Dynamic Con olled A mosphe e (DCA) s o age o
Con e ence pea e ains backg ound colo be e han
no mal CA s o age
H.M. Phan
1
, B.E. Ve linden
2a
, M. He og
1
, P. Ve bo en
1
, and B.M. Nicolai
1,2
1 KU Leu en, BIOSYST-MeBioS, Leu en, Belgium; 2 Flande s Cen e o Pos ha es Technology, Leu en, Belgium.
a e-mail: be . e linden@ cb .be
Abs ac
G een backg ound colo is a majo quali y a ibu e o ‘Con e ence’ pea (Py us
communis) de e mining i s comme cial alue. P ope con olled a mosphe e s o age
can slow down yellowing o he ui peel o se e al mon hs. A s o age expe imen
was ca ied ou wi h ‘Con e ence’ pea s, ha es ed in Sep embe 2023. Pea s we e
s o ed in CA condi ions o 3 kPa O
2
and 0.7 kPa CO
2
as well as dynamic con olled
a mosphe e (DCA) condi ions. When no mal CA condi ions we e eached, he oxygen
concen a ion was allowed o u he dec ease based on RQ- alues measu ed on he
whole s o age olume. Tes s we e ca ied ou a ha es and a e 4 and 8 mon hs o
s o age and shel -li e a 18°C. The backg ound colo was measu ed using a colo
spec ome e as well as wi h an image analysis sys em as CIE Lab alues 8 imes
sp ead o e 12 days. Besides hese expe imen s in lab condi ions, ials in an
indus ial se ing we e ca ied ou . Colo was measu ed du ing shel -li e a e
ha es and a e 5 mon hs s o age. The esul s o he lab expe imen s showed ha
he g een backg ound colo o ‘Con e ence’ pea was be e e ained when s o ed in
DCA condi ions compa ed o classical CA s o age. I was obse ed ha in he
indus ial ial i was mo e di icul o each lowe oxygen condi ions mainly because
he ooms we e no ai igh enough. Howe e , signi ican lowe oxygen condi ions
han no mal CA condi ions we e eached esul ing in be e quali y e en ion.
Keywo ds:
CA, RQ-DCA, Quali y change, Backg ound colo , Image analysis
INTRODUCTION
Nex o ui i mness, ui backg ound colo is a majo comme cial quali y a ibu e
o ‘Con e ence’ pea (
Py us communis
L.). Consume s, especially in No he n Eu ope,
p e e a g een backg ound colo deno ing a esh like appea ance. Once ha es ed, esh
ui unde go con inuous deg ada ion in quali y due o a ious oxida i e p ocesses, such
as espi a ion and e hylene p oduc ion, occu ing wi hin he cells. Pos ha es handle s
a emp o limi hese eac ions implica ed in quali y b eakdown by s o ing esh ui a
low empe a u e in con olled a mosphe e wi h educed O
2
and ele a ed CO
2
concen a ions, hence p olonging hei s o age li e (Kade , 1986). Du ing s o age, he
g een skin backg ound colo u ns yellow as he chlo ophyll pigmen s wi hin he skin a e
b oken down o colo less p oduc s, and in he p ocess he unde lying yellow pigmen s
a e unmasked (Mülle e al., 2007; K äu le , 2008). P ope con olled a mosphe e s o age
can slow down so ening and yellowing o he ui peel o se e al mon hs. The O
2
concen a ions, howe e , should no go below he sa e ma gin o he anae obic
compensa ion poin (ACP) o p e en igge ing e men a ion leading o he p oduc ion
o o - la ou s and s o age diso de s (F anck e al., 2007). Addi ionally, many pea
cul i a s, such as ‘Con e ence’, a e highly suscep ible o high CO
2
concen a ions causing
in e nal b owning, while low O
2
concen a ions o en esul in ca i ies. As he ACP may
a y wi h season, o igin and picking da e, a sa e, bu likely subop imal oxygen
concen a ion, is o en main ained in CA s o age, esul ing in mo e quali y loss han
expec ed. Dynamic con olled a mosphe e (DCA) s o age sea ches o he lowe oxygen
limi , by adap ing oxygen le els dynamically as a unc ion o s o age ime, based on a
measu emen o a biological esponse o he s o ed ui o low oxygen le els (Zanella,
2003). As oxygen le els each lowe le els, me abolism slows down and hea o
espi a ion p oduced by he ui dec eases esul ing in less hea needed o be emo ed
by he cooling sys em. This may esul in ene gy sa ings o s o age up o 10% (Ve linden
e al., 2023).
