Distribution of 3-Isobutyl-2-methoxypyrazine across Rachis Components of Vitis vinifera Shiraz and Cabernet Sauvignon

Rootstock can signifcantly alter the concentration of methoxypyrazines (MPs) in the bunch stem (rachis) of Vitis vinifera L. cv. Cabernet Sauvignon and Shiraz, which has implications for winemaking and wine style. Te distribution of MPs across the rachis is an important consideration, but such information was not available. Tis study aimed to address this research question by comparing MP concentrations in diferent rachis components throughout grape maturation and in the absence of ambient light. Shiraz and Cabernet Sauvignon bunches were sampled throughout development, segmented into four components (peduncle, top rachis, bottom rachis, and pedicel), and 3-isobutyl-2-methoxypyrazine (IBMP) was quantifed in each. For both cultivars, IBMP showed a negative correlation with grape maturity, with concentrations in pedicel at harvest being signifcantly higher than other rachis components. Additionally, light exclusion signifcantly increased IBMP concentrations in all rachis segments. Te concentration of IBMP varied signifcantly between diferent rachis components. Te greatest concentrations were measured in the pedicel, which also contributed the largest proportion among the components to total rachis by weight. Due to elevated IBMP concentrations in rachis and the difculties in excluding matter other than grape from a fermentor, the presence of pedicel during fermentation could produce Shiraz and Cabernet Sauvignon wines with higher concentrations of MPs, thereby potentially increasing vegetal sensory characteristics.


Introduction
Te chemical composition of berries is heterogeneous within a vineyard, vine and bunch, and this variability could alter the sensory properties of a wine if heterogeneous grape parcels are harvested [1]. Asynchronous berry development contributing to heterogeneity can be attributed to aspects of terroir, which encompass geographical and climatic diferences between grape growing regions, and the spatial variation of soil, sunlight, slope, and water availability within a vineyard [2].
Winemakers may opt to include rachis during whole-or partial-bunch fermentation to produce desirable tannin [9,10], colour [10,11], pH, or ethanol changes to the wine [9,12]. However, these techniques are avoided for varieties that are known to produce IBMP (an impact odorant with green capsicum character) in rachis and berry, such as Cabernet Sauvignon, because they can cause the perception of "stemmy" favours in wine [13]. Even for varieties where a portion of whole bunches is favourable in producing quality wines, such as Pinot noir, stem addition exceeding 60% can produce wines with sensory characteristics associated with MPs [9]. Rachis can also be unintentionally present during fermentation as a by-product of harvesting and destemming practices, as a component of matter other than grape (MOG). Historically, MOG levels of approximately 5% w/w have been found in machine harvested fruit [14]; however, recent technological advancements can decrease these levels to around 1% [15]. Even so, there should still be concern, given the potential for MOG in the form of rachis to impart undesirable and nonvarietal "green" sensory characteristics to a variety like Shiraz [16], which otherwise lacks the genetic ability to produce IBMP in the berry [17].
Being characteristic of certain grape varieties, IBMP is a "varietal" aroma compound that can contribute notes of "green capsicum" and "grassy" to red wine at concentrations of 10-15 ng/L [18]. Such sensory characteristics are desirable when in balance with an overall wine bouquet, but at elevated concentrations, IBMP can contribute "herbaceous" and "vegetative" aromas that can dominate the sensory experience, decreasing both consumer liking and positive emotions associated with the wine [19]. Furthermore, even at concentrations below its sensory threshold, IBMP can alter wine aroma through synergistic interactions increasing perception of "smoky" and "tar" notes or antagonistic interactions that decrease desirable "red berries" and "foral violet" aromas [20]. As such, understanding how to control the concentration of IBMP (and other MPs) in wine is essential.
Recent research has shown Shiraz [8,21] and Cabernet Sauvignon [22] vines grafted to rootstock can have significantly higher IBMP concentrations in rachis than those on own roots. Tis was attributed in part to rootstock-mediated vine vigour altering ambient light exposure of bunches [21,22]. Furthermore, in Shiraz rachis the concentrations of IBMP, IPMP, and SBMP increased throughout berry maturity [8], contrasting the negative relationship between these variables observed in the berry [23].
With respect to the contribution of rachis to MOG and likelihood of contributing unwanted sensory characters, it can be expected that the pedicel component of rachis could most easily enter a fermentor, due to its small size or attachment to the berry. Although studies have addressed MP concentrations in rachis overall, the MP concentration of diferent parts of the rachis remained to be investigated. Tis study aimed to fll that knowledge gap by determining the concentration of IBMP, IPMP, and SBMP in diferent rachis components throughout grape maturation for Vitis vinifera L. cv Shiraz on Ramsey rootstock and own roots grown in the Barossa Valley, and Cabernet Sauvignon on 110 Richter rootstock grown in the Coonawarra. Trials exploring the exclusion of ambient light on MP distribution in Shiraz bunches were also undertaken. MPs were quantifed by GC-MS/MS using an established stable isotope dilution assay and experimental data were analysed with linear mixed models (LMMs). Results from the study were intended to provide producers with an understanding of how the concentration of MPs across rachis components is infuenced by grape maturity and light exposure, thereby giving information that helps to estimate their potential infuence on wine sensory profles.  [24] with the value of 1.00 used for the length of day coefcient. Rainfall and Huglin index data are summarised in Table S1 [25].

