From the original inoculum sample, we identified ten Operational Taxonomic Units using linear discriminant analysis effect-size analysis that were significantly associated with particular genotypes in P1 and P2. We compared their presence/absence at the end of P4 to those OTUs that were not found to be associated with genotype. Interestingly, those OTUs that were significantly associated with particular genotypes at the start of the experiment were significantly more likely to be present at the end of the experiment than those not associated with genotype . In addition to genotype effects, we were interested in what other factors were driving our observed change in community composition. Bray-Curtis distances across all samples uncovered a significant effect of both passage number and sample type on bacterial communities . As this was an open system, we next sought to determine if there was a high degree of dispersal amongst plants within the greenhouse by directly comparing the communities of experimental and control plants. At every passage, cannabis drying rack control and experimental plants are found to host significantly different communities , suggesting minimal effects of dispersal within the greenhouse relative to our inoculations.
When inoculum and control samples are removed from analysis, there remains a significant effect of passage number and a significant overall effect of plant genotype on community composition . When variance is partitioned, passage can explain 51% of dissimilarity, whereas genotype explains only 4%. Replicate lines from accession 2934 were lost after P3 due to a stem rot fungal pathogen present in the original inoculum that seemingly only infected this genotype. However, the observed overall genotype effect was not driven by this accession, as there remains a significant effect of genotype after its removal , and passage number remains highly significant . To better understand how the original, diverse, field inoculum changed over four passages on plants in the greenhouse, we calculated the percentage of OTUs in the original inoculum that were detectable over the course of the experiment . At the end of P1, 92% of the field inoculum OTUs were still present on the plants, but by P4, this was reduced to 29%. We then calculated if the decrease in original community member diversity was the result of replacement by non-inoculum taxa . In this case, we observed that the proportion of sequencing reads representing the original inoculum OTUs remains above 78% . This suggests that a relatively small percentage of the community was made up of OTUs that colonized plants from the greenhouse environment.
Of note, some OTUs considered “non-inoculum” were likely present in the initial inoculum, but in too low of abundance to detect. In particular, there were 27 OTUs with reads in the spray inoculum sample in the non-rarefied dataset, but this was number was reduced to zero after data rarefaction. To account for the impact of the small percentage of arriving species on community composition, we re-analyzed the dataset using only those OTUs that were observed to be present in the initial inoculum . In this case, passage number remains a significant driver of community dissimilarity , as does genotype . We next measured changes in community diversity over the course of passaging and across lines. We found a significant decrease in both OTU richness and alpha diversity over time across all plant genotypes , including when only original spray inoculum OTUs are considered . Importantly, this drop in diversity from the start of the experiment does not correspond to a decrease in overall bacterial abundance on plants . Note that our measures of bacterial growth likely largely overestimate the starting densities and do not account for population turnover , and are therefore highly conservative. In P1, we also estimated fold change of bacterial abundance on control plants that were sprayed with heatkilled inoculum, and found an average change of 0.76, which is significantly lower than the averaged 11-fold change for experimental plants which received live inoculum .
Finally, although passaging was performed in a control temperature greenhouse, outside high and low temperatures and humidity all varied significantly across passages , which may have impacted the observed differences in both abundance and growth across passages. With the knowledge that communities were drastically changing over time, we sought to determine if the rate at which the communities were changing was consistent. To do this, we calculated Bray-Curtis distances of microbiomes in each passage to P1 microbiomes . As we similarly observed through ordination plots in Figure 1, the communities become less similar to P1 over time. We then fit both a linear and quadratic regression to these data, and we found a better fit of a quadratic model than linear. We determined this through comparing both R2 values and calculating Akaike information criterion values . Both models were highly significant . Taken together, this suggests that community change may be slowing down, although it appears to have not entirely stopped. We next observed changes in relative abundance of specific taxa within lines over time . At each passage, there are numerous taxa that are differentially abundant compared to other passages. In some cases, there was evidence for replacement of OTUs within taxonomic groups. Specifically, in the top 10 most differentially abundant taxa as determined by using a Kruskal-Wallis test, three of them are in the family Pseudomonadaceae. OTUs 0010 and 0004 are in significantly higher relative abundance in P1 than in P4 , and gradually decreased in relative abundance, whereas OTU0002, an unclassified Pseudomonadaceae, is significantly more abundant in P4 as compared to other passages . All three OTUs are present in the initial spray inoculum, although OTU0002 represents only 0.03% of rarified spray inoculum reads whereas OTU0004 represents 27% and OTU0010 represents 21%. To better understand how bacterial community dynamics were changing over the course of the four passages, we utilized a recently developed cohesion metric to quantify connectivity of microbial community. In brief, community cohesion is a computational method used to predict within-microbiome dynamics by quantifying connectivity of microbial communities based on pairwise correlations and relative abundance of taxa. Changes in community cohesion over time are suggestive of biotic interactions, where connectivity can arise from either, or both, positive and negative interactions resulting from cross-feeding or competition as well as environmental co-filtering. When applied to our dataset , we find a mild but significant increase in positive cohesion values from P1 to P4 . Consistent with positive cohesion values showing increased biotic interactions, there are also increasingly negative cohesion values from P1 to P4, which again is mild but significant . To test if bacterial communities were changing due to neutral processes alone, we first applied the Sloan neutral community model and found that a neutral model is less correlated with observed communities on the plants over time . However our data violate a key assumption of the neutral model in that dispersal was experimentally constrained within lines, and thus we took the approach of generating a null prediction based on the known community composition of inocula applied at each passage and comparing our observed communities to the predicted neutral community using a recently developed approached. We found that Bray Curtis distances between predicted and observed communities moderately increases over time , suggesting that community change over the course of the passaging experimentis likely the result of deterministic rather than neutral processes. Further evidence for a shift away from neutrality can be observed using occupancy- abundance curves in which the occupancy, or proportion of individuals in which an OTU is found, is plotted against its relative abundance. A positive correlation between the two is expected to occur by chance, curing marijuana as in a neutrally assembled community, but a change in distribution of individuals may indicate a community shaped by deterministic processes. When our data are visualized in this manner , we see that in P1, the most abundant taxa also occupy the highest proportion of plants, as you would expect in a neutral community not undergoing niche selection.
