DEVELOPMENT OF CARBOHYDRATE ANALYSIS-BASED METHODS

FOR SUSTAINABLE WALNUT ORCHARD MANAGEMENT

Maciej Zwieniecki, Anna Davidson, Adele Amico Roxas

ABSTRACT

Tree growth and yield are dependent on a complex set of interactions involving genotype,

physiological and developmental processes, and the interaction of these processes with the

environment (abiotic and biotic stress). Non-structural carbohydrates (NSC; starch (ST) and

soluble carbohydrates (SC) like sucrose, glucose or fructose) are at the center of tree intrinsic and

extrinsic activities. As such, they are the key energy currency. NSC are produced during

photosynthesis and used to ‘pay’ for all biological services (growth, respiration, nutrient uptake,

defense against pathogens, reproduction, protection from abiotic stress). Evolved management of

NSC reserves is the key allowing the tree to survive adverse climatic conditions during vegetative

growth portion of the year (spring-summer-fall) and dormancy (winter). As the Central Valley

climate becomes more erratic and abiotic stresses more severe (summer like temperatures in the

fall and prolonged fall drought, loss of winter fog, large daily swings of temperature, and

increasing probability of winter frost), understanding the mechanisms responsible for tree energy

(NSC) management is crucial, yet still missing, is a piece of the puzzle to supplement existing

techniques (like nutrient analysis, water potential measurements) in mediating the impact of

current and future abiotic stresses and maintain high productivity.

Using a ‘citizen research’ approach, the Carbohydrate Observatory (CO), we are gaining a unique

insight to seasonal dynamics of NSC in walnut trees: (http://zlab-carb-

observatory.herokuapp.com). A large-scale approach and the ability to collect samples from a

diverse range of sites, tree ages, varieties and management practices provides opportunity to

determine the main factors linking productivity with NSC management. Currently, we have

determined that unlike almond and pistachio, walnut has a conservative strategy of NSC

management resulting in small swings of reserves between the summer productive phase and

dormancy. However, even these small variations seem important for yield capacity as indicated by

a significant positive correlation between NSC reserves in the winter and yield the following year.

Interestingly, we did not observe a significant negative correlation of NSC content in summer with

yield, as was observed in almond and pistachio. This suggests that trees with high NSC reserves

are always most productive.

In addition, analysis of the NSC seasonal pattern provides an opportunity to develop predictive

models of dormancy. The experimental approach to manipulate NSC availability in the fall by

defoliation and their redistribution by girdling revealed a complicated pattern that suggests a

significant impact of winter NSC storage on bloom success. In general, any disruption to natural

senescence and redistribution resulted in delayed bloom and reduced synchrony of the bloom.

However, the most pronounced effect was seen after early fall defoliation and girdling in the spring

– a situation that reduces NSC storage and does not allow for late spring retrieval of distally located

sugars. This insight provides prospect for the development of a model that uses climatic data and

trees’ NSC management to asses progression of dormancy and tree readiness for synchronous

bloom similar to an existing model for almond: http://zlab-chill-heat-model.herokuapp.com.

OBJECTIVES

Our main objective is to develop a carbohydrate analysis method as a tool to determine walnut’s

physiological status that would complement the currently used methods such as water potential

and nutrient analysis. The goal is to use carbohydrate analysis as a new option for sustainable

orchard management. The main objective represents a long-term goal with sub-yearly objectives

representing steps. Sub-objectives for the reported year were:

(1) Continue to maintain and possibly expand a network of walnut orchards to provide samples

to a large-scale, state-wide study of seasonal dynamics of carbohydrate using a ‘citizen

research approach’ that minimizes time and costs for research.

(2) Provide easy and informative access to NSC information for walnut growers by further

improvements to our online platform: http://zlab-carb-observatory.herokuapp.com

(3) Describe and publish seasonal patterns of NSC dynamics within the walnut crown that

links NSC management with tree phenology.

