Edited by: Philipp Zerbe, University of California, Davis, United States
Reviewed by: Trent Russell Northen, Lawrence Berkeley National Laboratory (LBNL), United States; Sibongile Mafu, University of Massachusetts Amherst, United States
This article was submitted to Plant Metabolism and Chemodiversity, a section of the journal Frontiers in Plant Science
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Previous reports regarding rhizodeposits from apple roots are limited, and complicated by microbes, which readily colonize root systems and contribute to modify rhizodeposit metabolite composition. This study delineates methods for collection of apple rhizodeposits under axenic conditions, indicates rootstock genotype-specific differences and validates the contributions of vegetative activity to rhizodeposit quantity. Primary and phenolic rhizodeposit metabolites collected from two apple rootstock genotypes, G935 and M26, were delineated 2 months after root initiation by utilizing gas chromatography/liquid chromatography—mass spectrometry (GC/LC-MS), respectively. Twenty-one identified phenolic compounds and 29 sugars, organic acids, and amino acids, as well as compounds tentatively identified as triterpenoids were present in the rhizodeposits. When adjusted for whole plant mass, hexose, erythrose, galactose, phloridzin, kaempferol-3-glucoside, as well as glycerol, and glyceric acid differed between the genotypes. Phloridzin, phloretin, epicatechin, 4-hydroxybenzoic acid, and chlorogenic acid were among the phenolic compounds found in higher relative concentration in rhizodeposits, as assessed by LC-MS. Among primary metabolites assessed by GC-MS, amino acids, organic acids, and sugar alcohols found in relatively higher concentration in the rhizodeposits included L-asparagine, L-cysteine, malic acid, succinic acid, and sorbitol. In addition, putative ursane triterprenoids, identified based on accurate mass comparison to previously reported triterpenoids from apple peel, were present in rhizodeposits in high abundance relative to phenolic compounds assessed via the same extraction/instrumental method. Validation of metabolite production to tree vegetative activity was conducted using a separate set of micropropagated trees (genotype MM106) which were treated with a toxic volatile compound (butyrolactone) to inhibit activity/kill leaves and vegetative growth. This treatment resulted in a reduction of total collected rhizodeposits relative to an untreated control, indicating active vegetative growth contributes to rhizodeposit metabolites. Culture-based assays indicated an absence of bacterial or fungal endophytes in roots of micropropagated G935 and M26 plants. However, the use of fungi-specific primers in qPCR indicated the presence of fungal DNA in 30% of the samples, thus the contribution of endophytes to rhizodeposits cannot be fully eliminated. This study provides fundamental information for continued research and application of rhizosphere ecology driven by apple rootstock genotype specific rhizodeposition.
Rhizodeposits, which is the term used to encompass all plant metabolites entering the rhizosphere and originating from roots, are an important determinant of the rhizosphere microbiome composition and function which contributes to plant protection from pathogens and nutrient uptake (Berendsen et al.,
Previous work demonstrated that rhizodeposits can differ among apple rootstock genotypes (Leisso et al.,
The importance of differentiating apple rhizodeposits stems from research indicating genotype-specific apple rhizosphere microbiomes (Mazzola and Manici,
The objective of this study was to determine the influence of rootstock genotype on qualitative and quantitative attributes of apple rhizodeposits. Rhizodeposits from axenically reared apple rootstocks were assessed for phenolic compounds, organic acids, amino acids, sugar alcohols, sugars, and triterpenoids and the relative differences between the two apple rootstock genotypes was determined.
The first experiment assessed overall method viability and compared rhizodeposit metabolic profiles of 5 replicates each of M26 and G935; only phenolic compounds were assessed in this experiment. One replicate was removed from each genotype from the first experiment, due to microbial contamination. In the second experiment, experimentation was set up for 9 replicates for each of M26 and G935; two replicates were removed from each genotype due to microbial contamination in two replicates of the M26 rhizodeposit collection. Phenolic compounds, certain triterpenoids, as well as sugars, amino acids, organic acids, and sugar alcohols were assessed in this second experiment. A validation experiment (experiment three), assessed root tissue from micropropagated trees for endophytes with 5 separate G935 and M26 plantlets from the same initiation lines, due to the need to retain whole plantlets from second experiment for metabolic dry weight correction. A further validation experiment (experiment four) to ascertain the contribution of live plant metabolic activity to rhizodeposits (as opposed to sloughed off callus and root cells) was performed with twelve MM106 plantlets due to limited availability of G935 and M26. Micropropagation and rhizodeposit collection were similar for all experiments as described in the following sections.