Materials and Methods
Samples from the Coonawarra region were collected from a premium commercial vineyard (37°15′47.4″S, 140°49′58.7″E). Te vineyard was established in 2012 and consists of Vitis vinifera L. cv. Cabernet Sauvignon Reynella and SA125 clones on 110 Richter rootstock.
Te parentage of rootstocks present in the trials is summarised in a recent publication [21]. No signifcant pest or disease pressures were observed during the experimental seasons.

Maturity Variation Experiment.
In 2019/20, Shiraz samples were taken at fowering (80% cap fall) on the 26 th of November, 50% veraison on the 8 th of January, and harvest on the 9 th of March. Sampling locations at each time point were chosen to provide a representative sample from the southern, centre, and northern regions of the vineyard. At fowering, twenty-four bunches were collected from each sampling location and used to create six biological replicates for each rachis component (peduncle, top rachis, and bottom rachis). No pedicel material was retained at fowering due to its small size at this phenological stage and sampling limitations. At veraison and maturity, twelve bunches were collected from each sampling location and used to create six biological replicates of each rachis component (pedicel, peduncle, top rachis, and bottom rachis) ( Figure 1). Shiraz sampled in 2021/22 was taken from the same locations within the vineyard as for 2019/20 at fowering (80% cap fall) on the 25 th of November, 50% veraison on the 27 th of January, and harvest on the 15 th of March. In 2021/22, each sampling location was further divided into two six-vine subregions designated east and west. At all time points, twenty-four bunches (twelve from each east and west subregion) were collected from each vineyard location and used to create eight biological replicates (four from each east and west subregion) of each rachis component (peduncle, top rachis, bottom rachis, and pedicel) ( Figure 1). As for 2019/20, pedicel material was only collected at veraison and maturity. All rachis material was transported to the laboratory on ice where berry material was removed, and rachis was segmented into components, cut into approximately 1 cm pieces, frozen in liquid nitrogen, and stored at −80°C until analysis. An overview of the sampling methodology for Shiraz from Barossa Valley can be found in Figure S1 of the Supplementary Information.
Cabernet Sauvignon was sampled in the 2019/20 season at 80% veraison on 13 th February and at harvest on the 20 th of March. Sampling occurred prior to commercial harvest (15 th of April) due to complications arising from the COVID-19 pandemic. Sampling locations were the same for each time point and were chosen to provide a representative sample from the southern, centre, and northern regions of the vineyard. From each sampling location, six bunches were chosen at random and used to create two biological replicates of each rachis component (peduncle, top rachis, bottom rachis, and pedicel) ( Figure 1) per vineyard location. All rachis material was transported to the laboratory on ice where berry material was removed, and rachis was segmented into components, cut into approximately 1 cm pieces, frozen in liquid nitrogen, and stored at −80°C until analysis.
Shiraz and Cabernet Sauvignon grape bunches were collected at harvest from each sampling location and total soluble solids (TSS) (°Brix) were measured from two biological replicates of homogenate prepared by manually crushing 50 fresh berries randomly selected from bunches.