(4) Use data already collected to determine impact of NSC on yield.

(5) Experimentally test the role of NSC reserve levels and their redistribution on bud

development and bloom time.

SIGNIFICANT FINDINGS

(1) We retain a number of orchards that provide samples for analysis across a wide range of

climatic conditions although continuation of sampling from specific orchards was affected

by resent economic perturbances. We further improve and streamline our protocol for

sample analysis to determination content of soluble sugars and starch in walnut wood and

bark. Again, our methodology allows for cost effective large-scale analyses of

carbohydrate that provide means for a citizen research approach. In-house cost of analysis

is under $8 per sample compared to over $100 in commercial analytical labs.

(2) Online platform for sharing information on near real time of NSC content in walnut trees

(http://zlab-carb-observatory.herokuapp.com) has undergone significant improvement.

This included new capacity to compare orchard performance with state average and other

specific farms. In addition, it is now possible to compare NSC content in different

rootstocks, scions, age groups and geographical locations at the level of counties.

(3) A manuscript describing the impact of NSC management on tree phenology is under its

second round of revisions in Scientific Reports: Comparison of phenological traits, growth

patterns, and seasonal dynamics of non-structural carbohydrate in Mediterranean tree

crop species. 2019. Aude Tixier, Paula Guzmán-Delgado, Or Sperling, Adele Amico

Roxas, Emilio Laca, Maciej Zwieniecki. Scientific Reports (in revision).

(4) An analysis of NSC content in twigs was used to determine their impact on yield. Findings

suggest a significant positive correlation between NSC content in January and March on

following year yield. Surprisingly and unlike other nut species, no post-harvest NSC

exhaustion was found suggesting a highly conservative reproductive pattern.

(5) Detailed determination of temporal dynamics of NSCs content in walnut trees suggests that

early spring and late fall are the most significant periods affecting NSCs content and

redistribution. During late fall, walnut trees restore levels of NSCs required to survive

winter. During the spring, carbohydrates are mobilized for bud break. Experimentally

induced perturbances to these patterns significantly delayed or even halted flower and bud

development. Both early defoliation and girdling of phloem reduced bloom capacity and

delayed bloom.

PROCEDURES

The Carbohydrate Observatory is the research initiative providing analytical service to growers

interested in a better understanding of NSC management of their orchards. Growers provide three

twig samples (with xylem and phloem separated) per orchard that are subsequently analyzed for

soluble sugars and starch content in wood xylem – water conducting tissue) and bark (phloem –

sugar conducting tissue). Analytical results are being published on line at http://zlab-carb-

observatory.herokuapp.com.

Received samples are processed in the lab following the procedure described below:

• Each sample is ground into powder. A small amount (25 mg) is then washed in 1 mL of

pH buffer to dissolve all soluble carbohydrates.

• Using a colorimetric method (a spectrophotometer), the concentration of sugars is

measured (Anthrone method) in sample of buffer (50 uL) and recalculated to express SC

concertation per g of dry matter.

• The remaining material in the buffer is treated with two different enzymes that digest starch

to form the soluble sugars. These are again measured in a spectrophotometer using the

Anthrone method.

• For further details of the procedure please refer to published articles by the Zwieniecki lab,

on our website or request the procedure vial email (mzwienie@ucdavis.edu).

Additional information related to each orchard is provided by growers on voluntary bases that

includes specific management practices, age, scion/rootstock combination and yield. This part

needs improvement due to a low participation from walnut growers. The quality of analysis

requires a high number of samples. Our lab is currently communicating with many growers to

explain the reasoning for the need of additional data. In addition, CIMIS and NOAA weather data

is used in the analysis. The additional site related information is crossed referenced with the NSC

analysis database and used in subsequent analyses of the role of NSCs in annual orchard

performance.