The objec i e o his wo k was o demons a e ha RQ-DCA s o age can main ain
be e quali y ui o longe s o age imes compa ed o he classical CA s o age o
‘Con e ence’ pea . Besides se ing up expe imen s in he lab whe e condi ions can be
s ic ly con olled, ui was also s o ed a RQ-DCA condi ions in a comme cial s o age
plan o assess pe o mance in non-lab condi ions. Peel backg ound colo was used as he
quali y a ibu e o in e es . Two colo measu emen me hods, one based on poin
measu emen using a colo spec ome e and a second one based on image analysis o
pho og aphs aken in a con olled se up we e compa ed.
MATERIALS & METHODS
Pea ui s o age and shel -li e assessmen
The ‘Con e ence’ pea ui we e ha es ed on Sep embe 5, 2023 a op imal
comme cial ipeness in Bie beek, Belgium and s o ed a -0.6 °C in egula ai o h ee
weeks. A e his cooling pe iod, he pea s we e ans e ed in o ai - igh polyp opene
ui s o age con aine s o abou 300 L. Wi hin he con aine s, he a mosphe e was
con olled. In he CA ea men , he a mosphe e was con olled a 3 kPa oxygen and 0.7
kPa ca bon dioxide. Fo he RQ-DCA ea men , he ca bon dioxide concen a ion was
main ained a 0.7 kPa, while he oxygen concen a ion was dynamically adjus ed ( ange
be ween 3 and 0.3 kPa) acco ding o au oma ed RQ measu emen s on he whole ui
s o age con aine (Bessemans e al., 2016). When no mal CA condi ions we e eached,
he oxygen concen a ion was allowed o u he dec ease based on RQ- alues measu ed
on he whole s o age olume. As long as RQ alues we e below 1.3, oxygen concen a ions
we e educed. When RQ alues inc eased o 1.3 o abo e, he oxygen concen a ion was
inc eased again.
Pea samples om he same ha es ba ch we e also s o ed in a comme cial CA-
oom oge he wi h he pea s o he same o cha d. In he comme cial oom, pea s we e
also kep in egula ai o 3 weeks be o e slowly es ablishing CA condi ions o 3 kPa
oxygen and 0.7 kPa ca bon dioxide. A 95 days a e ha es RQ-DCA con ol was s a ed.
The ealised gas condi ions can be ound in Figu e 1.
Quali y o ui s o ed in lab con aine s was assessed a ha es and a e 4 and 8
mon hs o s o age du ing a shel -li e pe iod in egula ai (RA) a 18°C. Measu emen s
we e ca ied ou a ew hou s a e aking he ui ou s o age, and 7 addi ional
measu emen s sp ead o e a pe iod o 12 days. Quali y o ui s o ed in he comme cial
DCA- oom was assessed a e 5 mon hs o s o age.
F ui quali y measu emen s
Assessmen o peel backg ound colo was ca ied ou a ha es and a e 4 and 8
mon hs o s o age. A ha es , 30 ui s we e measu ed o each assessmen du ing shel -
li e. A e s o age 15 ui we e measu ed o each s o age condi ion a each assessmen
o shel -li e.
Colo measu emen
The colo o he ui was measu ed using a CM-2600d spec opho ome e (Konica
Minol a Sensing L d., Singapo e). Fo each ui , he skin backg ound colo was measu ed
i e imes on g een po ions as CIE Lab alues. The alues we e a e aged o ob ain he
mean alue o ha ui .
Pic u es o pea ui we e aken using a digi al came a se up in small cabine wi h
s anda d LED ligh ing and ixed dis ances. Using image analysis, backg ound colo o he
pea ui was segmen ed om he es o he pic u es. The RGB alues o he pixels
segmen ed as pea backg ound colo we e con e ed in o CIE Lab alues and a e aged
pe pea .
S a is ical analysis
To deno e s a is ical di e ences a e 4 and 8 mon hs o s o age, a wo-way ANOVA
was ca ied ou compa ing days o shel -li e and s o age condi ion (CA o DCA). A ull
ac o ial design was applied using JMP®, Ve sion 17. SAS Ins i u e Inc., Ca y, NC, 1989–
2023. To judge di e ences be ween s o age condi ions a speci ic shel -li e days, Tukey’s
HDS mul i compa ison es was applied.
Figu e 1. Realized s o age condi ions in he comme cial s o age oom ( op le ), he lab
s o age con aine wi h cons an CA condi ion, 3 kPa O2 and 0.7 kPa CO2 ( op igh ),
and he wo DCA con aine s wi h a ying O2 condi ions (bo om). Tempe a u e
was kep a ound -0.6°C.
RESULTS AND DISCUSSION
Colo change du ing CA and DCA s o age
A e ha es , ‘Con e ence’ pea s we e, i s , cooled down and kep in RA o 3 weeks.
This is a s anda d p o ocol ha will p e en de elopmen o co e b owning and ca i y
0
2
4
6
8
O
2
(kPa)
Comm.DCA
Lab CA
Lab DCA1
Lab DCA2
o ma ion la e in s o age. A e hese i s 3 weeks, a i s assessmen o he backg ound
colo was made and CA o DCA s o age p o ocols s a ed. The ealised O
2
and empe a u e
condi ions can be seen in Figu e 1. In he lab, pea s we e s o ed up o 268 days a e
ha es (8 mon hs) in CA and in RQ-DCA condi ions. The s o age in comme cial RQ-DCA
condi ions las ed only 182 days (5 mon hs) as he owne o he ui decided o open he
oom and ma ke he ui . While in lab condi ions, i was ai ly easy o s a he DCA
p o ocol a e eaching he classical CA condi ions (
3 kPa O2 and 0.7 kPa CO2
), i ook some
con incing and unexpec ed p ac ical p oblems be o e his could be s a ed in he
comme cial oom. E en ually he RQ-DCA p o ocol was s a ed a ound day 95 a e
ha es . As can be seen in Figu e 1 he O
2
concen a ion g adually dec eased un il le els
o abou 1.3 kPa while, in lab condi ions, he lowe limi se by he sys em o 0.3 kPa was
eached a abou day 80 and main ained o he es o he s o age pe iod. The eason why
he comme cial oom did no each he same low O
2
alues as he lab DCA con aine s was
due o he ela i e leakiness o he comme cial oom. The RQ measu emen s in he
comme cial oom we e s ill a ound 1 deno ing ae obic espi a ion condi ions bu he
lowe ed espi a ion o he ui was no able o compensa e he inc ease in O
2
concen a ion esul ing om leakage. A he end o he s o age, backg ound colo was
measu ed again as can be seen in Figu e 2. The le pane shows he a* alues 21 days a e
ha es and a he end o s o age o he 3 s o age condi ions. Lowe nega i e a* alues
co espond o g eene colo while highe , less nega i e alues co espond o mo e yellow
colo s. The Lab DCA s o age was able o main ain he same alues as a ha es while he
CA s o age showed yellowing. Comme cial DCA alues we e in e media e, as expec ed,
because a la ge pa o he s o age was ac ually cons an CA. In he igh pane,
co esponding Hue alues a e shown. A highe Hue alue co esponds wi h a g eene
colo , while a low Hue alue is mo e yellow. The same ends we e seen he e. Pea s om
CA s o age we e he mos yellow a e s o age ollowed by he comme cial DCA. The Lab
DCA was able o main ain simila alues as a ha es .
Figu e 2. Colo change o backg ound colo du ing s o age in cons an CA condi ion, 3 kPa O2
and 0.7 kPa CO2 (blue), DCA in he lab ( ed) and DCA in comme cial en i onmen .
Colo measu ed wi h a Minol a spec ome e exp essed as a* (le ) and Hue alues
( igh ). Ba s deno e s anda d e o s.
Colo change du ing shel -li e a e s o age
A e 4 and 8 mon hs o s o age in he lab con aine s, samples we e aken ou and
pu in o shel -li e condi ions in RA a 18°C. Peel backg ound colo was measu ed 8 imes
0 50 100 150 200 250 300
days a e ha es
-7.8
-7.6
-7.4
-7.2
-7
-6.8
-6.6
a*
Lab CA
Lab DCA
Comm.DCA
0 50 100 150 200 250 300
days a e ha es
100.5
101
101.5
102
102.5
103
103.5
Hue (°)
Lab CA
Lab DCA
Comm.DCA
du ing 12 days, bo h wi h he colo spec ome e and image analysis sys em. Resul s a e
shown in Figu e 3. Un o una ely, measu emen o colo spec ome e du ing he i s
week o shel -li e a e 4 mon hs o s o age we e los due o echnical p oblems. As was o
be expec ed, colo change du ing shel -li e was much highe han he change du ing he
whole 8 mon hs o (D)CA s o age. As can be seen in Figu e 2, he colo change du ing 8
mon hs in no mal CA, he wo s pe o ming s o age condi ions, was less han 1 a* uni
and abou 1.5° o Hue, while du ing 12 days o shel -li e he colo changed 5 a* uni s and
15° o Hue. These changes co esponded wi h he peel backg ound colo going om
esh g een o ipe yellow. As he change wi h shel -li e du a ion was e y signi ican ,
nei he pea s s o ed in no mal CA no RQ-DCA showed any signi ican di e ence a e 4
mon hs o s o age (Figu e 3 le panes). A e 8 mon hs o s o age, signi ican e ec o
s o age condi ion we e de ec ed (Figu e 3 igh panes). Hue alues o DCA s o ed pea s
(dashed lines) s a ed wi h highe alues, meaning mo e g een colo han CA s o ed pea s
and consis en ly kep highe alues un il 6 days (day 274 a e ha es ) o shel -li e.
Simila ly, a* alues o DCA s o ed pea s we e consis en ly lowe and so mo e g een han
CA s o ed pea s.
Figu e 3. Colo change du ing shel -li e in RA a 18°C o backg ound colo a e 4 (le column)
and 8 mon hs ( igh column) s o age in cons an CA condi ion, 3 kPa O2 and 0.7 kPa
CO2 ( ull lines) and DCA (dashed lined). The op ow con ains he measu emen s
ca ied ou wi h a Minol a spec ome e and he bo om ow shows he esul s
measu ed wi h he image analysis sys em. Ba s deno e s anda d e o s.
-8
-6
-4
-2
0
2
4
a*
shel li e a e 4 mon hs s o age - Minol a
shel li e a e 8 mon hs s o age - Minol a
shel li e a e 4 mon hs s o age - Colou Lab
shel li e a e 8 mon hs s o age - Colou Lab
Figu e 4 shows he pho og aphs aken wi h he image analysis sys em igh a e 8
mon hs s o age ( op ow) and on day 6 in shel -li e (bo om ow). A he s a o shel -li e,
i was no ob ious o isually de ec di e ences o he peel backg ound colo be ween he
wo s o age condi ions, bu a e 6 days, di e ences we e ob ious as was con i med by a
mul i compa ison es o bo h he a* and Hue alues measu ed wi h bo h me hods.
S a shel -li e a e 8 mon hs CA s o age
S a shel -li e a e 8 mon hs DCA s o age
Day 6 shel -li e a e 8 mon hs CA s o age
Day 6 shel -
li e a e 8 mon hs DCA
s o age
Figu e 4. Pho og aphs o ‘Con e ence’ pea s s o ed o 8 mon hs in CA and DCA condi ions
igh a e s o age and a e 6 days o shel -li e in RA a 18°C.
Compa ing colo measu emen using spec ome y and image analysis
Measu ing colo o ood objec i ely is no s aigh o wa d (MacDougall, 2002). The
pe cep ion o colo by he eye no only depends on he ‘objec i e’ colo o he pigmen s
bu also on ligh en i onmen o he angle wi h which we look o he colo ed su ace and
physical p ope ies o ha su ace ha de e mine how he ligh will be e lec ed. To
measu e colo o ui objec i ely, ligh and angle o obse a ion mus be s anda dised.
This was he case in bo h colo measu emen me hods used in his s udy bu no in he
same way, which is pa o he explana ion o why he wo me hods ga e di e en alues
o he same ui . As can be seen in Figu e 3, he spec ome e measu emen esul ed in
lowe alues o Hue angle, abou 25 deg ees, and highe alues o a°, abou 10 uni s. As
can be judged om he e o ba s, he a iabili y and noise on he wo measu emen
me hods was abou he same. Bo h colo measu emen me hods whe e able o
demons a e he same di e ences a ibu able o s o age condi ions. Howe e , he wo
me hods showed a signi ican o se in colo alues (abou 5 a* uni and 15° Hue) ha was
la ge compa ed o he di e ence be ween he wo s o age condi ions o 1 a* uni and 5°
Hue. This migh be u he explained by he pa icula peel ea u es o ‘Con e ence’ pea .