Micropropagated rootstock trees (M26, G935, and MM106) were grown axenically with methods similar to Dobranszki and da Silva (
Trees were removed from root elongation media, and roots were rinsed in sterile distilled water to remove loose callus tissue. The tree was then placed in a 2″ circular neoprene float (Ehydroponics.com) which had been disinfested by soaking in 95% ethanol for 24 h, and dried on a sterile Petri plate in a laminar flow hood. The tree was then transferred to a sterile Magenta box containing 85 mL sterile distilled water to allow remaining callus tissue to slough off. After 24 h, the tree was transferred again to a new Magenta box containing 85 mL sterile distilled water, where it remained for 4 additional days (Figure
Example experimental unit: rooted axenically micropropagated rootstock on sterilized neoprene float. Trees remained in water for 5 days after which water was filtered, flash frozen, and subsequently processed for metabolite analysis.
Rhizodeposit samples were processed using previously described methods (Leisso et al.,
For the first and fourth experiment, samples were analyzed for phenolic compounds using a method previously described in Leisso et al. (
Additionally, samples in the fourth experiment (
Methoxyanimation/trimethysilylation for analysis of sugars, organic acids, amino acids, and sugar alcohols was performed essentially as described by Rudell et al. (
Samples were analyzed a second time with a 20:1 split injection in order to better resolve the major sugars.
Agilent data files were converted to mzdata files from MassHunter software. Mzdata files were analyzed in MZmine (v 2.27) (Pluskal et al.,
Derivitized sample datafiles from GC-MS analysis were processed in Chemstation software (Agilent Inc., Palo Alto, CA, USA). Separate analysis methods were applied to split vs. splitless injections, with split injections focusing on peaks which did not have good separation in the splitless injection due to high abundance, specifically malic acid, fructose, glucose, sorbitol, and myo-inositol. The library from the splitless injection included amino acids, ursane triterpenoids, organic acids, sugar alcohols, and other sugar species; libraries employed were previously reported in Leisso et al. (
Fructose, glucose, myo-inositol, and L-alanine were found in high concentrations in extraction of fresh REM media (in which plantlets had been grown; the concern was that small amounts could remain on the roots, despite rinsing procedures) and were excluded from further analyses.
Roots of micropropagated G935 and M26 cultivars were assessed to determine the presence of endophytes in the axenically micropropagated trees which could influence metabolic profiling results. As trees were initiated in micropropagation media from actively growing foliar shoot tips, endophytes present in the roots would have had to originate from foliar tissue.
Trees were in root elongation media longer than the usual rooting window (3 months) in order to further allow any endophyte or contaminant growth.
Tissue collection was performed in the laminar hood. Trees were extracted from root elongation media, and any remaining media was removed from the roots with flamed-sterilized forceps. Roots and shoots of trees (5 per genotype M26 and G935) were axenically divided and placed in sterile specimen cups and flash frozen by placing the closed container in liquid nitrogen. Root tissue was chopped finely and DNA was extracted from a 50 mg sample using the PowerPlant Pro DNA isolation kit (MoBio Laboratories Inc.,) and a final elution volume of 50 μl. DNA was extracted from roots of a micropropagated tree exhibiting fungal contamination and used as a positive control in PCR reactions detailed below.
Total fungal and bacterial populations in root tissue were evaluated using a real-time qPCR procedure using the primers and amplification conditions as previously described (Reardon et al.,
The presence of culturable fungi or bacteria in roots of asymptomatic (no hyphae or bacterial colonies obvious in the media) micropropagated trees of cultivar G935 and M26 was assessed using full strength PDA (24 g L−1 potato dextrose broth + 15 g L−1 agar) and full strength TSA (30 g L−1 tryptic soy broth + 15 g L−1 agar), respectively. Five root segments from different roots were plated to TSA and PDA; the main stem was cut into 5 segments; each segment was streaked on TSA and then gently placed in the surface layer of the PDA media.