In winter of 2018/2019, a field experiment on the impact of NSC content and redistribution was

developed to study synchrony and timing of bloom in walnut. The study was performed during

2018-2019 in the experimental orchard owned by University of California, Davis (38°22’ N,

121°47’ W). For this study, we selected 20 homogenous mature walnut (Juglans regia L.) trees.

The experiment started in October and ended in May, to enable us to study the pre-dormancy,

dormancy and bud-break stages. Also, climatic data were collected from a data logger located in

the experimental site.

On the twenty selected trees, we applied experimental treatments. We defoliated the trees in

October and November to large limbs and girdled smaller branches three times (October,

December and April). Defoliation x girdling interaction treatments were also applied. Control

Branches were left intact. Girdling was done by removing ~1 cm of bark at the base of twigs using

a dull razorblade. The exposed xylem section was covered with Parafilm in order to avoid

desiccation. Every month we checked that the Parafilm was still in good condition. The sample

collection for carbohydrates measurement (NSC) was performed monthly starting October and

concluded in May. During every sample collection, twigs from the experimental orchard were

quickly transported to the laboratory. Bark was removed from ~3 cm of the twig and fresh weight

of the bark and wood was weighed. Samples were then dried at 75°C for 48 hours. Once dry,

samples were weighed; then ground into a fine and homogeneous powder using a ball mill for

NSC quantification. Soluble carbohydrates (SC) were extracted by incubating 25 mg of dry

material in 1 mL of sodium acetate buffer (0.2 M and 5.5 pH) for 15 minutes at 70 °C followed by

centrifugation (10 minutes at 21,000 g). The supernatant was diluted in deionized water (1:20,

v/v); the remaining pellet with the buffer was kept for starch analysis. The Pellet was incubated

for 10 min at 100°C to allow starch gelatinization and then digested with 0.7 U of amylase and 7

U of amyloglucosidase for 4 hours at 37 °C. Once the digestion was finished, samples were

centrifuged for 10 minutes at 21,000 g. The supernatant was diluted in deionized water (1:20, v/v).

SC and starch were quantified at the end of the extraction using Anthrone as a reagent (0.1 % (m/v)

in 98 % sulfuric acid) by reading absorbance at 620 nm (Leyva et al., 2008). In order to measure

the carbohydrate content in leaves, weekly we randomly selected five trees in October and

November. Leaf collection ended in the middle of November.

Phenological observation and bud measurement were performed on five randomly selected trees.

The phenology of walnut buds was described in the following stages: T0= dormant bud, T1= bud-

break (swollen bud with visible leaf primordial), T2= visible petiole and leaflets, T3= expanded

leaves (unfolded leaves but not yet photosynthetically active) (Fig. 1). Phenological observations

started before the bud-break in early April. Pictures were taken every ~4 days. Once the buds

reached stage T3 twigs were harvested for the last sample collection. At every twig collection we

also measured the weight of the buds (top and lateral position).

Figure 1. Phenological

stages of walnut

All data were analyzed by using the R package (R core team, 2003). For the analysis of NSC

concentration, phenological development and bud weight over the season a linear mixed effect

model was used with defoliation treatments (October and November), girdling treatments

(October, December, and March) and interaction (defoliation x girdling) as fixed factors and trees

as a random factor. Tukey’s HSD tests were performed on each model to separate means when ANOVA results were significant (P< 0.05).

RESULTS

Determination of carbohydrate content in walnut trees in relation to yield management by the

walnut tree: Over the period of 2016-19 we received and collected samples from over 100 orchards

resulting in 8000 analysis. The data is presented on the website (Figure 2) for orchard owners and

managers to review the performance of their sites. The website allows for comparative analysis of

multiple orchards, NSC content to each other and against additional parameters that include:

rootstock, scion, age, and county locations. Permanently displayed values are: running average of

the NSC content for the entire state and all data points collected/analyzed so far. Features include

the capacity to zoom to any portion of the graph. In addition, it is possible to look at a specific

type of NSC i.e. soluble sugars and starch across wood and bark.

