To determine if the effects of PhylloStart bacteria on plant reproductive success would be seen in an environment with more potential sources of phyllosphere bacteria and/or whether early inoculation of plants changed subsequent microbiome assembly in the field, we included a field component in the third trial experiment. After inoculation, we transferred both PhylloStart-inoculated and control plants into the field. These plants were sampled concurrently with the plants from the same cohort that remained in the greenhouse , and their phyllosphere communities were sequenced. We analyzed greenhouse and field locations separately and found a significant effect of PhylloStart inoculation density on bacterial abundances in the greenhouse =10.17, p=0.006, but no effect in the field =4.07, p=0.13. A Dunn Post-Hoc test showed that PhylloStart High treated plants had significantly higher bacterial abundance than the control plants in the greenhouse . Further, while we see that community composition is influenced by PhylloStart treatment in the greenhouse, we see no such effect in the field grown plants. When looking at a PCoA of bacterial community similarity using Bray-Curtis distance we see that the PhylloStart-treated greenhouse plants clearly separate out from the control plants, with the plants treated with high concentrations of PhylloStart distinct from the controls, and the plants treated with low concentrations of PhylloStart falling between the controls and the high inoculation. Meanwhile, the communities associated with the field grown plants differ from those grown in the greenhouse, plant pot with drainage but do not otherwise separate by PhylloStart treatment. When analyzing dissimilarity using an Adonis PERMANOVA, we see a significant effect of PhylloStart treatment , location; ie. field vs. greenhouse , as well as a significant PhylloStart by location interaction .
When analyzing each location separately with a pairwise Adonis, we see that there are significant differences between PhylloStart high and control treated plants in the greenhouse , but no difference between PhylloStart high and low , or PhylloStart low and control in the greenhouse. The plant microbiome is increasingly recognized for its role in shaping plant health, but most work to date has focused on below ground associations between plant-microbe interactions. Thus far,most evidence for an impact of the above-ground, phyllosphere microbiome on their hosts has focused on disease or herbivore resistance . Our initial experiment established that greenhouse-grown plants develop a significantly more abundant microbial community when inoculated with a synthetic microbial community . When inoculated onto plants early in development, we found that our taxa represent the dominant members of the phyllosphere and that overall bacterial densities were far higher in amended plants relative to controls . With two additional greenhouse studies, we determined that these microbial associations lead to a significant increase in the total number of fruit produced by greenhouse-grown tomato plants . We also found that, in a growth chamber trial, the bacterial community provided nutrient status dependent protection from P. syringae establishment. In contrast, plants that were transplanted into a field environment did not appear to benefit from the initial inoculation of PhylloStart bacteria . Given the minimal development of the phyllosphere community in non-treated greenhouse control plants , our findings suggest that greenhouses present an ideal location to study the effect of microbial amendments on agriculturally relevant plant traits, and support previous work finding that greenhouse-grown plants develop bacterial communities distinct from outdoor environments .
The extraordinarily low background levels of bacterial colonization allowed us to examine the importance of phyllosphere bacteria to plant fitness, where we found that the application of PhylloStart bacteria was associated with increased total fruit production . Further, that we do not see a fruit number/size tradeoff in this study suggests that the microbial amendment is increasing investment in above-ground biomass, rather than simply redirecting resources from fruit size to number, as is commonly observed in seeds for example . Our results add to a body of work describing how fruit yield can be affected by both local nutrient conditions and microbial associations but extend the latter to the above ground tissues. In contrast to the greenhouse, we did not see evidence for either establishment or impact of PhylloStart amendment in the field. In this case, PhylloStart bacteria were not found at significant abundances on these plants after a month in the field , and their initial community structure did not seem to shape the future composition of the phyllosphere communities . While this would seem to contradict results finding initial colonizers dominate plant microbiome assembly , priority effects often depend on the identities of the earlycolonizing species and their environments. For example, when wood disks were pre-colonized with fungi and placed in a field for six months, retention of initial colonizers in the eventual community varied between ascomycetes and basidiomycetes, and from season to season . In the tomato plant phyllosphere, PhylloStart bacteria may have been overwhelmed by dispersal from neighboring plants , or from other sources. Further, the field conditions during which we ran our trials may have differed from those under which we initially quantified the phyllosphere composition to design PhylloStart. Previous work has shown that community composition will depend on both plant host genotype as well as local environmental conditions .
It remains possible that ‘local’ or otherwise well-adapted isolates may have yielded better performance. Further field trials under a broad spectrum of conditions and locations, as well as with larger sample sizes, are needed to determine whether bacterial amendments to the phyllosphere can potentially confer benefits to commercial field tomato production. There are various mechanisms by which the phyllosphere community might provide the observed benefits to its host. These include: 1) altering plant hormone signaling, either directly by producing phytohormones or indirectly through the elicitation of a plant response; 2) by increasing the nutrients available to the plant either through enhanced nutrient fixation or availability; and 3) through reducing stress, either environmental or due to pathogen pressure . While our study does not seek to explain the mechanism underlying observed biostimulant effects, it likely relies on a combination of these factors. However, that the effects of Azomite fertilization and PhylloStart inoculation acted primarily in an additive fashion suggests that altered nutrient acquisition is not a particularly dominant force. Like many phyllosphere microbiome studies, our experimental design did not specifically exclude the possible movement of bacteria to the soil , and it thus remains possible that some fraction of the inoculated bacteria colonized the below-ground compartment. However, recent studies have found that phyllosphereand rhizosphere-associated bacteria are predominately adapted for survival in their respective niches , and, when paired with our 16S rRNA sequencing results showing robust and long-term establishment of the community on the leaf surfaces, we think it is more likely that the effect is mediated through phyllosphere interactions directly. With this in mind, future work exploring the mechanisms of phyllosphere associated growth promotion should specifically differentiate any effects impacting above versus below ground responses to the PhylloStart bacteria, either through reciprocal translocation of the species or by physically preventing inoculation of or migration to the soil. One specific potential mechanism for the increased reproductive success is linked to the phytohormone auxin , which is a major regulator of plant growth, is commonly produced by bacteria inhabiting the phyllosphere , and has been linked to increased biomass accumulation in rice and corn . In this context, increased fruit yield could be mediated by the action of auxin in decreasing flower abscission , potentially leaving more flowers available to set. Of note, we did not observe a significant change in flower number across the first trial. Using BLAST to search the genomes of the PhylloStart bacteria, we found that several members have matches for idpC , a key protein in auxin production . Future research is needed to confirm that these bacteria can produce auxin in planta, growing blueberries in pots and if this may explain some of their plant-beneficial effects. It is also possible that the PhylloStart bacteria alter the plant’s response to environmental cues, allowing the plant to better optimize its growth strategy and invest more resources in reproduction.
Recent work has focused on the phenomenon of microbiome-dependent ontogenic timing , by which the presence of certain bacterial species acts as essential cues in the developmental timing of their host organism . For example, the composition of the Boechera stricta soil-associated bacterial community has been found to significantly alter the timing and duration of flowering . Further research is needed to assess the potential role that host-associated microbes play in developmental timing. Throughout these trials we saw no evidence of an interaction between the nutrient status of the plant and the effect of the PhylloStart bacteria; instead, the bacterial and Azomite treatments additively increased the total yield. Given these observations, we were curious if the PhylloStart community would show the nutrient dependent pathogen protection found in our lab’s previous work . Indeed, we found, in a growth chamber experiment, that the addition of this community limited the growth of the pathogen P. syringae in nutrient-limited plants, but that this effect disappeared when organic phosphorus fertilizer was added . These results are in line with the stress gradient hypothesis, which posits that inter-species interactions should become more facilitative under adverse conditions , and highlight the important role that phyllosphere bacteria play in stress response. Indeed, previous work in this system has shown that the PhylloStart bacteria up-regulate defense responses in Arabidopsis and subsequently reducing pathogen colonization, . In summary, we find that the presence of phyllosphere-associated bacteria benefit their plant host when grown in a microbially depauperate greenhouse environment, through an increase in reproductive success as measured by total fruit production, with further evidence for pathogen resistance. These results are important for understanding the role of microbial communities in host outcomes and are broadly relevant in an agricultural context where, for example, 32% of domestic and 56% of imported tomatoes in the United States are grown in greenhouses that may not provide adequate colonization of phyllosphere bacteria . Further, we show that bacterial inoculation provides an additive increase in fruit production when applied with a common supplement containing micronutrients, opening avenues for further optimization of agricultural production by harnessing the biostimulant properties of phyllosphere microbes.The detachment of a grape berry from its pedicel generally damages the berry because the vascular tissues and associated parenchyma, collectively known as “the brush”, remain attached to the pedicel and are pulled out of the berry on detachment, leaving an open wound sometimes called a “wet” stem scar on the berry’s stem-end. Berry detachment may also remove pieces of skin or cause the whole berry to rupture. Such mechanical damage can reduce the yield and quality of machine-harvested grapes for wine or raisins. Stem-end picking damage also limits the quality and storage life of stemless table grapes. Certain plant growth regulators known as “abscission agents” activate an abscission zone at the pedicel-fruit boundary. The activation of this abscission zone reduces fruit detachment force and promotes the development of dry stem scars . Abscission agents could reduce picking damage and thereby serve as harvest aides if treated grapes can be harvested after the abscission zone is activated, but before the fruit abscises. However, once the abscission zone has been activated, development proceeds quickly and may lead to excessive preharvest fruit drop .The first compound tested as an abscission agent for grape was ethephon, a phosphonic-acid compound that decomposes to release the gaseous plant hormone ethylene. Ethephon can induce the abscission of mature grape berries within 7 to 14 days after treatment, but high dosages are needed. The use of ethephon as an abscission agent for grapes would require an application dosage which is higher, and a preharvest interval that is shorter, than those for the current registered use of ethephon on grapes in order to enhance berry color. Such changes could be expected to increase ethephon residues on treated fruit, and it seems unlikely that regulatory agencies would approve a use that could increase ethephon residues on grapes since existing residues are already a concern. However, it should be noted that Ferrara et al. found that effective dosages of ethephon did not result in excessive residues. Jasmonates, including methyl jasmonate, a natural product, have also been shown to induce the abscission of various fleshy fruits, including blueberry , orange , and tomato.