To assess the potential contribution of sloughed off cells to the rhizodeposit metabolome, a follow-up validation experiment was performed, where plants were treated with a toxic volatile compound to slow growth and destroy functionality of vegetative tissue. The study employed MM106 rootstock due to insufficient quantities of G935 or M26 plants. Twelve micropropagated explants (MM106) in water were rinsed and transferred from media to neoprene floats as describe above; after a day of soaking to remove callus tissue, explants were transferred to fresh sterile distilled water, and water collected for rhizodeposit analysis after an additional day. Following the initial collection, explants were transferred to fresh collection boxes, and for half (6) of the trees, a sterile filter disc treated with 10 μl butyrolactone (Sigma-Aldrich), a toxic volatile compound produced by
Validation of rhizodeposit metabolites collected via root dip originating in connection with vegetative growth. The concern was that sloughed off root or callous cells would contribute more greatly to rhizodeposits than other modes of rhizodeposition. After rinse and soaking steps, 12 micropropagated trees (genotype MM106) were divided into treatment groups (butyrolactone [butyr] and untreated control [ctl]), incubated for another day in water, then rhizodeposits collected (d1). Following rhizodeposit collection (d1) and transfer to new water, micropropagated trees were treated with the toxic volatile compound butyrolactone to kill/reduce growth of vegetation, and after 4 days in new water, rhizodeposits collected (d4). Differing letters (a–c) above bars indicate statistically significant means.
Peak areas were normalized using the ibuprofen internal standard as a reference feature, and data were scaled using autoscaling (mean-centered and divided by the standard deviation of each variable) in MetaboAnalyst 3.0 (Xia et al.,
Sugars, organic acids, amino acids, and sugar alcohols were analyzed via GC-MS following derivatization of lyophilized samples (Table
Metabolites assessed via derivatization and GC-MS analysis included amino acids, organic acids, sugars, sugar alcohols, and other compounds.
L-asparagine [3TMS] | 11.2 | 159 | 2891380 | 2764796 | |
L-cysteine [3TMS] | 11.4 | 406 | 2157648 | 1412312 | |
L-aspartic acid [3TMS] | 10.7 | 232 | 848542 | 196042 | * |
B-alanine [3TMS] | 10.0 | 160 | 557381 | 175848 | |
L-valine [2TMS] | 8.4 | 144 | 479378 | 366937 | |
Malic acid [3TMS] | 10.5 | 335 | 6127902 | 5718066 | |
Succinic acid [2TMS] | 9.1 | 147 | 1179148 | 497640 | * |
Malonic acid [2TMS] | 8.3 | 147 | 617007 | 1337261 | * |
Oxalic acid [3TMS] | 7.8 | 235 | 325887 | 615729 | |
Galacturonic acid [1TMS] | 15.6 | 375 | 217022 | 186028 | |
Lactic acid [2TMS] | 6.9 | 117 | 210143 | 171185 | |
Glyceric acid [3TMS] | 9.3 | 315.0 | 104137 | 71080 | * |
Pyruvic acid [1TMS] | 7.0 | 695 | 92595 | 51891 | |
Hexose [5TMS] | 13.8 | 319 | 574733 | 214959 | * |
Gentiobiose [8TMS] | 17.8 | 361 | 463491 | 1467215 | |
Xylose [4TMS] | 11.7 | 307 | 386442 | 429090 | |
Gulose [5TMS] | 16.0 | 217 | 382891 | 183284 | |
Ribose [4TMS] | 11.8 | 307 | 381419 | 468254 | |
Rhamnose [4TMS] | 12.1 | 217 | 377152 | 408940 | |
Galactose [5TMS] | 15.2 | 319 | 116357 | 230342 | * |
Erythrose [5TMS] | 10.2 | 317 | 108654 | 81221 | * |
Ribofuranose [4TMS] | 16.8 | 509 | 13591 | 165 | |
Sorbitol [6TMS] | 13.4 | 319 | 41071653 | 31450752 | |
Erythritol [4TMS] | 13.7 | 423 | 1228877 | 892587 | |
Threitol [1TMS] | 10.9 | 307 | 737590 | 542915 | |
Galactitol [6TMS] | 15.6 | 421 | 57379 | 47484 | |
(Urea |
8.5 | 189 | 12951401 | 12407485 | |
Glycerol |
8.8 | 307 | 6408323 | 2320712 | * |
Ursolic acid [2TMS] | 22.4 | 42 | 1278650 | 2681073 | |
(Phosphoric acid |
8.9 | 189 | 659084 | 1338218 | * |
(Ursane2of4) [3TMS] | 21.4 | 320 | 142450 | 479239 | |
(Octadecanoic acid |
14.8 | 341 | 78556 | 91641 | |
(Ursane3of4) [2TMS] | 22.1 | 571 | 16241 | 40512 | |
(Oleonitrile |
16.0 | 319 | 14467 | 20213 | |
(Ursane1of4) [2TMS] | 21.0 | 482 | 3816 | 26582 |
PCA/cluster analysis of combined metabolic profiles of sugars, organic acids, amino acids, and sugar alcohols indicated a general difference in relative quantity of many compounds assessed according to rootstock genotype (Figure
Metabolic analysis of compounds analyzed via metabolite derivitization (tri-methylsilylation) and GC-MS analysis, which includes organic acids, amino acids, sugars, and sugar alcohols. Principal components analysis
Several metabolites detected in rootstock rhizodeposits compounds analyzed were also found in the agar-based media in which trees rooted prior to rhizodeposit collection, with the major metabolites including fructose, glucose, myo-inositol, and L-alanine. Every effort was made to remove media from the roots, including an overnight “soak” which had the dual purpose of callus tissue removal. However, although these metabolites could also be part of the apple rhizodeposit suite of metabolites based on reports of rhizodeposits from other plant species (Badri and Vivanco,
In the first experiment, only one phenolic compound (phloridzin) statistically differed in relative quantity between rootstocks; it was higher in M26 (data not shown).