Figure 2. Snapshot of the website allowing for comparative analysis of NSC in walnut trees.

This level of insight allows individual growers participating in the study to compare their orchards against

all specific management practices they use or orchard properties that were not revealed to us and make

future decisions on how to explore the data.

Analysis of carbohydrates in bark and wood reveal that major swings in the NSC content occur in

wood. Specifically, there is a relatively slow but progressive use of reserves from maximum

content in January until July, then content is constant until harvest and the sudden recovery is

observed between harvest and leaf senescence. These findings in association with NSC content in

the fall and yield (see below) underlines the necessity of post-harvest management aiming at

recovery of NSC reserves prior to dormancy.

A small number of farms that revealed their yields to us gave us the first insight to determine the

role of branch NSC in crop performance (Fig. 3). Higher number of yield reports would improve

this portion of research exploration and possibly provide further support to presented results.

However, even with limited information, we can provide insight that suggests a positive correlation

between NSC content in late fall and in the spring preceding bloom with the following summer

yield. This is an important finding that suggest potential use of the fall information as a tool to

either predict future yield or apply fall management aiming at restoration of high NSC content in

trees.

Figure 3. Shown are respective slopes of correlation between carbohydrate content and yield (coefficient

of linear function formulated as: yield= coefficient*NSC Type + offset). Region ‘A’ shows a generally

positive correlation (coefficient>0) of winter NSC levels with the following summer yield. Region ‘B’

shows coefficients in summer as generally not correlated with the current year’s yield. ‘*’ denote

significance with p<0.05 for the correlation for at least one NSC type.

How NSC reserves impact yield is not yet known. This might happen via multiple pathways. (1)

Survival of dormancy by providing energy to maintain high vitality of meristems (both apical in

buds and secondary in cambium); (2) Recovery of xylem transport capacity by providing energy

for restoration of hydraulic continuity in xylem vessels – i.e. generation of positive pressure in

xylem during spring; (3) assuring a high level of reserves at bloom time thus, providing energy to

develop healthy flowers and sustain fast growth of new leaves before they achieve photosynthetic

independency. An experiment to determine if indeed manipulation of NSC content prior to

dormancy and their redistribution, suggests that there is a significant impact on bloom time and

synchrony of bloom. Specifically, we have found several interesting aspects of both branch

defoliation (aiming at the reduction of NSC content in branches) and girdling aiming at reducing

the redistribution of NSC potential that shows unique aspects of walnut biology.

Defoliation surprisingly, resulted in higher content of NSC in wood and bark late in winter

(January) then in control branches, suggesting that winter redistribution of NSC is very active in

walnut trees, equalizing NSC access across a tree (Figure 4). This is especially evident if we

compare the content of defoliated branches with branches where redistribution was affected by

girdling of the phloem. In general, girdling resulted in a lower content of NSC and smaller if not

existing peak of NSC in January (Figure 5). April NSC content in wood and bark of defoliated

branches was on average ~130 and ~110 NSC mg/DW respectively, and in girdled branches ~110

and 75 NSC mg/DW in wood and bark. This translates to much lower availability of NSC during

bloom time in branches with reduced reallocation of NSC. Impact of both defoliation and girdling

was similar to the impact of girdling only, thus again, suggesting that redistribution of NSC in the

winter is very active and can control levels of NSC availability during spring.

Figure 4. Impact of

defoliation time on NSC

content in walnut twigs.

NSC – non-structural

carbohydrates, SC –

soluble sugars, St - starch

Figure 5. Impact of

girdling time on NSC

content in walnut twigs.

NSC – non-structural

carbohydrates, SC –

soluble sugars, St - starch

Early defoliation despite the insignificant impact on NSC content in branches leads to significant

delay and asynchrony of bud break (Figure 6). At the same time, the impact of girdling (i.e.

affecting the redistribution of the NSC) was observed only for late (March) girdling (Figure 7).