However, this trend collapses by P4 with many abundant taxa occupying far fewer individuals than would be expected under an assumption of neutrality. We next designed an experiment to which we could apply Sloan’s model of neutral theory . All lines from the end of P4 were pooled together and re-inoculated onto tomato plants, mimicking the inoculation procedure from the first passage. Plants that received the P4-pooled inoculum had significantly different bacterial community composition than the P4 plants themselves . We did not observe an effect of genotype on the communities assembled from this combined inoculum . We also found that the majority of the variation between samples was driven by an exceptional situation of introduction of a greenhouse taxon to the plants . To test if neutral processes were driving community structure in this experiment, we again examined fit to a neutral model using the Sloan model approach. In this case, as with P1, the assumption of equal dispersal potential among plants is met. In 200 iterative predictions, the fit of the neutral model is significantly higher in P1 than P4 Combined , suggesting that neutral processes are dictating the community structure after the first passage, but not in the P4 Combined experiment . We also see the occupancy-abundance relationship breakdown in P4-Combined when compared to P1 directly .The similarity of changes in community structure both across replicates and genotypes over the course of the passaging experiment led us to predict that these microbiomes were adapting to the local plant and greenhouse environment. To further determine if the community changes we observed from P1 to P4 were due to habitat selection rather than neutral processes, we employed a community coalescence competition experiment. In this experiment , phyllosphere communities from the end of P1 and the end of P4 were inoculated onto a new cohort of plants, either on their own or in an approximately 50:50 mixture of live cells . To ensure that our method for the mixed inoculum was effective, we sequenced multiple replicates of the P1, P4, and Mix inoculums and found that source of inoculum explains 88% of dissimilarity amongst samples . To ensure that the Mix inoculum was significantly different than both P1 and P4 separately, we compared P1 and Mix inocula directly and found that 75% of difference between samples can be explained by this variable . Similarly, when P4 and Mix are compared directly, 74% of variation in the community is explained . This consistent difference among the three inocula allowed us to compare the communities colonizing plants from each treatment. We first measured final bacterial abundance and found that colonization was lower on these plants than in previous experiments, but does not significantly differ among treatments , apart from control plants, where bacterial colonization was greatly reduced . We then compared bacterial communities again using 16S amplicon sequencing and ordinated samples on a PCoA based on Bray-Curtis distances. Plants that received P1 inoculum had distinctly different communities than those that received either P4 or the Mixed inoculum. Plants that received the Mixed inoculum clustered together with those receiving P4 and were relatively indistinguishable. Using ADONIS tests, we determined that inoculum source can explain 45% of Bray-Curtis dissimilarity amongst samples , and there was no effect of plant genotype . In a pairwise analysis between P1 and Mixed, inoculum source explains 31% of the community dissimilarity . In contrast, inoculum source does not explain any significant variation in dissimilarity amongst P4 and Mixed inoculum plants . Together, these results suggest that the plants receiving the 50:50 mixed inoculum were indistinguishable in community composition from those receiving the pooled, P4 adapted microbiomes, and that these selected communities were not invadable by the microbial communities from the start of the experiment. Consistent with our results from the passaging experiment itself, alpha diversity was highest in P1 plants compared to both P4 and Mixed plants . Alpha diversity did not differ amongst communities colonizing plants from the P4 and Mixed inoculums, despite being different between the two inocula themselves. We also examined compositional makeup of the communities , and consistent with P1 to P4 passaging results, we see differentially abundant taxa between groups . Again, two Pseudomonas OTUs are more abundant in P1 plants as compared to P4 and Mix, in which there was an unclassified Pseudomonaceae that was higher in relative abundance.