At that point, the start inoculum was divided into 6 aliquots and stored in glycerol freezing buffer. For each inoculation in the first passage, an aliquot was thawed and cellspelleted for 10 mins at 4000 X G. Cells were re-suspended in 200mL 10mM MgCl2 buffer. Of this, 40mL were and heat killed in an autoclave for a 30 minutes at 121°C. Inoculum was plated, and an absence of growth confirmed that the heat-kill was effective. To get initial concentration of inoculum, dilution plating was performed on Kings Broth agar plates . Soil from each site, which had been stored at -20°C, was combined in a sterile Nalgene bucket and thoroughly mixed before inoculation.Soil inoculation: The top layer of every pot was supplemented with 40 grams of UC Davis Farm Soil. Soil inoculation was only performed once and only for the first passage of plants. Spray inoculation: Each plant was sprayed, using misting spray tops placed in 15mL conicals, with approximately 4.5mL of inocula. Control plants from passage 1 were inoculated with the heat-killed inocula. Control plants from subsequent experiments were inoculated with sterile 10mM MgCl2. Immediately after inoculation, plants were placed in a random order in a high humidity misting chamber for 24 hours. After 24 hours, the plants were moved to a greenhouse bench. Plants were inoculated once per week in the same manner and were placed in the misting chamber for 24 hours after every inoculation. Passage one plants received 5 weeks of inoculation, P2-P4: four weeks, and the fifth cohort: five weeks.Ten days after the final spray inoculation, plants were sampled.

With the exception for plant cohort 5, all plants were cut off at the base and immediately placed into sterile 1L bottles individually. By the end of cohort 5, dry marijuana the plants had grown too large to sample the entire plant, and instead, roughly 2/3 of the plant material was sampled from each plant, with care taken to sample the same age of branches from every plant. After collection, plant material was weighed, and 200mL of sterile 10mM MgCl2 were added to each bottle containing the plant material. The bottles were submerged in a sonicating water bath, sonicated for 5 minutes, vortexed, and sonicated for another 5 minutes. Half of the volume from each plant was pelleted for 10 mins at 4200 X G, re-suspended in ~1mL of 1:1 KB Broth Glycerol, divided into aliquots, and stored at -80°C for inoculation of the subsequent passage. The other half of the volume was pelleted in the same manner and then stored as a pellet at -20°C for DNA extractions. To prepare inoculation of the next passage, microbiome glycerol stocks were thawed, briefly pelleted to remove glycerol, and re-suspended in sterile 10mM MgCl2. Volume of re-suspension depended slightly on the size of the plants, but in general ranged from 5-10mL. Microbiomes were never pooled.Genotypes 2706, 3472, and 2934 were used for this experiment, and four plants of each genotype received each treatment . One control plant of each genotype wasspray inoculated with MgCl2 as a control. To prepare the inoculum, microbiomes from the end of passage one and the end of passage four were used. All aliquots were thawed and combined. The same was done for all of the individual microbiomes that came off of passage 4 plants.

To remove the glycerol, the samples were spun down and re-resuspended in 10mM MgCl2. In order to generate the 50/50 mix of P1 and P4 microbiomes, live/dead PCR with PMA treatment was used, adapted from the following method. Briefly, serial dilutions of P1 and P4 were performed in MgCl2. Each sample then received PMA at a final concentration of 100uM and vortexed. Samples were incubated in the dark at room temp for 5 minutes. Then they were placed in ice on a tray exactly 10cm away from a 700 watt halogen lamp. The light was turned on for 30 seconds, and turned off for 30 seconds. During the 30 seconds without light, the samples were all vortexed. This was repeated three more times. Samples were then pelleted for 10 minutes at 5000 X G. The supernatant including the excess PMA was removed, and cells were re-suspended in sterile 10mM MgCl2. Droplet Digital PCR was then utilized to quantify bacteria from each sample, and concentration was matched to 7.7 x 106 cells/mL. P1 and P4 were aliquotted separately and then recombined for the mixed inoculum so that each plant received ~9 x 104 bacteria each week that they were inoculated. Plants were inoculated for three weeks and harvested 10 days after the final inoculation as described previously.The neutral model was proposed by Sloan et al. to describe both microbial diversity and taxaabundance distribution of a community. Burns et al. have developed a R package based on Sloan’s neutral model to determine the potential importance of neutral process to a community assembly. In brief, the neutral model creates a potential neutral community by a single free parameter describing the migration rate, m, based on two sets of abundance profiles – a local community and meta communities. The local community describes the observed relative abundance of OTUs, while the meta community is estimated by the mean relative abundance across all local communities. The estimated migration rate is the probability of OTU dispersal from the meta community to replace a randomly lost individual in the local community.