This suggests that the impact of spring supply and redistribution is most important for synchrony

of bloom, as suggested before (see Tixier, A., Sperling, O., Orozco, J., Lampinen, B., Roxas, A.A.,

Saa, S., Earles, J.M., Zwieniecki, M.A. 2017). Spring bud growth depends on sugar delivery by

xylem and water recirculation by phloem Münch flow in Juglans regia. Planta 246:495-508). In

addition, any combination of defoliation with girdling leads to delayed bud break and significant

asynchrony of bloom.

Figure 6. Impact of defoliation on bloom time.

Early (October) defoliation significantly

delayed the bud break and bloom asynchrony.

Figure 7. Impact of girdling on bloom time.

Late (March) girdling lead to a significant

delay in bud break and bloom asynchrony.

Grower implications:

• Twigs NSC content in the fall is positively correlated with yield in the following summer.

This implies that post-harvest management aiming at the reconstitution of NSC content can

improve orchard yield performance.

• Use of carbohydrate analysis can significantly improve yield prediction models.

• Fall practices interrupting NSC accumulation or climatic impacts on the accumulation of

NSC, as well as disruption to phloem activity during dormancy, can lead to a significant

delay in budbreak time and asynchronous bloom. Specifically, early leaf drop can

negatively delay budbreak.

• NSC dynamics in the winter and NSC impacts on budbreak timing can be used in models

assessing the progression of dormancy and help to provide insights to the use of

budbreaking chemicals.

Specific findings are published in the following papers:

Tixier, A., Amico Roxas, A., Godfrey, J., Saa, S., Lightle, D., Maillard, P., Lampinen, B.,

Zwieniecki, M.A. 2017. Role of Bark Color on Stem Temperature and Carbohydrate Management

during Dormancy Break in Persian Walnut (Juglans regia L.). Journal of American Society for

Horticultural Science 142(6):454–463. doi: 10.21273/JASHS04216-17

Abstract: Temperature is assumed to be the principal regulatory signal that determines the end of

dormancy and resumption of growth. Indirect evidence that stem temperature interferes with

phenology comes from the common orchard practice of painting stems to protect them from

disease. This work studies the effects of the application of white paint to the stems of persian

walnut (Juglans regia) trees on winter stem temperature, carbohydrate content, and spring

phenology. Painting bark resulted in the delay of budbreak by several days, higher nonstructural

carbohydrate (NSC) concentrations in the bark and wood of painted extension shoots and changes

in the spatial gradients of NSC during budbreak. The demands of maintenance respiration

exceeded mobilization from local carbon pools during bud development suggesting a potential role

of carbohydrate transport during spring budbreak in persian walnut. Painting provides an exciting

perspective for mitigating effects of milder winter in orchards. The effect of reducing diurnal and

spatial temperature variability limits early budbreak, NSC depletion associated with intense

maintenance respiration, freeze–thaw cycles and frost dehardening.

Tixier, A., Sperling, O., Orozco, J., Lampinen, B., Roxas, A.A., Saa, S., Earles, J.M., Zwieniecki,

M.A. 2017. Spring bud growth depends on sugar delivery by xylem and water recirculation by

phloem Munch flow in Juglans regia. Planta 246:495-508

Abstract: Main conclusion: During spring, bud growth relies on long-distance transport of

remotely stored carbohydrates. A new hypothesis suggests this transport is achieved by the

interplay of xylem and phloem. During the spring, carbohydrate demand of developing buds often

exceeds locally available storage, thus requiring the translocation of sugars from distant locations

like limbs, stems and roots. Both the phloem and xylem have the capacity for such long-distance

transport, but their functional contribution is unclear. To address this ambiguity, the spatial and

temporal dynamics of carbohydrate availability in extension shoots of Juglans regia L. were

analyzed. A significant loss of extension shoot carbohydrates in remote locations was observed

while carbohydrate availability near the buds remained unaffected. This pattern of depletion of

carbohydrate reserves supports the notion of long-distance translocation. Girdling and dye

perfusion experiments were performed to assess the role of phloem and xylem in the transport of

carbohydrate and water towards the buds. Girdling caused a decrease in nonstructural carbohydrate

concentration above the point of girdling and an unexpected concurrent increase in water content

associated with impeded xylem transport. Based on experimental observations and modeling, we

propose a novel mechanism for maintenance of spring carbohydrate translocation in trees where

xylem transports carbohydrates and this transport is maintained with the recirculation of water by

phloem Munch flow. Phloem Munch flow acts as a pump for generating water flux in xylem and

allows for transport and mobilization of sugars from distal locations prior to leaves photosynthetic

independence and in the absence of transpiration.

Predicting bloom dates by temperature mediated kinetics of carbohydrate metabolism in deciduous

trees. 2019. Or, S., Kamai, T., Tixier, A., Davidson, A., Jarvis-Shean, K., Raveh, E., DeJong, T.,

Zwieniecki, M.A. Agricultural and Forest Meteorology (in press):

https://doi.org/10.1016/j.agrformet.2019.107643

Abstract: Trees in seasonal climates gauge winter progression to assure vital and productive

blooming. However, how dormant plants asses environmental conditions remains obscure. We

postulated that it involves the energetic reserves required for bloom, and therefore studied winter

carbohydrate metabolism in deciduous trees. We quantified non-structural carbohydrates

throughout winter in almond, peach, and pistachio trees in California and Israel and characterized

winter metabolism. We constructed a carbohydrate-temperature (C–T) model that projects changes

in starch and soluble carbohydrate concentrations by temperature mediated kinetics. Then, we

tested the C–T model projections of bloom times by 20 years of temperature and phenology records

from California. The C–T model attributes winter carbohydrate regulation in dormant trees to

continuous updates of metabolic pathways. The model projects a surge in starch synthesis at the

end of winter, and critically low concentrations of soluble carbohydrates, that trigger bloom. This

is supported by field measurements of starch accumulation at the end of winter (˜50 mg g−1 DW

in almonds) that preceded bloom by ̃ 10 days. The C–T model provides a physiological framework

for bloom forecasts in deciduous orchards. It integrates contrasting notions of chill and heat and

elucidates why abnormal winter temperatures may compromise bloom in deciduous orchards.

Comparison of phenological traits, growth patterns, and seasonal dynamics of non-structural

carbohydrate in Mediterranean tree crop species. 2019. Aude Tixier, Paula Guzman Delgado, Or

Sperling, Adele Amico Roxas , Emilio Laca , Maciej Zwieniecki. Scientific Reports (in second

revision)

Abstract: Despite non-structural carbohydrate (NSC) importance for tree productivity and

resilience, little is known about their seasonal regulations and trade-off with growth and

reproduction. We characterize the seasonal dynamics of NSC in relation to the aboveground

phenology and temporal growth patterns of three deciduous Mediterranean species: almond

(Prunus dulcis (Mill.) D. A. Webb), walnut (Juglans regia L.) and pistachio (Pistacia vera L.).

Seasonal dynamics of NSC were synchronous between wood tissues from trunk, branches and

twigs. Almond had almost identical levels and patterns of NSC variation in twigs, branches and

trunks whereas pistachio and walnut exhibited clear concentration differences among plant parts

whereby twigs had the highest and most variable NSC concentration, followed by branches and

then trunk. While phenology had a significant influence on NSC seasonal trends, there was no

clear trade-off between NSC storage and growth suggesting that both were similarly strong sinks

for NSC. A temporal trade-off observed at the seasonal scale was influenced by the phenology of

the species. We propose that late senescing species experience C allocation trade-off at the end of

the growing season because of C-limiting thermal conditions and priority allocation to storage in

order to survive winter.