Exclusion of Sunlight on the Distribution of Methoxypyrazines in Grape
Rachis. Experiments were performed over the 2021/22 growing season in the Barossa Valley and utilised opaque boxes designed to eliminate ambient light from the grape bunches while preventing temperature and humidity changes [26]. Te experiment comprised boxed (box) and nonboxed (control) vines, with the treatment vines in each vineyard plot chosen to be representative samples of the southern, centre, and northern regions of the vineyard. Control bunches were those used in the maturity variation experiment and were sourced from vines located next to experimental vines. Te box treatment was applied at 1 week postfowering (wpf ) on the 23 rd of November to whole bunches on own roots (n � 24) and Ramsey (n � 20). Samples were harvested at 14 wpf on the 15 th of March and processed as described previously [22]. For box and control samples, TSS (°Brix) were measured from a homogenate prepared by manually crushing 15 fresh berries randomly selected from each box. TSS values for control were taken as outlined above.

Weight of Individual Rachis Components in Shiraz.
Shiraz material (n � 26) from Wrattonbully, South Australia was sampled at commercial harvest in 2022, frozen at −20°C, transported in Styrofoam boxes on ice, and stored at −30°C for 4 months until processing. Rachis material from each bunch was segmented into peduncle, top rachis, bottom rachis, and pedicel, and individual weights of each component were determined.

Measures of Canopy Architecture.
A surrogate measure of vine vigour was obtained in the Barossa Valley on the 17 th of January 2022 using a LICOR LAI-2200C Plant Canopy Analyser. For every six-vine subregion (east and west) at each sampling location, one above-canopy reading (ambient light), and four below-canopy readings were taken to provide an estimate of leaf area index (LAI).

Top rachis
Bottom rachis Pedicel Figure 1: Schematic outlining the components that the rachis material was segmented into prior to extraction and analysis of methoxypyrazines.

Quantitation of Methoxypyrazines.
MPs were quantifed in grape and rachis tissue using a stable isotope dilution assay with headspace SPME-GC-MS/MS [22] with modifcations to sample preparation. Briefy, the modifcations involved frozen rachis tissue (0.5-2 g, dependent on component) being ground to a fne powder with a cryomill (Retsch, Germany) in liquid nitrogen. Approximately 200 mg of rachis from the sunlight exclusion experiment and 500 mg of rachis from the maturation trials was accurately weighed for extraction and analysis as per the previous report [22]. Te respective limit of detection (LOD) and limit of quantitation (LOQ) values (ng/kg) were 0.13 and 0.44 for IBMP, 0.11 and 0.37 for IPMP, and 0.15 and 0.48 for SBMP [22]. . Mixed-efect linear regression models (linear mixed model, LMM) were used to determine treatment efects on log-transformed MP concentrations with the "lmerTest" package. For the Shiraz maturity variation experiment, rootstock, vintage, berry maturity, and rachis component were set as fxed efects, and vineyard block was set as a random factor. For the light exclusion trials, rachis component, light, and rootstock were set as fxed efects, and vineyard row and°Brix were included as random factors. For the Cabernet Sauvignon maturity variation experiment, rootstock, berry maturity, and component were set as fxed efects and vineyard row was set as a random factor. Estimated marginal means (statistically modelled variable mean response for each level of a predictor variable) and standard error (SE) values of the models were calculated on backtransformed values using the "emmeans" package and compared using Bonferroni-adjusted pairwise comparisons. Summary statistics for variables with measurements below the LOD and/or LOQ were calculated using the Kaplan-Meier technique with Efron's bias correction using the "NADA" package. Summary plots were produced using "ggplot2".