The migration rate can be interpreted as dispersal limitation. In each microbiome passage, half of the samples were randomly selected and the relative abundance profile at the OTU level was used. The neutral model fit and migration rate were estimated in the resolution results of 200 iterations for P1, P2, P3, P4, and P4 Combined.We applied a null model approach on the serial passaging data P1-P4 to characterize the changes of stochastic process driving the assembly of plant microbiome over time. Lines that had high quality sequencing data at every time point , were used for this analysis. The null scenario for each line at each passage was generated using the data for that same line at the previous passage. The null scenario of P1 was generated using the original field inoculum sample. The null model approach was based on community pairwise dissimilarity proposed by Chase and Myers and extended by Stegen et al. to incorporate species abundance . Chase and Myers proposed a degree of species turnover by a randomization procedure where species probabilistically occur at each local community until observed local richness is reached. However, the estimated degree of turnover does not include species abundance. To take full advantage of our dataset, we also incorporated species relative abundance into the procedure proposed by Stegen et al. Zinger et al. has developed R code for the null model and applied the null model approach on the soil microbiome. This approach does not require a priori knowledge of the local community condition and determines if each plant microbiome at the current passage deviates from a null scenario generated by that same microbiome at the previous passage. In brief, the null scenario of each was generated by random resampling of OTUs and remained the same richness and number of reads with the original sample. Total OTUs observed in the sample and the corresponding relative abundance were used as probabilities of selecting an OTU and its associated number of reads, respectively. The BrayCurtis distance is used to calculate dissimilarities across null communities with 1,000 permutations. The null deviation shows the differences between average null expectation and the observed microbiome of the same line.Bacteriophage viruses that infect bacteria are both ubiquitous and abundant, so much so that they are estimated to largely outnumber bacterial cells in the environment. Their role in controlling bacterial populations has been studied since their discovery as “bacteria-killers” in the early 1900s, and there are countless studies investigating bacteria–phage pairwise dynamics. Even the fundamental Luria–Delbruck fluctuation experiment that demonstrated mutations arise in the absence of selection was conducted using phage as the selective pressure. In addition to fundamental research, marijuana grow system lytic phages have been widely studied for their use as biological control and therapeutic agents, and have been successfully used in control of plant pathogens, including in tobacco, tomato, detached flowers of Rosaceae trees, and even some ready-to-eat foods such as hot dogs and lettuce leaves. Beyond just controlling bacterial abundance, phage predation in the ocean is thought to impact bio-geochemical cycling and food web processes through bacterial lysis, converting biomass into dis- solved organic matter and contributing to the dissolved organic carbon pool. Recently, increased interest in the human microbiome is also beginning to acknowledge a key role of lytic phages in these communities, but in contrast to free-living microbiota, little empirical work has been done to uncover the role of phages in these systems. Specifically lacking is an understanding of when and how phages shape the abundance and composition of host-associated bacteria. Phages can have important impacts on the competitive dynamics and structure of bacterial communities, and they are predicted to maintain bacterial diversity through a variety of mechanisms. As bacteria evolve to escape phage predation and phages counter-adapt to these resistances, evolutionary and co-evolutionary dynamics can drive important phenotypic and genotypic variation within the bacterial community . One idea that is frequently put forth in marine systems, known as “kill-the- winner” dynamics, suggests the most abundant bacteria should also be the most susceptible to phage predation. In this model, an increase in bacterial abundance is followed by an increase in the associated phage population and a subsequent decrease in bacterial abundance, effectively preventing one type of bacteria from ever dominating the community. Antagonistic co-evolution between bacteria and phage has recently been put forth as a driver of bacterial diversification and variation within the human gut microbiome, with potential impacts on microbiome function and human health through phage-mediated homeostasis or dysbiosis.

The impact of lytic phages on bacterial cell density and community diversity may in part be the result of cell lysis, which not only has a direct effect on the population of cells, but also has an indirect effect on competition among bacterial strains and species within a community. Phages may also increase bacterial density and diversity by releasing nutrients into the environment via lysis byconferring metabolic or morphological traits to bacteria upon integration into the genome . Furthermore, culturing phages can be difficult, as it is reliant on having an isolate of a suitable bacterial host. Despite these difficulties, studying phages in host-associated microbiomes may reveal interactions between bacteria and phages that are different from those occurring in free-living microbial populations. For example, recent work has suggested that microbiomes may be dominated by temperate phages that integrate into the host genome as opposed to lytic phages that lyse their host cells . In this case, changes in bacterial density or composition due to lysis would likely represent only one small way in which phages impact their microbiome. Furthermore, host factors such as age, immunity and health are likely to change the dynamics between bacteria and phages within the host environment . Another key difference in host-associated microbiomes compared to free- living bacterial communities is the process of colonization of a new host. Phages may play a key role in early microbiome establishment, the importance of which will be specific to the mode of microbiome transmission and the diversity and composition of colonizing bacteria. In this work, we sought to investigate the role of lytic phages in shaping bacterial abundance and community composition during colonization of the phyllosphere . We used a filtration method to deplete phages from the microbial community associated with tomato leaves, modified from an approach that has previously proven effective for separation of phage and bacteria in seawater. Through inoculation of a size fractionated field-grown tomato microbiome onto juvenile, growth chamber-grown plants, we are able to test whether the lytic phage fraction of the phyllosphere has an impact on bacterial abundance, composition and diversity during microbiome establishment. We find an impact of phages on overall bacterial abundance and relative abundance of specific taxa that is measurable after 24 hr, but not at 7 days post-inoculation. We also find evidence for slightly higher alpha diversity after 7 days in those communities in which phages were initially present in the inoculum relative to those in which they were depleted.