The identified compounds that differed significantly (
Metabolites analyzed by LC-MS analysis included compounds tentatively identified as triterpenoids, flavonoids, chlorogenic acids, benzoic acids, procyanidins, and other phenolic compounds.
(Dihydroxy-urs-12-3n-28-oic acid) | 471.3474 | 471.3515 | −8.8 | 76.9 | _Triterpenoid_471_2 | 444964656 | 386003980 | |
(Trihydroxy-urs-12-en-28-oic acid) | 487.3424 | 487.3465 | −8.5 | 64.9 | _Esculentic acid | 183735512 | 300310905 | |
(Dihydroxy-urs-12-3n-28-oic acid) | 471.3474 | 471.3520 | −9.8 | 76.5 | _Triterpenoid_471_1 | 239898362 | 257066843 | |
(Trihydroxy-urs-12-en-28-oic acid) | 487.3423 | 487.3410 | 2.6 | 66.7 | _Triterpenoid_487_2 | 43279565 | 71329308 | |
(3-oxo-1?-hydroxy-urs-12-en-28-oic acid) | 469.3318 | 469.3317 | 0.2 | 75.1 | _Pomonic acid | 50216895 | 36395157 | |
Ursolic acid | 455.3524 | 455.3559 | −7.7 | 79.5 | Ursolic acid | 44839049 | 25577316 | |
(3-oxo-19α-dihydroxy-urs-12-en-28-oic acid) | 485.3267 | 485.3298 | −6.3 | 66.9 | _Triterpenoid_485 | 7363407 | 11435877 | |
(Oleaonolic acid) | 455.3524 | 455.3559 | −7.7 | 75.2 | _Oleonolic acid | 3884911 | 2638595 | |
(3-oxo-19α-dihydroxy-urs-12-en-28-oic acid) | 485.3266 | 485.3266 | 0.1 | 69.6 | _Annurcoic acid | 2791337 | 1458985 | |
Phloridzin | 435.1291 | 435.1325 | −7.8 | 27.9 | Phloridzin | 3597822 | 29849517 | * |
Phloretin | 273.0761 | 273.0704 | 20.9 | 42.2 | Phloretin | 8733671 | 21845333 | |
Phloretic acid | 165.0551 | 165.0567 | −9.9 | 10.1 | Phloretic acid | 343892 | 716204 | |
Phloroglucinol | 125.0238 | 125.0264 | −20.6 | 2.6 | Phloroglucinol | 308726 | 352251 | |
Epicatechin | 289.0711 | 289.0729 | −6.0 | 10.2 | Epicatechin | 27960574 | 20283468 | |
Hyperin | 463.0880 | 463.0890 | −2.1 | 21.8 | Hyperin | 6852133 | 3916945 | |
Quercitrin | 447.0931 | 447.0937 | −1.3 | 26.2 | Quercitrin | 2455612 | 933915 | |
Isoquercitrin | 463.0871 | 463.0891 | −4.4 | 21.8 | Isoquercitrin | 294231 | 649031 | |
Kaempferol-2-rutinoside | 593.1502 | 593.1535 | −5.6 | 25.8 | Kaempferol2rut | 182784 | 442723 | |
kaempferol-3-glucoside | 447.0930 | 447.0934 | −0.9 | 29.7 | Kaempferol3gluc | 17434 | 369714 | * |
Catechin | 289.0711 | 289.0725 | −4.8 | 5.9 | Catechin | 206357 | 303534 | |
(Reynoutrin) | 434.0850 | 434.0933 | −19.2 | 50.5 | _Reynoutrin | 0 | 297345 | |
Rutin | 609.1451 | 609.1472 | −3.5 | 21.8 | Rutin | 57712 | 32809 | |
Chlorogenic acid (3-O-caffeoylquinic acid) | 353.0872 | 353.0895 | −6.6 | 6.5 | Chlorogenic acid | 19523743 | 15623595 | |
Quinic acid | 191.0551 | 191.0572 | −11.2 | 1.6 | Quinic acid | 20978217 | 10633187 | |
(P-coumaroylquinic acid) | 337.0921 | 337.0943 | −6.4 | 10.3 | _Pcouquin acid | 18129527 | 7244580 | |
(P-coumaroylquinic acid) | 337.0921 | 337.0941 | −5.9 | 8.0 | _Pcouquin_unk1 | 3116356 | 5048331 | |
Caffeic acid | 179.0341 | 179.0373 | −17.8 | 7.3 | Caffeic acid | 1793336 | 1551952 | |
(P-coumaroylquinic acid) | 337.0921 | 337.0942 | −6.2 | 13.3 | _Pcouquin_unk2 | 6181379 | 1154323 | * |
4-p-coumaric acid | 163.0394 | 163.0411 | −10.3 | 13.9 | P-coumaric acid | 958879 | 583182 | |
(4-hydroxbenzoic acid) fragment | 137.0238 | 137.0260 | −16.1 | 15.6 | 4HBunk1 | 1789184 | 804328 | * |
(P-hydroxybenzoic acid alkyl ester) | 285.0611 | 285.0557 | 19.0 | 5.1 | _Uralenneoside | 1628920 | 1143132 | |
3,4 hydroxybenzoic acid | 153.0191 | 153.0220 | −19.0 | 2.8 | 34HB | 2225856 | 1639588 | |
4-hyrdroxybenzoic acid (authentic standard) | 137.0238 | 137.0291 | −38.6 | 3.2 | 4HB | 7519629 | 2557597 | |