Concentration and Distribution of 3-Isobutyl-2-Methoxypyrazine in Shiraz Rachis Components throughout Berry
Maturation. Shiraz bunches were collected at harvest in 2022 from the Wrattonbully wine region to determine the proportion (% w/w) of individual rachis components (peduncle, top rachis, bottom rachis, pedicel) of the total rachis fresh weight. Timing of sampling was an important consideration; rachis reaches its defnite size by veraison [27] but dehydration of the peduncle continues until harvest [28], which would decrease its fresh weight and cause a concentration efect of IBMP within this component. In addition, Shiraz rachis grown on Ramsey rootstock and own roots has been shown to increase in IBMP concentration through berry maturation, reaching its maximum at harvest [8].
Shiraz rachis fresh weight at harvest was calculated as 6.5% of the total bunch weight, which falls within the range of 3-7% for Vitis vinifera rachis [27]. Te proportion (% w/w) of individual rachis components of the total fresh rachis weight was calculated for peduncle (10%), top rachis (18%), bottom rachis (9%), and pedicel (63%) ( Figure S2 of the Supplementary Material). Tese values were used throughout this publication to estimate the contribution from each rachis component to the total IBMP concentration present within an average Shiraz rachis. To the best of our knowledge, this is the frst published segmentation data for rachis material, with a single previous study showing that peduncle was approximately 20% of Shiraz rachis fresh weight [28].  (Table S3 of the Supplementary Material) was dependent on rootstock and berry maturity (P < 0.001, Figure 2).
According to the LMM, concentrations of IBMP ranged at veraison from 2.51 ng/kg rachis for peduncle from own roots at harvest to 142 ng/kg rachis for Ramsey pedicel. At veraison, IBMP concentrations were signifcantly higher in Ramsey than own roots for pedicel (142 and 54.3 ng/kg rachis, respectively), top rachis (18.8 and 7.74 ng/kg rachis), and bottom rachis (12.9 and 3.58 ng/kg rachis), although this diference was no longer signifcant for any rachis component at harvest. Tis was somewhat surprising considering that own roots had a higher measure of vine vigour than Ramsey (P < 0.001, Table S2), and vigour has previously been positively correlated with rachis IBMP concentrations at harvest [21]. However, that prior research also showed no signifcant diference between IBMP concentrations in own roots and Ramsey Shiraz rachis at harvest for three vintages [21] that were sampled from the same vineyard used in the current work.
For both Ramsey and own roots, the concentration of IBMP in top rachis, bottom rachis, and peduncle signifcantly decreased between fowering and harvest (P ≤ 0.05) with a downward trend evident as grape maturity increased ( Figure 2). For pedicel, Ramsey was signifcantly lower at harvest (63.9 ng/kg rachis) than veraison (142 ng/kg rachis), and while own roots appeared to trend upwards from veraison (54.3 ng/kg rachis) to harvest (81.4 ng/kg rachis), there was no signifcant diference.
In contrast, previous research reported Shiraz rachis from the Murray Darling region of Victoria increased in MP concentration throughout the 2017/18 growing season, with IBMP, IPMP, SBMP being above the LOD by 8 wpf [8]. In the current study, however, SBMP remained below the LOD throughout both studied vintages, and IPMP was detected only sporadically and at low concentrations (data not shown). As rachis is composed of approximately 55 to 80% water [29], increased dehydration of rachis due to climate diferences between the regions may have elevated the concentrations of MPs in Murray Darling rachis relative to that from the Barossa Valley. As calculated by the Huglin index, the Murray Darling region was classifed as very warm (3181) for the 2017/18 season, whereas Barossa Valley was classifed as temperate warm for both 2019/20 (2379) and 2021/22 (2252).
Overall, the close agreement in trends and concentrations for Ramsey and own root rachis seen herein suggest that rootstock may have no bearing on the way that IBMP is distributed within rachis. Alternative explanations for the distribution of IBMP in diferent rachis components could relate to diferences in light environment due to berry or vine shading.

