 0.527 48.610 4.076 5.746 79.157 35.680 13.492 11.809 36.578 6.542 1.320 10.023 1.614 62.774 52.997 6.235 2.374 30.003 2.620 0.251 <.001 0.154 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 0.030 0.124 <.001 0.023 <.001 <.001 0.181 <.001 0.169 0.756 <.001 <.001 <.001 <.001 0.001 0.220 - - - - - - <.001 0.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 160 A) B) C) Figure A3.7 Effects of light availability and soil source on seedling traits. A) AMF colonization, B) EMF colonization, C) phenolics, D) lignin, and E) NSC. 161 Figure A3.7 (Cont’d) D) E) 162 Figure A3.8 Correlations between AMF colonization, EMF colonization (for Q. alba and Q. rubra), phenolics, lignin, and non-structural carbohydrates. 163 Table A3.9 Cox proportional hazards survival models evaluating the effects of light availability and soil source (sterilized versus live conspecific) on seedling survival. Individual models were performed for each species. A. saccharum P. serotina Q. alba Q. rubra p df LRc2 Response Parameters Light Soil source Light ´ Soil Bonferroni correction for species: alpha level is set to 0.0125. Bolded values are statistically significant at p < 0.0125. Underlined values are marginally significant at p < 0.05. 0.755 0.097 61.781 <.001 0.814 0.055 0.072 0.327 0.613 6.254 0.020 0.560 1.630 1.521 2.82 0.202 0.001 0.093 3.236 0.959 0.256 LRc2 LRc2 LRc2 1 1 1 p p 0.012 0.920 0.454 p Table A3.10 Cox proportional hazards survival models evaluating the effects of light availability and soil source (conspecific versus pooled heterospecific) on seedling survival. Individual models were performed for each species. A. saccharum LRc2 0.701 8.604 0.050 Parameters Light Soil source Light ´ Soil Bonferroni correction for species: alpha level is set to 0.0125. Bolded values are statistically significant at p < 0.0125. Underlined values are marginally significant at p < 0.05. p 0.383 0.176 0.237 p 0.402 0.003 0.823 LRc2 9.021 2.214 0.578 LRc2 4.094 1.015 0.192 LRc2 0.760 1.834 1.397 df 1 1 1 P. serotina p 0.043 0.314 0.192 Q. rubra p 0.003 0.137 0.447 Q. alba Table A3.11 Cox proportional hazards survival models evaluating the effects of light availability and soil source (conspecific versus unpooled heterospecific) on seedling survival. Individual models were performed for each species. Sterilized soil was excluded from the model. A. saccharum P. serotina Q. alba Q. rubra p df LRc2 Response Parameters Light Soil source Light ´ Soil Bonferroni correction for species: alpha level is set to 0.0125. Bolded values are statistically significant at p < 0.0125. Underlined values are marginally significant at p < 0.05. 