‘Con e ence’ pea has ough b own egions on hei skin which a e some imes di icul o
a oid using he spec ome e measu emen me hod as i measu es spo s o a ew
millime e in diame e . The image analysis sys em on he o he hand can easily exclude
he b ownish pixels om he g een peel.
CONCLUSIONS
Two colo measu emen echniques, one based on spec ome e measu emen s and
ano he on image analysis, we e used o assess he colo o ‘Con e ence’ pea s a e
s o age and subsequen shel -li e. Bo h me hods we e able o gene a e colo da a o he
same s a is ical quali y. Howe e , he image analysis me hod was belie ed o esul in
colo alues ma ching be e he ue colo o he un ough egions o he pea skin while
hose o he spec oscopic me hod we e comp omised by he colo o he ough egions
ypical o ‘Con e ence’ pea .
The esul s o he lab expe imen s showed ha he g een backg ound colo o
Con e ence pea was be e e ained when s o ed in RQ-DCA condi ions compa ed o
classical CA s o age. I was obse ed ha in he indus ial ial i was mo e di icul o
each lowe oxygen condi ions mainly because he ooms we e no ai igh enough.
Howe e , signi ican ly lowe oxygen condi ions han no mal CA condi ions we e eached
esul ing in be e quali y e en ion.
Adop ion o DCA echnology by indus y should be encou aged as be e quali y
e en ion would lead o less ui loss and lowe espi a ion hea p oduc ion leading o
lowe ene gy consump ion du ing s o age. Bo h ou comes a e bene icial o coun e ing
clima e change and ealizing a mo e sus ainable wo ld.
ACKNOWLEDGEMENTS
This p ojec has ecei ed unding om he Eu opean Union's Ho izon 2020 esea ch
and inno a ion p og amme unde g an ag eemen No 101036588, Eu opean ood chain
supply o educe GHG emissions by 2050 (ENOUGH).
Li e a u e Ci ed
Bessemans, N., Ve bo en, P., Ve linden, B.E., and Nicolaï, B.M. (2016). A no el ype o dynamic con olled a mosphe e
s o age based on he espi a o y quo ien (RQ-DCA). Pos ha es Biol. Technol. 115, pp. 91-102.
F anck, C., Lamme yn, J., Ho, Q.T., Ve bo en, P., Ve linden, and B., Nicolaï, B.M. (2007). B owning diso de s in pea
ui . Pos ha es Biol. Technol. 43, pp. 1-13.
Kade , A.A. (1986). Biochemical and physiological basis o e ec s o con olled and modi ied a mosphe es on ui s
and ege ables. Food Technol. 40,99–104.
K äu le , B. (2008). Chlo ophyll b eakdown and chlo ophyll ca aboli es in lea es and ui . Pho ochem. Pho obiol.
Sci. 7, 1114–1120.
MacDougall, D. (2002). Clou in ood: Imp o ing quali y. Woodhead Publishing Se ies in Food Science. Else ie
Science, Camb idge, GBR Else ie Science.
Mülle , T., Ul ich, M., Ongania, K.H., and K äu le , B. (2007). Colou less e apy olic chlo ophyll ca aboli es ound in
ipening ui a e e ec i e an ioxidan s. Angew. Chem. In . Ed. 46, 8699–8702.
Ve linden, B.E., Bessemans, N., Ve bo en, P., and Nicolai, B.M. (2023). Sa ing ene gy using RQ-based dynamic
con olled a mosphe e s o age o bluebe y ui . P oceedings o he 26 h IIR In e na ional Cong ess o Re ige a ion:
Pa is , F ance, Augus 21-25, 2023. h ps://doi.o g/10.18462/ii .ic .2023.0954
Zanella, A. (2003). Con ol o apple supe icial scald and ipening—a compa ison be ween 1-me hylcyclop opene
and diphenylamine pos ha es ea men s, ini ial low oxygen s ess and ul a low oxygen s o age. Pos ha es Biol.
Technol. 27, 69–78 h p://dx.doi.o g/10.1016/s0925-5214(02) 00187-4.