Vintage and Stage of Grape Development.
Te concentration of IBMP in Shiraz rachis components from the Barossa Valley was shown to be dependent on the simple main efect of vintage (P < 0.001) according to the LMM, with an average IBMP concentration of 103 ng/kg and 57.6 ng/kg in the 2019/20 and 2021/22 growing seasons, respectively. Grape maturity at harvest (as a surrogate measure of rachis maturity) varied signifcantly (P < 0.001) between vintages, with 2019/20 (27.4°Brix) being signifcantly lower than 2021/22 (28.9°Brix). Broadly these results suggest a negative relationship between IBMP concentration in rachis and grape maturity, similar to that observed for Cabernet Sauvignon berries [23].
In addition, the concentration of IBMP in rachis was dependent upon a three-way interaction between component, vintage, and berry maturity (P � 0.03) according to the LMM (Figure 3). IBMP concentrations ranged from a high of 110 ng/kg rachis for pedicel at veraison in 2019/20 down to 0.87 ng/kg rachis for peduncle at harvest in the same season. IBMP concentrations in pedicel were higher than all other rachis components at veraison and harvest, but pedicel did not signifcantly difer at either maturity time point in 2019/20 or 2021/22 (Figure 3). In 2021/22, top rachis (18.4 ng/kg rachis) and bottom rachis (9.59 ng/kg rachis) were diferent (P < 0.05) at veraison, which may suggest that the regulation of IBMP distribution in these organs is somewhat variable. Light exposure can signifcantly alter MP accumulation in Shiraz rachis material [21], so it was theorised that top rachis could experience higher levels of ambient light than bottom rachis throughout the growing season, due to its more exposed position in the bunch. As pedicel was not separated by bunch position, it is feasible that pedicel material could also difer in IBMP content based on its location within a bunch. However, due to shading from the berry, the variability in pedicel light exposure across the bunch should be lower than for top versus bottom rachis, although this aspect could be evaluated in future.
While the average IBMP concentration was signifcantly higher in 2019/20 (103 ng/kg) than 2021/22 (57.6 ng/kg), all components at harvest except for pedicel were signifcantly higher in IBMP concentration in 2021/22 than 2019/20. Vintage efects in rachis IBMP are attributed to diferences in growing season temperature, with a negative correlation between temperature and IBMP in rachis proposed for Shiraz [21] and Cabernet Sauvignon [22]. However, the vintage efects seen within the current results were not readily explained due to climatic variables. Te Huglin index values of 2379 and 2252 the 2019/20 and 2021/22 growing seasons, respectively, were similar and both were classifed as "temperate warm," suggesting minimal temperature variation overall.
Notably, the IBMP concentration in pedicel material at harvest from either vintage was signifcantly higher than all other rachis components on a ng/kg of rachis basis (Figure 3), which suggests that pedicel is the most substantial source of rachis IBMP. As concentrations of IBMP in pedicel are not signifcantly diferent between veraison and harvest, testing the pedicel at veraison could provide valuable information about potential wine sensory outcomes for winemakers who are considering partial-or whole-bunch fermentation or mechanically harvesting fruit without any sorting.