Benzoic acid | 121.0289 | 121.0310 | −17.5 | 14.5 | Benzoic acid | 825223 | 3678816 | |
(Phenolic glycoside) | Unknown | 461.1672 | 7.2 | _Dicaffeoylacteoside | 2126495 | 1275547 | ||
Gallic acid | 169.0141 | 169.0166 | −14.9 | 1.7 | Gallic acid | 493283 | 910044 | |
Hydrocaffeic acid | 181.0501 | 181.0531 | −16.7 | 6.8 | Hydrocaffeic acid | 700079 | 602926 | |
Salicin | 285.0971 | 285.1004 | −11.6 | 4.8 | Salicin | 122619 | 182760 | |
(Esculin) | 339.0712 | 339.0747 | −10.3 | 5.2 | _Esculin | 213443 | 660061 | * |
Scopoletin | 191.0350 | 191.0371 | −11.0 | 15.2 | Scopoletin | 56206 | 61686 | |
(Eriobofuran) | 243.0740 | 243.0663 | 31.7 | 53.1 | _Eriobo2 | 2365538 | 2857828 | |
(Noraucuparin) | 215.0712 | 215.0711 | 0.5 | 27.3 | _Noraucuparin | 388993 | 883580 | |
(Aucuparin) | 229.0861 | 229.0861 | 0.2 | 2.1 | _Aucuparin | 130917 | 72510 | |
(Eriobofuran) | 243.0661 | 243.0671 | −4.1 | 41.3 | _Eriobo1 | 74616 | 656321 | * |
Procyanidin B2 | 577.1345 | 577.1367 | −3.8 | 8.0 | Procyanidin B2 | 511461 | 774127 | |
(Procyanidin) | 577.1345 | 577.1367 | −3.8 | 6.4 | _Procyanidin_6_55 | 2743098 | 758511 | |
(Procyanidin) | 865.1981 | 865.2016 | −4.0 | 12.3 | _ProcyC1 | 169764 | 142725 | |
(Procyanidin) | 577.1345 | 577.1370 | −4.3 | 21.4 | _Procyanidin_21_48 | 0 | 58335 |
Principal components analysis
Tentative identification of triterpenoids was based on agreement of accurate mass with ursenoic acids in apple peel reported by McGhie et al. (
Summed peak areas from analysis of polar compounds (zicpHILIC) and phenolic compounds indicated that trees treated with the toxic compound butyrolactone had reduced production of metabolites (Figure
Fungi were detected by qPCR in 30% of the samples (Supplemental Table
Previous work indicated that apple rootstock genotype impacts rhizodesposit quantity and composition, and is also influenced by environmental, physiological, and ontological factors (Leisso et al.,
In the current study, several organic acids were detected at relatively high concentration, including malic acid, succinic acid, malonic acid, and oxalic acid. Functionally, organic acids released by roots into the rhizosphere can influence nutrient availability through changes in soil pH. For example, salicylic and citramalic acid are found in high concentrations in sugar beet rhizodeposits and are able to solubilize phosphorus (Khorassani et al.,
Urea levels were unexpectedly high in rhizodeposits but did not differ significantly between genotypes; this metabolite was not assessed in the previous study (Leisso et al.,
Among identified phenolic compounds, phloridzin, kaemferol-3-glucoside, and esculin differed in relative quantity between G935 and M26. Phloridzin has been reported repeatedly as a major phenolic component of apple roots and bark and yet remains ambiguous in terms of impact on pathogen and other microbial populations. Some studies have reported phloridzin to possess detrimental allelopathic effects on apple seedlings (Yin et al.,
Chlorogenic acid, quinic acid, and epicatechin were additional phenolic compounds found in relatively high levels in rootstock rhizodpeposits. Epicactechin has been reported in high levels in other rosaceous plants (Oszmianski et al.,
An intriguing result is the relatively high levels of putative triterprenoids present in micropropagated tree rhizodeposits. The structure of the compounds detected in the present study were not fully elucidated but the masses correspond to triterpenoid compounds reported in apple peel tissue (McGhie et al.,