0.422 0.644 29.913 <.001 0.962 0.002 4.655 5.698 6.860 0.331 0.246 0.853 9.046 5.121 0.806 0.031 0.017 0.009 0.933 1.346 0.034 LRc2 LRc2 LRc2 1 1 1 p p 0.003 0.024 0.369 p 164 Figure A.3.9 Survival curves for each species, soil source, and light availability. 165 Figure A4.1 Differences in predicted seedling survival when grown at high light availability. Data are means with 95% credible intervals; credible intervals that do not overlap with the zero line are statistically significant. Differences in survival above the zero line indicate a positive PSF (higher survival in conspecific than heterospecific soils); differences in survival below the zero line indicate negative PSFs (lower survival in conspecific than heterospecific soils). 166 Figure A4.2 Differences in predicted seedling survival when grown at low light availability. Data are means with 95% credible intervals; credible intervals that do not overlap with the zero line are statistically significant. Differences in survival above the zero line indicate a positive PSF (higher survival in conspecific than heterospecific soils); differences in survival below the zero line indicate negative PSFs (lower survival in conspecific than heterospecific soils). 167 Table A4.2 Summary data (mean ± standard deviation) for percent mycorrhizal colonization by AMF and EMF, phenolics, lignin, and nonstructural carbohydrates. Data are provided for each species ´ soil source ´ light level. EMF Species ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACSA ACSA ACSA Soil ACRU ACRU ACRU ACSA ACSA ACSA POGR POGR POGR PRSE PRSE PRSE QUAL QUAL QUAL QURU QURU QURU StCon StCon StCon ACRU ACRU ACRU Light Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High AMF 38.82 ± 4.07 42.62 ± 3.64 46.27 ± 1.9 36.73 ± 3.13 31.45 ± 3.14 31.67 ± 7.34 33.08 ± 2.72 25.62 ± 3.1 30.38 ± 6.33 29.92 ± 5.98 38.58 ± 8.28 35 ± 8.26 31.36 ± 2.82 28.57 ± 6.8 34.31 ± 8.02 32.71 ± 4.95 33.64 ± 7.72 35.57 ± 8.43 0 ± 0 0.15 ± 0.38 0 ± 0 60.5 ± 6.54 68.31 ± 5.65 74 ± 5.51 168 Phenolics 0.15 ± 0.06 0.3 ± 0.16 0.34 ± 0.12 -0.26 ± 0.26 -0.16 ± 0.21 -0.14 ± 0.31 -0.13 ± 0.28 -0.25 ± 0.27 0.08 ± 0.22 -0.35 ± 0.23 -0.1 ± 0.23 0.06 ± 0.24 -0.31 ± 0.22 -0.06 ± 0.28 0.07 ± 0.28 -0.25 ± 0.18 -0.12 ± 0.16 -0.01 ± 0.14 -0.07 ± 0.05 0.07 ± 0.16 0.19 ± 0.06 -0.37 ± 0.32 -0.01 ± 0.25 -0.1 ± 0.39 Lignin 5.1 ± 0.23 10.1 ± 0.86 8.6 ± 1.16 5.37 ± 1.15 9.09 ± 1.5 8.8 ± 1.08 5.43 ± 0.73 8.17 ± 0.91 8.3 ± 0.77 6.34 ± 1.06 8.47 ± 1.56 8.44 ± 1.07 5.11 ± 0.92 9.04 ± 1.89 8.14 ± 1.6 5.6 ± 1.4 8.61 ± 2.02 8.26 ± 1.41 5.71 ± 0.61 7.41 ± 0.93 9.61 ± 1.18 11.59 ± 1.4 13.7 ± 0.57 13.87 ± 0.73 NSC 6.97 ± 0.69 12.14 ± 1.68 12.61 ± 0.71 8.96 ± 2.18 11.67 ± 1.5 11.96 ± 1.11 7.84 ± 1.14 10.98 ± 2.4 11.85 ± 1.12 8.97 ± 1.53 13.56 ± 1.74 12.18 ± 1.32 8.51 ± 1.97 12.96 ± 0.91 12.04 ± 1.35 8.42 ± 1.54 11.24 ± 1.56 12.57 ± 1.46 8.68 ± 1.29 10.61 ± 0.89 12.64 ± 0.75 11.77 ± 0.94 12.11 ± 0.29 12.54 ± 0.34 Table A4.2 (Cont’d) ACSA ACSA ACSA ACSA ACSA ACSA ACSA ACSA ACSA ACSA ACSA ACSA ACSA ACSA ACSA ACSA ACSA ACSA PRSE PRSE PRSE PRSE PRSE PRSE PRSE PRSE PRSE ACSA ACSA ACSA POGR POGR POGR PRSE PRSE PRSE QUAL QUAL QUAL QURU QURU QURU StCon StCon StCon ACRU ACRU ACRU ACSA ACSA ACSA POGR POGR POGR Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High 66.64 ± 5.68 67.83 ± 5.41 74.93 ± 4.46 49.85 ± 7.72 65 ± 3.65 67.29 ± 4.7 61.54 ± 3.62 67 ± 2.45 71.53 ± 5.64 57.64 ± 5.05 64.93 ± 5.31 68.43 ± 4.11 56.5 ± 5.08 68.67 ± 5.46 70.29 ± 5.01 0 ± 0 0 ± 0 0.07 ± 0.26 61 ± 3.19 59.91 ± 3.51 69.62 ± 8.61 59.73 ± 5.1 62.55 ± 4.52 70.08 ± 3.7 56.08 ± 3.68 57.18 ± 5.13 64.62 ± 5.12 -0.15 ± 0.15 0.33 ± 0.18 0.4 ± 0.12 -0.42 ± 0.3 0.01 ± 0.23 -0.08 ± 0.52 -0.28 ± 0.32 -0.11 ± 0.19 0.13 ± 0.52 -0.18 ± 0.41 -0.09 ± 0.25 -0.1 ± 0.46 -0.28 ± 0.12 -0.15 ± 0.46 0.1 ± 0.39 -0.34 ± 0.13 -0.02 ± 0.09 0.21 ± 0.13 0.51 ± 0.14 0.73 ± 0.12 0.99 ± 0.08 0.5 ± 0.28 0.6 ± 0.28 0.97 ± 0.21 0.36 ± 0.19 0.66 ± 0.16 1.08 ± 0.24 11.35 ± 0.95 13.19 ± 1.31 13.83 ± 0.3 11.87 ± 1.25 13.83 ± 0.53 13.71 ± 0.5 11.67 ± 1.56 13.23 ± 0.39 14.01 ± 0.67 11.06 ± 1.75 13.17 ± 0.66 13.2 ± 0.62 11.11 ± 2.78 13.56 ± 0.59 13.8 ± 0.76 9.89 ± 1.3 11.84 ± 1.69 13.53 ± 1.5 5.57 ± 1.23 9 ± 1.21 10.22 ± 1.67 5.8 ± 1.14 10.03 ± 1.68 9.2 ± 1.62 6.76 ± 1.02 8.31 ± 1.7 8.34 ± 1.92 11.18 ± 1.1 12.17 ± 0.42 12.67 ± 0.14 10.79 ± 1.13 12.12 ± 0.45 12.53 ± 0.33 10.87 ± 0.97 12.12 ± 0.46 12.65 ± 0.27 10.84 ± 1.05 12.28 ± 0.42 12.66 ± 0.33 11.49 ± 1.06 12.14 ± 0.35 12.5 ± 0.28 10.92 ± 3.22 12.13 ± 0.48 12.15 ± 1.46 8.87 ± 1.12 12.25 ± 1.39 11.91 ± 1.57 7.98 ± 2.16 12.94 ± 0.71 12.2 ± 1.22 8.04 ± 2.09 11.77 ± 2.12 12.75 ± 1.01 169 Table A4.2 (Cont’d) PRSE PRSE PRSE PRSE PRSE PRSE PRSE PRSE PRSE PRSE PRSE PRSE QUAL QUAL QUAL QUAL QUAL QUAL QUAL QUAL QUAL QUAL QUAL QUAL QUAL QUAL QUAL PRSE PRSE PRSE QUAL QUAL QUAL QURU QURU QURU StCon StCon StCon ACRU ACRU ACRU ACSA ACSA ACSA POGR POGR POGR PRSE PRSE PRSE QUAL QUAL QUAL Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High 62.2 ± 4.73 68 ± 9.87 76.58 ± 7.22 57.45 ± 4.89 56.77 ± 5.73 68.23 ± 3.09 56.75 ± 7.39 56.5 ± 6.27 64.2 ± 8.41 0 ± 0 0 ± 0 0.21 ± 0.58 39.33 ± 5.31 41.15 ± 4.79 36.92 ± 7.89 33.64 ± 8.63 41.47 ± 4.17 38.07 ± 6.66 40.07 ± 5.4 41.4 ± 7.46 43.73 ± 2.46 39.08 ± 2.5 41.71 ± 3.73 37.86 ± 4.62 40.57 ± 2.24 44.13 ± 6.32 49.87 ± 6.07 0.36 ± 0.13 0.39 ± 0.1 0.65 ± 0.11 0.49 ± 0.25 0.74 ± 0.19 0.9 ± 0.21 0.51 ± 0.1 0.62 ± 0.16 0.94 ± 0.36 0.27 ± 0.1 0.05 ± 0.09 -0.06 ± 0.1 4.16 ± 0.33 4.8 ± 0.42 5.51 ± 0.31 3.98 ± 0.37 5.09 ± 0.4 5.41 ± 0.45 4.16 ± 0.63 5.04 ± 0.27 5.42 ± 0.76 4.73 ± 0.4 4.69 ± 0.37 5.05 ± 0.56 4.62 ± 0.35 4.89 ± 0.38 5.32 ± 0.32 5.27 ± 0.08 9.9 ± 1.67 9.84 ± 0.92 5.71 ± 1.1 9.37 ± 1.28 9.3 ± 1.95 5.5 ± 1.1 9.57 ± 2.37 8.9 ± 1.81 5.66 ± 0.53 7.9 ± 1.64 8.71 ± 0.31 21.33 ± 2.53 20.09 ± 2.48 21.49 ± 1.48 21.8 ± 2.3 21.24 ± 2.56 21.84 ± 2.57 20.76 ± 1.38 20.26 ± 1.56 22.27 ± 1.86 19.7 ± 2.78 20.13 ± 1.97 21.91 ± 1.46 22.64 ± 2.46 22.8 ± 1.64 24.1 ± 2.49 7.03 ± 0.57 12.06 ± 1.87 13 ± 0.79 9.05 ± 2.13 12.1 ± 1.85 11.36 ± 1.53 7.86 ± 2.76 12.43 ± 1.28 12.17 ± 1.68 8.99 ± 2.25 11.4 ± 0.99 12.22 ± 0.7 18.29 ± 2.53 21.14 ± 2.74 20.63 ± 1.6 18.77 ± 2.49 20.13 ± 1.7 23.21 ± 1.46 19.15 ± 2.41 21.75 ± 2.63 20.58 ± 2.23 20.25 ± 2.11 22.43 ± 2.41 22.3 ± 3.81 17.81 ± 4.22 17.55 ± 2.4 21.29 ± 2.27 27.71 ± 16.85 27.43 ± 9.91 32.43 ± 13.96 35.86 ± 14.39 33.71 ± 17.52 34.86 ± 12.05 34.29 ± 9.76 22.14 ± 11.75 33 ± 11.22 30.14 ± 20.32 28.29 ± 13.41 37 ± 11.97 35 ± 15.28 35.29 ± 8.85 34.29 ± 7.48 170 Table A4.2 (Cont’d) QUAL QUAL QUAL QUAL QUAL QUAL QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU QURU StCon StCon StCon ACRU ACRU ACRU ACSA ACSA ACSA POGR POGR POGR PRSE PRSE PRSE QUAL QUAL QUAL QURU QURU QURU StCon StCon StCon Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High 39.57 ± 6.94 46 ± 3.14 43.57 ± 5.29 0 ± 0 0 ± 0 0 ± 0 37.77 ± 5.83 44.71 ± 6.28 43.29 ± 5.5 37.54 ± 3.71 40.46 ± 4.7 46.5 ± 3.57 40.07 ± 4.57 44.29 ± 4.94 45.27 ± 4.98 38.58 ± 4.89 42.69 ± 5.6 46.79 ± 4.82 38.07 ± 5.91 45.43 ± 5.21 50.53 ± 4.82 40.57 ± 6.22 44.67 ± 3.44 49 ± 3.95 0 ± 0 0 ± 0 0 ± 0 27.57 ± 10.97 30.86 ± 10.24 39.86 ± 6.74 3.71 ± 1.5 6.43 ± 1.51 10.29 ± 5.12 24 ± 14.11 31.57 ± 15.15 26.57 ± 19.46 28.57 ± 17.31 35.57 ± 19.12 34.14 ± 12.16 37.57 ± 15.86 29.43 ± 16.67 42.29 ± 18.39 26.86 ± 17.32 31.29 ± 9.18 32.14 ± 22.65 32.14 ± 24.29 22.29 ± 23.98 40 ± 20.12 23 ± 18.57 33.71 ± 14.35 35.43 ± 9.24 4.57 ± 3.87 6.14 ± 2.61 6.43 ± 2.44 171 4.52 ± 0.32 4.89 ± 0.42 5.08 ± 0.27 2.52 ± 0.37 3.2 ± 0.28 4.27 ± 0.12 3.22 ± 0.43 3.55 ± 0.35 3.38 ± 0.45 3.26 ± 0.64 3.55 ± 0.38 3.43 ± 0.48 3.05 ± 0.36 3.45 ± 0.28 3.87 ± 0.38 3.33 ± 0.48 3.65 ± 0.38 3.52 ± 0.68 3.25 ± 0.33 3.76 ± 0.22 3.66 ± 0.47 2.49 ± 0.56 3.25 ± 0.35 4.27 ± 1.1 2.78 ± 0.33 3.39 ± 0.25 3.71 ± 0.1 20.39 ± 1.42 19.8 ± 1.74 23.37 ± 1.64 19.33 ± 1.42 21.04 ± 1.42 22.53 ± 1.55 23.41 ± 3.59 25.13 ± 2.5 24.53 ± 2.84 23 ± 4.39 24.03 ± 2.37 26.91 ± 1.97 21.69 ± 3.42 23.36 ± 2.34 24 ± 1.46 23.06 ± 2.87 23.71 ± 1.76 26.29 ± 2.72 25.06 ± 5.22 26.3 ± 2.44 23.93 ± 3.68 25.73 ± 2.15 24.8 ± 2.26 25.41 ± 3.85 27 ± 1.63 24.4 ± 0.4 24.51 ± 1.42 19.53 ± 3.18 20.85 ± 1.78 21.76 ± 2.76 18.95 ± 1.68 17.99 ± 2.3 21.42 ± 1.88 14.15 ± 2.61 14.73 ± 3.67 15.41 ± 1.93 14.97 ± 3.51 14.16 ± 1.68 16.06 ± 3.5 14.38 ± 2.43 16 ± 1.78 15.42 ± 3.66 15.06 ± 1.84 16.19 ± 1.44 17.27 ± 2.04 15.79 ± 3.75 14.68 ± 1.42 17.64 ± 3.19 11.27 ± 1.46 14.82 ± 2.91 17.98 ± 1.49 10.97 ± 2.26 15.01 ± 0.85 17.7 ± 1.29 Table A4.3 Summary of predicted seedling survival (mean ± standard deviation). Data are provided for each species ´ soil source ´ light level. Species Soil Light ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACSA ACSA ACSA ACSA ACSA ACSA ACSA PRSE PRSE PRSE PRSE PRSE PRSE PRSE QUAL QUAL QUAL QUAL QUAL QUAL QUAL QURU QURU QURU QURU QURU QURU ACRU ACSA POGR PRSE QUAL QURU ST-CON ACRU ACSA POGR PRSE QUAL QURU ST-CON ACRU ACSA POGR PRSE QUAL QURU ST-CON ACRU ACSA POGR PRSE QUAL QURU ST-CON ACRU ACSA POGR PRSE QUAL QURU Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Lower CI Upper CI 0.3064 0.3991 0.5034 0.4702 0.6885 0.6738 0.728 0.5723 0.2195 0.4524 0.5012 0.5444 0.7043 0.7173 0.3024 0.2442 0.3637 0.175 0.4599 0.4922 0.755 0.4688 0.5706 0.7576 0.512 0.8362 0.7621 0.8811 0.5278 0.6023 0.6735 0.3815 0.7987 0.8681 0.6853 0.7636 0.8343 0.8148 0.9223 0.9338 0.9485 0.9544 0.8554 0.9361 0.9391 0.9524 0.9797 0.9813 0.7467 0.7295 0.8171 0.6547 0.8457 0.8706 0.9678 0.9049 0.9547 0.9843 0.9076 0.9952 0.9867 1.005 0.9432 0.956 0.977 0.8943 0.9947 0.9996 Predicted Survival 0.5031 0.5785 0.6797 0.6569 0.8138 0.8277 0.8524 0.8133 0.6054 0.7701 0.7722 0.798 0.8821 0.8836 0.5498 0.4924 0.6162 0.4066 0.677 0.711 0.8804 0.7333 0.819 0.9076 0.7501 0.9413 0.912 0.97 0.7816 0.8146 0.87 0.6901 0.9381 0.9679 172 Table A4.3 (Cont’d) QURU ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACSA ACSA ACSA ACSA ACSA ACSA ACSA PRSE PRSE PRSE PRSE PRSE PRSE PRSE QUAL QUAL QUAL QUAL QUAL QUAL QUAL QURU QURU QURU QURU QURU QURU QURU ACRU ACRU ST-CON ACRU ACSA POGR PRSE QUAL QURU ST-CON ACRU ACSA POGR PRSE QUAL QURU ST-CON ACRU ACSA POGR PRSE QUAL QURU ST-CON ACRU ACSA POGR PRSE QUAL QURU ST-CON ACRU ACSA POGR PRSE QUAL QURU ST-CON ACRU ACSA Low Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med Med High High 0.7742 0.4141 0.484 0.589 0.5714 0.741 0.7505 0.7816 0.6744 0.3997 0.5965 0.6144 0.644 0.7699 0.7808 0.4335 0.3739 0.5154 0.2514 0.556 0.5906 0.7996 0.6491 0.7567 0.8531 0.688 0.8936 0.8819 0.9433 0.6908 0.7646 0.8143 0.6228 0.8878 0.9366 0.883 0.4819 0.5597 0.9954 0.7643 0.8129 0.8703 0.854 0.9422 0.9478 0.9619 0.9671 0.9036 0.9497 0.9548 0.9631 0.9834 0.9821 0.8109 0.7849 0.8479 0.7412 0.8824 0.9053 0.9744 0.9678 0.9827 0.9946 0.9738 1.003 0.9969 1.01 0.973 0.98 0.9871 0.952 0.9965 1.009 1 0.8144 0.8621 0.9311 0.589 0.6562 0.7435 0.7246 0.854 0.8652 0.8844 0.8541 0.6765 0.8179 0.8195 0.8428 0.9102 0.9099 0.639 0.5873 0.6961 0.5071 0.7466 0.7732 0.9092 0.8637 0.9111 0.9558 0.8713 0.971 0.959 0.9855 0.8751 0.8972 0.9295 0.8188 0.9666 0.9837 0.9643 0.6741 0.731 173 Table A4.3 (Cont’d) ACRU ACRU ACRU ACRU ACRU ACSA ACSA ACSA ACSA ACSA ACSA ACSA PRSE PRSE PRSE PRSE PRSE PRSE PRSE QUAL QUAL QUAL QUAL QUAL QUAL QUAL QURU QURU QURU QURU QURU QURU QURU POGR PRSE QUAL QURU ST-CON ACRU ACSA POGR PRSE QUAL QURU ST-CON ACRU ACSA POGR PRSE QUAL QURU ST-CON ACRU ACSA POGR PRSE QUAL QURU ST-CON ACRU ACSA POGR PRSE QUAL QURU ST-CON High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High 0.6585 0.6258 0.7913 0.7898 0.8093 0.8782 0.7485 0.8426 0.8438 0.8415 0.9149 0.91 0.5046 0.4607 0.5746 0.3763 0.6157 0.6629 0.8377 0.6889 0.7775 0.8898 0.7131 0.908 0.8654 0.9416 0.8366 0.8597 0.8986 0.7608 0.9393 0.9645 0.9433 0.9128 0.8952 0.9588 0.9627 0.973 0.9954 0.977 0.9887 0.9901 0.9932 1.002 1 0.8564 0.8316 0.8816 0.7754 0.9171 0.9285 0.9825 0.9641 0.9826 0.9961 0.9697 1.004 0.997 1.008 0.9915 0.9942 0.999 0.9864 1.007 1.009 1.007 0.802 0.7864 0.8894 0.8983 0.9119 0.9558 0.8935 0.9454 0.9449 0.9515 0.9737 0.9735 0.7149 0.673 0.7637 0.6033 0.8026 0.8266 0.931 0.8668 0.9112 0.9568 0.8725 0.9712 0.9589 0.9862 0.9455 0.9543 0.9691 0.9171 0.9853 0.9928 0.9844 174 Table A5.1 Sample size (n) for each seedling species ´ soil source ´ light level in the greenhouse experiment. ACRU ACSA PRSE QUAL QURU Seedling species CON-ST Low Med High ACRU Low Med High ACSA Low Med High POGR Low Med High PRSE Low Med High QUAL Low Med High QURU Low Med High 14 13 15 8 9 10 8 10 12 10 12 12 10 11 13 12 14 14 13 14 14 15 16 15 12 13 13 13 15 14 14 15 16 11 14 14 15 15 16 15 15 15 15 15 16 12 14 15 13 14 15 14 14 16 11 13 14 15 15 16 15 16 15 14 14 16 13 13 15 10 10 15 12 13 15 13 12 15 12 14 15 14 15 15 14 14 15 8 10 11 7 9 11 10 10 12 5 9 10 11 11 12 11 13 12 175 Table A5.2 Sample size (n) for each seedling species ´ soil source ´ light level in the field experiment. † Indicates a sample size that below 5. Seedling species QUAL QURU CON-ST Low Med High ACRU Low Med High ACSA Low Med High POGR Low Med High PRSE Low Med High QUAL Low Med High QURU Low Med High 11 18 4† 8 16 5 9 21 3† 10 12 3† 12 22 3† 8 16 4† 8 13 6 13 18 5 10 32 2† 10 17 3† 12 28 3† 15 28 0 7 11 5 16 20 7 176 Table A5.3 Summary of seedling biomass in the field experiment. Data is presented for each species ´ soil source ´ light level. Soil Species QUAL ACRU QUAL ACRU QUAL ACRU QUAL ACSA QUAL ACSA QUAL ACSA QUAL CON-ST QUAL CON-ST QUAL CON-ST POGR QUAL POGR QUAL POGR QUAL PRSE QUAL PRSE QUAL QUAL PRSE QUAL QUAL QUAL QUAL QUAL QUAL QUAL QURU QUAL QURU QUAL QURU QURU ACRU QURU ACRU QURU ACRU QURU ACSA QURU ACSA QURU ACSA QURU CON-ST QURU CON-ST QURU CON-ST Light level High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med Biomass 3666.01 ± 12.51 1689.22 ± 231.1 3575.54 ± 33.04 1712.08 ± 146.55 4042.85 ± 184.55 5027.56 ± 115.16 1087.85 ± 437.61 3617.4 ± 269.16 3520.27 ± 123.15 1792.78 ± 124.88 4000.65 ± 253.16 1769.46 ± 146.5 4012.82 ± 263.91 2928.33 ± 55.41 1383.87 ± 116.58 2281 ± 131.29 3626.45 ± 413.01 1704.09 ± 133.76 3965.82 ± 210.26 3468.59 ± 126.44 990.08 ± 92.34 3467.45 ± 217.89 3639.76 ± 73.42 988.03 ± 162.09 3497.36 ± 184.74 3417.24 ± 62.63 1475.39 ± 552.04 2225.94 ± 319.93 177 Table A5.3 (Cont’d) QURU POGR High 3542.57 ± 126.24 QURU POGR Low 1051.93 ± 124.16 QURU POGR Med 3426.28 ± 258.21 QURU PRSE High 3691.07 ± 57.85 932.17 ± 98.61 QURU PRSE Low QURU PRSE Med 3501.95 ± 223 QURU QUAL High 3571.43 ± 148.14 QURU QUAL Low 1030.77 ± 107.57 QURU QUAL Med QURU QURU High 5399.9 ± 324.45 QURU QURU Low 1556.8 ± 276.23 QURU QURU Med 4008.48 ± 188.21 3439.6 ± 189 178 Table A5.4 Summary of seedling biomass in the greenhouse experiment. Data is presented for each species ´ soil source ´ light level. Soil Species ACRU ACRU ACRU ACRU ACRU ACRU ACRU ACSA ACRU ACSA ACRU ACSA ACRU CON-ST ACRU CON-ST ACRU CON-ST POGR ACRU POGR ACRU POGR ACRU PRSE ACRU PRSE ACRU ACRU PRSE ACRU QUAL ACRU QUAL ACRU QUAL ACRU QURU ACRU QURU ACRU QURU ACSA ACRU ACSA ACRU ACSA ACRU ACSA ACSA ACSA ACSA ACSA ACSA ACSA CON-ST ACSA CON-ST ACSA CON-ST POGR ACSA POGR ACSA POGR ACSA PRSE ACSA PRSE ACSA PRSE ACSA Biomass 435.72 ± 80.16 10.16 ± 1.66 140.22 ± 29.78 371.78 ± 120.84 14.66 ± 2.25 193.12 ± 49.32 414.16 ± 73.51 11.52 ± 2.27 160.34 ± 56.57 320.91 ± 77.12 13.42 ± 1.77 167.96 ± 54.84 373.73 ± 80.1 14.31 ± 1.84 156.03 ± 32.25 346.43 ± 58.94 13.58 ± 2.3 162.43 ± 53.13 359.05 ± 95.87 13.72 ± 1 175.69 ± 62.14 710.14 ± 203.54 80.72 ± 11.93 313.55 ± 76.56 883.42 ± 102.86 80.15 ± 8.16 383.22 ± 85.6 934.28 ± 119.87 74.21 ± 10.55 425.59 ± 52.74 594.84 ± 159.14 91.07 ± 9.46 270.06 ± 74.27 782.61 ± 133.41 76.41 ± 14.07 319.43 ± 96.5 Light level High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med High Low Med 179 Table A5.4 (Cont’d) ACSA QUAL High 646.41 ± 100.28 89.19 ± 14.5 ACSA QUAL Low ACSA QUAL Med 265.99 ± 69.72 ACSA QURU High 605.47 ± 127.97 89.22 ± 14.72 ACSA QURU Low ACSA QURU Med 274.01 ± 55 PRSE ACRU High 741.15 ± 184.93 74.14 ± 13.59 PRSE ACRU Low PRSE ACRU Med 301.04 ± 55.39 PRSE ACSA High 715.52 ± 213.05 78.11 ± 10.83 PRSE ACSA Low 278.63 ± 93.25 PRSE ACSA Med 841.51 ± 95.01 PRSE CON-ST High 81.09 ± 12.17 PRSE CON-ST Low 442.48 ± 55.8 PRSE CON-ST Med 554.95 ± 142.7 POGR High PRSE 80.48 ± 16.93 POGR Low PRSE 271.53 ± 82.5 POGR Med PRSE 808.4 ± 116.55 PRSE High PRSE 60.54 ± 7.66 PRSE PRSE Low 423.63 ± 77.77 PRSE Med PRSE 587.39 ± 86.63 PRSE QUAL High 88.4 ± 11.68 PRSE QUAL Low 291.2 ± 76.29 PRSE QUAL Med PRSE QURU High 641.48 ± 102.49 82.06 ± 13.79 PRSE QURU Low 259.74 ± 76.08 PRSE QURU Med QUAL ACRU High 3386.46 ± 229.5 QUAL ACRU Low 1055.42 ± 129.87 QUAL ACRU Med 2529.42 ± 297.94 QUAL ACSA High 3232.49 ± 200.79 QUAL ACSA Low 1060.3 ± 116.89 2352.63 ± 344 QUAL ACSA Med QUAL CON-ST High 3781.74 ± 229.71 QUAL CON-ST Low 1165.79 ± 221.2 QUAL CON-ST Med 2604.63 ± 265.12 POGR High 3448.65 ± 188.59 QUAL POGR Low 1124.04 ± 124.86 QUAL 180 Table A5.4 (Cont’d) POGR Med 2463.69 ± 364.43 QUAL PRSE High 3321.23 ± 217.76 QUAL PRSE Low 990.15 ± 158.15 QUAL QUAL PRSE Med 2327.46 ± 342.79 QUAL QUAL High 3883.71 ± 158.21 QUAL QUAL Low 1266.15 ± 227.99 QUAL QUAL Med 2615.77 ± 211.74 QUAL QURU High 3336.68 ± 157.96 QUAL QURU Low 1124.6 ± 93.69 QUAL QURU Med 2532.21 ± 254.17 QURU ACRU High 3360.74 ± 184.55 QURU ACRU Low 1046.38 ± 97.72 QURU ACRU Med 2500.16 ± 242.63 QURU ACSA High 3321.63 ± 155.43 QURU ACSA Low 1057.27 ± 149.69 QURU ACSA Med 2496.75 ± 276.45 QURU CON-ST High 3724.07 ± 162.16 QURU CON-ST Low 1109.21 ± 198.79 QURU CON-ST Med 2483.38 ± 223.65 POGR High 3393.5 ± 278.17 QURU POGR Low 1114.29 ± 102.92 QURU POGR Med 2319.26 ± 347.3 QURU PRSE High 3348.89 ± 156.41 QURU Low 1022.26 ± 131.25 PRSE QURU QURU PRSE Med 2329.37 ± 260.71 QURU QUAL High 3291.16 ± 250.07 QURU QUAL Low 1116.29 ± 101.97 QURU QUAL Med 2453.07 ± 283.18 QURU QURU High 3884.31 ± 239.72 QURU QURU Low 1275.73 ± 247.61 QURU QURU Med 2681.42 ± 260.84 181 Figure A5.1. Log response ratio ± standard error of seedling biomass in conspecific versus heterospecific soils at medium light availability in the field experiment. Values < 0 indicate negative PSF and values > 0 indicate positive PSF and values < 1 indicate negative PSF. Values that are statistically different from 0 (p < 0.05) are indicated with a star *. Figure A5.2. Log response ratio ± standard error of seedling biomass in conspecific versus heterospecific soils at high light availability in the field experiment. Values < 0 indicate negative PSF and values > 0 indicate positive PSF and values < 1 indicate negative PSF. Values that are statistically different from 0 (p < 0.05) are indicated with a star *. 182