Ambient Light Exclusion throughout Maturation.
As with berry, light exclusion is known to yield signifcantly higher concentrations of IBMP, IPMP, and SBMP in Shiraz [21] and Cabernet Sauvignon [22] rachis, but the efect of light on MP distribution across the rachis was unknown.
Addressing the hypothesis that diferences in natural light exposure could be responsible for the observed diferences in IBMP concentration between rachis components (peduncle, top and bottom rachis, and pedicel), light exclusion boxes were applied to Shiraz grape bunches at 1 wpf on Ramsey rootstock and own roots (box) with nonboxed bunches (control) taken from nearby vines at harvest. Tere was no signifcant diference in maturity (°Brix) for Shiraz berries obtained from Ramsey (28.2 ± 1.53) and own roots (29.7 ± 1.41) box samples at harvest. Te concentration of IBMP in Shiraz rachis at harvest (Table S4 of the Supplementary Material) for box and control samples was signifcantly dependent upon the simple main efect of rootstock (P < 0.001) (Figure 4) according to the LMM. Te estimated marginal means were 72.2 ng/kg rachis (own roots) and 143 ng/kg rachis (Ramsey) for control bunches, and 780 ng/kg rachis (own roots) and 1219 ng/kg rachis (Ramsey) for box bunches. Although slightly lower, these values were refective of previous research involving light exclusion trials on Shiraz, with box bunches being substantially higher than controls [21]. IPMP and SBMP were not detected in any control samples but were above the LOD for 100% and 96% of Ramsey box samples (Table S4 of the Supplementary Material). However, due to their low concentrations, the data were not analysed further.
An interaction between rachis component and light (P < 0.001) signifcantly afected the concentration of IBMP according to the LMM. Te impact of light exclusion on IBMP is visualised on a ng/kg rachis ( Figure 5(a)) and a ng/kg component ( Figure 5(b)) basis.
On a per kilogram of rachis basis ( Figure 5(a)), the marginal means for IBMP concentration ranged from 6.58 ng/kg rachis (bottom rachis, control) to 597 ng/kg rachis (pedicel, box). Pedicel box was signifcantly higher in IBMP concentration than all other components, independent of light conditions. IBMP concentrations in pedicel control (64.5 ng/kg rachis) were equivalent to peduncle box (53.98 ng/kg rachis), and higher (P ≤ 0.05) than top rachis control (9.85 ng/kg rachis), peduncle control (8.95 ng/kg rachis), and bottom rachis control (6.58 ng/kg rachis). Box and control samples trend in a similar manner, suggesting that bunch light exposure is an important consideration for controlling IBMP within a vineyard. On a per kilogram of component basis ( Figure 5(b)), the marginal means of IBMP varied from 53.8 ng/kg component (top rachis, control) to 1260 ng/kg component (bottom rachis, box). Concerning light excluded components, bottom rachis was signifcantly higher than top rachis (827 ng/ kg component) and peduncle (469 ng/kg component), but not signifcantly diferent to pedicel (961 ng/kg component). Pedicel was equivalent to bottom and top rachis but higher than peduncle. Notably, the pattern of distribution across the rachis was dissimilar to the control samples, although there were still signifcant diferences between rachis components ( Figure 5(b)), suggesting that bunch light exposure might contribute to the regulation of IBMP movement, biosynthesis, and/or storage in various components of rachis, perhaps in unison with other regulatory processes.
IBMP biosynthesis from the precursor hydroxypyrazine (IBHP) is regulated in berry by the VvOMT gene family, primarily through the activity of the methyltransferase enzyme VvOMT3 [30,31]. Expression ofVvOMT3 is upregulated in the berry when ambient light is excluded, signifcantly increasing concentrations of IBMP [23]. Te exclusion of light also increases the concentration of IBMP in rachis [21,22], but the molecular basis remains unknown. A study of Shiraz rachis found that IBMP biosynthesis throughout the growing seasons was not correlated with the levels of VvOMT3 expression [8]. Instead, those researchers proposed that translocation from other vines organs, particularly the roots where there are elevated concentrations of IBMP, could explain IBMP concentrations in the rachis.
However, while the expression of genes in the VvOMTfamily has been shown to vary between vine organs [32], the differential expression of VvOMT3 across and within vine components remains unexplored. Te signifcantly elevated concentrations of IBMP in the bottom rachis under light exclusion conditions ( Figure 5(b)) may suggest that a targeted approach measuring gene expression in the diferent rachis components should be a consideration for future work that aims to elucidate IBMP biosynthesis in rachis.
Although light does not afect IBMP distribution in rachis in a uniform manner, the clear relationship between light and absolute IBMP concentration, established herein and elsewhere [21,22], highlights that bunch light exposure might remain an important tool for grapegrowers to regulate IBMP not only in berry but also in rachis.

Methoxypyrazine Distribution in Cabernet Sauvignon
Rachis. Preliminary experiments were conducted over the 2019/20 growing season to quantify IBMP distribution in the rachis of Cabernet Sauvignon clones (Reynella and SA125) grown on 110 Richter rootstock in Coonawarra (Table S5 of the Supplementary Material). Te concentration of IBMP in Cabernet Sauvignon rachis was signifcantly diferent between rachis components (P < 0.001) according to the LMM. Te estimated marginal means ranged from 52.4 ng/kg component for peduncle to 229 ng/kg component for pedicel ( Figure 6(a)). Pedicel had signifcantly higher concentrations than bottom rachis (140 ng/kg component) and top rachis (124 ng/kg component), which themselves were equivalent, and signifcantly higher than peduncle (52.4 ng/kg component). Additionally, IBMP concentrations in Cabernet Sauvignon rachis varied signifcantly with berry maturity (P < 0.001) (Figure 6(b)), from 148 ng/kg at veraison to 98 ng/kg at harvest. Tis suggests that IBMP concentrations in rachis are negatively correlated with berry maturity in a similar manner to IBMP in not only Cabernet Sauvignon berries [31] but also Shiraz rachis, as described in an earlier section.
Although  Figure S4 of the Supplementary Material). As the Cabernet Sauvignon and Shiraz samples were sourced from diferent regions, the diference in IBMP concentration in rachis seen herein was difcult to attribute solely to varietal diferences. Previous work has shown large variations in IBMP concentration for Shiraz rachis across multiple regions in a single vintage [21] and a regional infuence could not be discounted in the present study.
IBMP concentration in Cabernet Sauvignon rachis components was dependent on an interaction efect between clone and berry maturity (P � 0.006) with values for SA125 varying signifcantly (P ≤ 0.05) from veraison (155 ng/kg) to harvest (81.5 ng/kg) (Figure 6(c)). In comparison, Reynella did not vary signifcantly (P > 0.05) in IBMP concentration from veraison to harvest (141 ng/kg and 117 ng/kg, respectively). Tere was no statistical diference in IBMP between SA125 and Reynella rachis at harvest. Tis contrasted with previous fndings, in which IBMP concentrations in the berry of Carménère [33] and Sauvignon blanc [34] clones, and in the rachis of Shiraz clones 1654 and BVRC [21], varied signifcantly at harvest. Such variability is proposed to arise due to genetic variation between clones [34], but further research is required to understand the biological mechanism that leads to clonal variation in IBMP concentration in berry or rachis.
As vintage and rootstock have been shown to regulate MP accumulation in Cabernet Sauvignon rachis [22], further research encompassing these variables is necessary to support the preliminary trends in MP distribution observed in the present study with Cabernet Sauvignon rachis grown on 110 Richter over a single vintage. Furthermore, the elevated IBMP concentration in pedicel at harvest implies that the presence of Cabernet Sauvignon pedicels in a fermentor has the potential to increase the concentration of IBMP and alter wine sensory characteristics. Pedicel is the most likely MOG to enter a fermentor, so it would be interesting to determine the proportion of MOG attributable to pedicel as a result of diferent crushing and sorting techniques, to ascertain the likely impact on wine style.

Conclusion
Tis research signifcantly expands upon existing knowledge by showing that IBMP distribution throughout Shiraz rachis is not equivalent and can be signifcantly impacted by rootstock, stage of grape development, and vintage. As the concentration of IBMP was not signifcantly diferent between veraison and harvest in pedicel, the main contributor of overall rachis IBMP, quantifcation of IBMP in pedicel at veraison could inform winemakers about potential sensory outcomes related to rachis presence during fermentation. Tis trial also reinforced the importance of bunch light exposure throughout the growing season in mediating IBMP concentrations in rachis at harvest, providing knowledge that will be useful for managing IBMP in the vineyard or winery. It remains to be determined how the variation in IBMP across Shiraz rachis components occurs. Additional research considering the impact of berry shape and cluster compactness on pedicel light exposure throughout the growing season may help to further explain the elevated concentrations of IBMP in pedicel observed in the present study.
In addition, the impact of viticultural region remained unexplored and, in conjunction with assessing other rootstock and scion combinations, would be an opportunity for future research. Furthermore, a preliminary trial suggested that IBMP concentrations in Cabernet Sauvignon rachis may be afected by grape maturity and clone, which has implications for wine quality and style. Additional work could be useful to confrm these results and determine the impact of vintage and rootstock on the distribution of MPs across the bunch stem.
Putting the work into a practical context, in a best-case scenario, if only 1% w/w of MOG is permitted in top grade fruit with a maximum of 50% being rachis, such unintentional inclusion of rachis during fermentation would be insufcient to exceed the detection threshold of IBMP in red wine based on the results of this study. A direct efect on the sensory profle of wine would therefore be unlikely, even in the worst-case scenario of complete bunch light exclusion throughout the growing season. However, subthreshold concentrations of IBMP can alter the sensory profle of wine, so the presence of rachis components during fermentation should remain a primary consideration for winemakers.

Data Availability
Te data used to support the fndings of this study are included within the article and the supplementary information fle.