The question remains as to which compounds or classes of compounds has the most influence on the rhizosphere microbiome in terms of functions that impact tree health and apple fruit production. In
Another point in consideration of rhizodeposit assessment is the biochemical/physical relationship roots have with their external media in terms of rhizodeposit release. The large solute concentration gradient in the root, relative to the external medium, likely causes molecules with no charge, like sugars, to be lost through passive diffusion (Jones et al.,
Endophyte analyses were inconclusive, which was somewhat unexpected, as trees had been propagated from actively growing shoot tips of trees, and because until recently (Liu et al.,
Our findings confirmed that vegetative activity contributes to the levels of bulk rhizodeposits, indicating both that rhizodeposits may be translocated from photosynethic activity in the leaves and that sloughed off root tissue or cells are not likely to be the only source of rhizodeposits. This concurs with our previous results where total abundance of rhizodeposit metabolites was well correlated with total leaf area (Leisso et al.,
This study reports methods for axenic collection of apple rhizodeposits, and describes the relative levels of a number of compounds in collected rhizodeposits from two apple rootstock genotypes (G935 and M.26), including phenolic compounds, sugars, sugar alcohols, amino acids, organic acids, and several putative triterpenoids. Potential impacts of these metabolites in terms of plant health include their function as substrate sources for beneficial microorganisms or in demonstrating inhibitory activity toward various plant pathogens. A prospective long-term vision for this line of research is to integrate apple genotype-specific rhizodeposition patterns or capacity with particular soil characteristics to enable site-suitability decisions for apple rootstock genotypes or appropriate site modifications to enhance horticultural output. To achieve this, additional research is needed to (1) further delineate rhizodeposit metabolites that are consistently produced at levels high enough to have functional impacts on both soil chemistry and microbial ecology, (2) determine the effects of major rhizodeposits on microbial communities, including identifying endophytes and their contributions to rhizodeposits, (3) further determine the primary environmental, ontological, and physiological drivers that change levels of rhizodeposits, and (4) integrate rootstock genotype, rhizodeposition, rhizosphere microbiome, and soil chemistry to enable precision agricultural determination of appropriate system modifications.
RL designed and carried out the research. DR contributed methods for instrumental analysis and MM contributed to experimental design and writing.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We gratefully acknowledge Natalie O'Hara for assistance with apple tree micropropagation. This work was supported by the Washington Tree Fruit Research Commission (project number AP-16-106), USDA-National Institute of Food and Agriculture (NIFA)-Agriculture Food and Research Initiative project 2017-67012-26093, USDA-NIFA - Specialty Crop Research Initiative project 2016-51181-25406.
The Supplementary Material for this article can be found online at: