Embed Size (px)
Citation preview
The elusive nature of minerality in
white wine
Wendy V Parr Ph DLincoln University Christchurch New Zealand
Minerality
Dominique Valentin Jordi Ballester Dominique Peyron
Claire Grose
Jason Breitmeyer Rob Sherlock Brett Robinson
Philippe Darriet
Minerality ie what do we experience in wine that evokes
geological metaphors
bull Flinty silex
bull Smoky
bull Chalky calcareous
bull Matchstick gun flint struck match
bull Oyster shell iodine fossils
bull Earthy
bull Wet stones stony
bull Pencil lead graphite
Minerality in wine
bull Fashionable notion
ndash Increasingly reported as a wine descriptor
bull Positive concept
ndash Associated with cool climate wines
ndash Associated with higher-priced wines
bull Good for marketing
ndash Links with the concept of terroir or provenance
However geologists amp plant scientists tell us that it is unlikely that we are tasting directly the vineyard rocks amp soil (Maltman 2013)
So is perceived minerality all in the
mind
Have we been seduced by smart marketing
ldquoAll in the mindrdquo
bull Are we sensing specific characters in wine that evoke geological metaphors
ndash Data-driven perception based on the phenomenological properties of the wine
bull Or is minerality primarily a mental construction
ndash Top-down or knowledge-based perception
bull And is our experience of perceived minerality in wine shared by others
ndash Within a culture
ndash Across cultures
bull Influence of wine composition
Qs arising from industry media amp anecdotal reports
bull Can minerality be smelled or is it a palate sensation only
bull How is perceived minerality related to
ndash Acidity
ndash Relative absence of varietal flavour
ndash Reductive characteristics
bull What other wine characteristics associate with or drive perception of mineral character
How did we go about it(Parr et al FQAP 2015)
bull 32 French amp 31 NZ wine professionals evaluated 16 100 Sauvignon wines (8 French 8 NZ)ndash Via 3 modes of perception
bull Smelling only
bull Full lsquotastingrsquo
bull Palate only taste amp trigeminal (mouthfeel)
bull Wines were categorised amp describedndash Varietal characters
ndash Mineral characters (flinty smoky calcareous)
ndash Acidity freshness
ndash Reductive characters (graphite burnt rubber sulphide)
ndash Overall mineral intensity complexity familiarity liking
bull Physico-chemical analyses on the wines
What did we find
French and NZ wines were judged similarly in terms of overall intensity of mineral character
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFCFSHB
NZAVFB
NZAVV
NZLWB
NZLWWH
NZRS
NZRSL
NZSVANZSVCH
Mineral NC
Mineral OMineral G
-2
-1
0
1
2
3
-4 -3 -2 -1 0 1 2 3
F2
(24
83
)
F1 (6944 )
Biplot (axes F1 et F2 9427 )
How were the wines characterised French Free SortingColours represent MDS clusters
pyrazine
good balance
acidic
varietal character
boxwood
fruit
tropical
mineral
asparagus
expressive aromatics
reductive
cat pee
citrusfloral
herbaceous
bitter
faulty
terroir expressioncreamy
oxidised
unripe
oak
smokey
NZRSL
FSHB
NZAVFB FSFC
FBCLMFSBI
NZRS
FSBAGA
FLCH
FLPCNZSVA
NZLWWH
FBCBB
NZLWB
NZAVV
NZSVCH
-15
-1
-05
0
05
1
-1 -05 0 05 1 15 2 25 3
F2 (
20
50
)
F1 (3900 )
FR axes F1 et F2 5950
How were the wines characterised NZ Free SortingColours represent MDS clusters
FSHB
FBCBB
NZSVCH
FLCH
NZLWBNZLWWH
NZRSNZAVFB
NZAVV
FSBI
FSBAGA
FSFC
FLPC
NZRSL
NZSVA
FBCLM
tropical
herbaceous
acidic reductive
pyrazine
boxwood
ripe
agedcitrus
faulty
complex
unripe
good balance
simplethin oak
varietal character
gooseberry
expressive aromatics
good weight
mineral lack of varietal character
oxidised
grapefruit
lack acidity
-15
-1
-05
0
05
1
-15 -1 -05 0 05 1 15 2
F2 (
15
33
)
F1 (4813 )
NZ_axes F1 et F2 6345
Can mineral be smelled
bull Both cultures reported perceiving minerality via each mode of perception ie
ndash Mineral can be smelled and
ndash Is a palate experience (taste amp mouthfeel)
bull French lsquotastersrsquo relied more on their noses than NZers in their full-tasting judgments of minerality NZers relying equally on olfaction (smelling) and palate phenomena
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Minerality
Dominique Valentin Jordi Ballester Dominique Peyron
Claire Grose
Jason Breitmeyer Rob Sherlock Brett Robinson
Philippe Darriet
Minerality ie what do we experience in wine that evokes
geological metaphors
bull Flinty silex
bull Smoky
bull Chalky calcareous
bull Matchstick gun flint struck match
bull Oyster shell iodine fossils
bull Earthy
bull Wet stones stony
bull Pencil lead graphite
Minerality in wine
bull Fashionable notion
ndash Increasingly reported as a wine descriptor
bull Positive concept
ndash Associated with cool climate wines
ndash Associated with higher-priced wines
bull Good for marketing
ndash Links with the concept of terroir or provenance
However geologists amp plant scientists tell us that it is unlikely that we are tasting directly the vineyard rocks amp soil (Maltman 2013)
So is perceived minerality all in the
mind
Have we been seduced by smart marketing
ldquoAll in the mindrdquo
bull Are we sensing specific characters in wine that evoke geological metaphors
ndash Data-driven perception based on the phenomenological properties of the wine
bull Or is minerality primarily a mental construction
ndash Top-down or knowledge-based perception
bull And is our experience of perceived minerality in wine shared by others
ndash Within a culture
ndash Across cultures
bull Influence of wine composition
Qs arising from industry media amp anecdotal reports
bull Can minerality be smelled or is it a palate sensation only
bull How is perceived minerality related to
ndash Acidity
ndash Relative absence of varietal flavour
ndash Reductive characteristics
bull What other wine characteristics associate with or drive perception of mineral character
How did we go about it(Parr et al FQAP 2015)
bull 32 French amp 31 NZ wine professionals evaluated 16 100 Sauvignon wines (8 French 8 NZ)ndash Via 3 modes of perception
bull Smelling only
bull Full lsquotastingrsquo
bull Palate only taste amp trigeminal (mouthfeel)
bull Wines were categorised amp describedndash Varietal characters
ndash Mineral characters (flinty smoky calcareous)
ndash Acidity freshness
ndash Reductive characters (graphite burnt rubber sulphide)
ndash Overall mineral intensity complexity familiarity liking
bull Physico-chemical analyses on the wines
What did we find
French and NZ wines were judged similarly in terms of overall intensity of mineral character
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFCFSHB
NZAVFB
NZAVV
NZLWB
NZLWWH
NZRS
NZRSL
NZSVANZSVCH
Mineral NC
Mineral OMineral G
-2
-1
0
1
2
3
-4 -3 -2 -1 0 1 2 3
F2
(24
83
)
F1 (6944 )
Biplot (axes F1 et F2 9427 )
How were the wines characterised French Free SortingColours represent MDS clusters
pyrazine
good balance
acidic
varietal character
boxwood
fruit
tropical
mineral
asparagus
expressive aromatics
reductive
cat pee
citrusfloral
herbaceous
bitter
faulty
terroir expressioncreamy
oxidised
unripe
oak
smokey
NZRSL
FSHB
NZAVFB FSFC
FBCLMFSBI
NZRS
FSBAGA
FLCH
FLPCNZSVA
NZLWWH
FBCBB
NZLWB
NZAVV
NZSVCH
-15
-1
-05
0
05
1
-1 -05 0 05 1 15 2 25 3
F2 (
20
50
)
F1 (3900 )
FR axes F1 et F2 5950
How were the wines characterised NZ Free SortingColours represent MDS clusters
FSHB
FBCBB
NZSVCH
FLCH
NZLWBNZLWWH
NZRSNZAVFB
NZAVV
FSBI
FSBAGA
FSFC
FLPC
NZRSL
NZSVA
FBCLM
tropical
herbaceous
acidic reductive
pyrazine
boxwood
ripe
agedcitrus
faulty
complex
unripe
good balance
simplethin oak
varietal character
gooseberry
expressive aromatics
good weight
mineral lack of varietal character
oxidised
grapefruit
lack acidity
-15
-1
-05
0
05
1
-15 -1 -05 0 05 1 15 2
F2 (
15
33
)
F1 (4813 )
NZ_axes F1 et F2 6345
Can mineral be smelled
bull Both cultures reported perceiving minerality via each mode of perception ie
ndash Mineral can be smelled and
ndash Is a palate experience (taste amp mouthfeel)
bull French lsquotastersrsquo relied more on their noses than NZers in their full-tasting judgments of minerality NZers relying equally on olfaction (smelling) and palate phenomena
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Minerality ie what do we experience in wine that evokes
geological metaphors
bull Flinty silex
bull Smoky
bull Chalky calcareous
bull Matchstick gun flint struck match
bull Oyster shell iodine fossils
bull Earthy
bull Wet stones stony
bull Pencil lead graphite
Minerality in wine
bull Fashionable notion
ndash Increasingly reported as a wine descriptor
bull Positive concept
ndash Associated with cool climate wines
ndash Associated with higher-priced wines
bull Good for marketing
ndash Links with the concept of terroir or provenance
However geologists amp plant scientists tell us that it is unlikely that we are tasting directly the vineyard rocks amp soil (Maltman 2013)
So is perceived minerality all in the
mind
Have we been seduced by smart marketing
ldquoAll in the mindrdquo
bull Are we sensing specific characters in wine that evoke geological metaphors
ndash Data-driven perception based on the phenomenological properties of the wine
bull Or is minerality primarily a mental construction
ndash Top-down or knowledge-based perception
bull And is our experience of perceived minerality in wine shared by others
ndash Within a culture
ndash Across cultures
bull Influence of wine composition
Qs arising from industry media amp anecdotal reports
bull Can minerality be smelled or is it a palate sensation only
bull How is perceived minerality related to
ndash Acidity
ndash Relative absence of varietal flavour
ndash Reductive characteristics
bull What other wine characteristics associate with or drive perception of mineral character
How did we go about it(Parr et al FQAP 2015)
bull 32 French amp 31 NZ wine professionals evaluated 16 100 Sauvignon wines (8 French 8 NZ)ndash Via 3 modes of perception
bull Smelling only
bull Full lsquotastingrsquo
bull Palate only taste amp trigeminal (mouthfeel)
bull Wines were categorised amp describedndash Varietal characters
ndash Mineral characters (flinty smoky calcareous)
ndash Acidity freshness
ndash Reductive characters (graphite burnt rubber sulphide)
ndash Overall mineral intensity complexity familiarity liking
bull Physico-chemical analyses on the wines
What did we find
French and NZ wines were judged similarly in terms of overall intensity of mineral character
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFCFSHB
NZAVFB
NZAVV
NZLWB
NZLWWH
NZRS
NZRSL
NZSVANZSVCH
Mineral NC
Mineral OMineral G
-2
-1
0
1
2
3
-4 -3 -2 -1 0 1 2 3
F2
(24
83
)
F1 (6944 )
Biplot (axes F1 et F2 9427 )
How were the wines characterised French Free SortingColours represent MDS clusters
pyrazine
good balance
acidic
varietal character
boxwood
fruit
tropical
mineral
asparagus
expressive aromatics
reductive
cat pee
citrusfloral
herbaceous
bitter
faulty
terroir expressioncreamy
oxidised
unripe
oak
smokey
NZRSL
FSHB
NZAVFB FSFC
FBCLMFSBI
NZRS
FSBAGA
FLCH
FLPCNZSVA
NZLWWH
FBCBB
NZLWB
NZAVV
NZSVCH
-15
-1
-05
0
05
1
-1 -05 0 05 1 15 2 25 3
F2 (
20
50
)
F1 (3900 )
FR axes F1 et F2 5950
How were the wines characterised NZ Free SortingColours represent MDS clusters
FSHB
FBCBB
NZSVCH
FLCH
NZLWBNZLWWH
NZRSNZAVFB
NZAVV
FSBI
FSBAGA
FSFC
FLPC
NZRSL
NZSVA
FBCLM
tropical
herbaceous
acidic reductive
pyrazine
boxwood
ripe
agedcitrus
faulty
complex
unripe
good balance
simplethin oak
varietal character
gooseberry
expressive aromatics
good weight
mineral lack of varietal character
oxidised
grapefruit
lack acidity
-15
-1
-05
0
05
1
-15 -1 -05 0 05 1 15 2
F2 (
15
33
)
F1 (4813 )
NZ_axes F1 et F2 6345
Can mineral be smelled
bull Both cultures reported perceiving minerality via each mode of perception ie
ndash Mineral can be smelled and
ndash Is a palate experience (taste amp mouthfeel)
bull French lsquotastersrsquo relied more on their noses than NZers in their full-tasting judgments of minerality NZers relying equally on olfaction (smelling) and palate phenomena
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Minerality in wine
bull Fashionable notion
ndash Increasingly reported as a wine descriptor
bull Positive concept
ndash Associated with cool climate wines
ndash Associated with higher-priced wines
bull Good for marketing
ndash Links with the concept of terroir or provenance
However geologists amp plant scientists tell us that it is unlikely that we are tasting directly the vineyard rocks amp soil (Maltman 2013)
So is perceived minerality all in the
mind
Have we been seduced by smart marketing
ldquoAll in the mindrdquo
bull Are we sensing specific characters in wine that evoke geological metaphors
ndash Data-driven perception based on the phenomenological properties of the wine
bull Or is minerality primarily a mental construction
ndash Top-down or knowledge-based perception
bull And is our experience of perceived minerality in wine shared by others
ndash Within a culture
ndash Across cultures
bull Influence of wine composition
Qs arising from industry media amp anecdotal reports
bull Can minerality be smelled or is it a palate sensation only
bull How is perceived minerality related to
ndash Acidity
ndash Relative absence of varietal flavour
ndash Reductive characteristics
bull What other wine characteristics associate with or drive perception of mineral character
How did we go about it(Parr et al FQAP 2015)
bull 32 French amp 31 NZ wine professionals evaluated 16 100 Sauvignon wines (8 French 8 NZ)ndash Via 3 modes of perception
bull Smelling only
bull Full lsquotastingrsquo
bull Palate only taste amp trigeminal (mouthfeel)
bull Wines were categorised amp describedndash Varietal characters
ndash Mineral characters (flinty smoky calcareous)
ndash Acidity freshness
ndash Reductive characters (graphite burnt rubber sulphide)
ndash Overall mineral intensity complexity familiarity liking
bull Physico-chemical analyses on the wines
What did we find
French and NZ wines were judged similarly in terms of overall intensity of mineral character
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFCFSHB
NZAVFB
NZAVV
NZLWB
NZLWWH
NZRS
NZRSL
NZSVANZSVCH
Mineral NC
Mineral OMineral G
-2
-1
0
1
2
3
-4 -3 -2 -1 0 1 2 3
F2
(24
83
)
F1 (6944 )
Biplot (axes F1 et F2 9427 )
How were the wines characterised French Free SortingColours represent MDS clusters
pyrazine
good balance
acidic
varietal character
boxwood
fruit
tropical
mineral
asparagus
expressive aromatics
reductive
cat pee
citrusfloral
herbaceous
bitter
faulty
terroir expressioncreamy
oxidised
unripe
oak
smokey
NZRSL
FSHB
NZAVFB FSFC
FBCLMFSBI
NZRS
FSBAGA
FLCH
FLPCNZSVA
NZLWWH
FBCBB
NZLWB
NZAVV
NZSVCH
-15
-1
-05
0
05
1
-1 -05 0 05 1 15 2 25 3
F2 (
20
50
)
F1 (3900 )
FR axes F1 et F2 5950
How were the wines characterised NZ Free SortingColours represent MDS clusters
FSHB
FBCBB
NZSVCH
FLCH
NZLWBNZLWWH
NZRSNZAVFB
NZAVV
FSBI
FSBAGA
FSFC
FLPC
NZRSL
NZSVA
FBCLM
tropical
herbaceous
acidic reductive
pyrazine
boxwood
ripe
agedcitrus
faulty
complex
unripe
good balance
simplethin oak
varietal character
gooseberry
expressive aromatics
good weight
mineral lack of varietal character
oxidised
grapefruit
lack acidity
-15
-1
-05
0
05
1
-15 -1 -05 0 05 1 15 2
F2 (
15
33
)
F1 (4813 )
NZ_axes F1 et F2 6345
Can mineral be smelled
bull Both cultures reported perceiving minerality via each mode of perception ie
ndash Mineral can be smelled and
ndash Is a palate experience (taste amp mouthfeel)
bull French lsquotastersrsquo relied more on their noses than NZers in their full-tasting judgments of minerality NZers relying equally on olfaction (smelling) and palate phenomena
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
So is perceived minerality all in the
mind
Have we been seduced by smart marketing
ldquoAll in the mindrdquo
bull Are we sensing specific characters in wine that evoke geological metaphors
ndash Data-driven perception based on the phenomenological properties of the wine
bull Or is minerality primarily a mental construction
ndash Top-down or knowledge-based perception
bull And is our experience of perceived minerality in wine shared by others
ndash Within a culture
ndash Across cultures
bull Influence of wine composition
Qs arising from industry media amp anecdotal reports
bull Can minerality be smelled or is it a palate sensation only
bull How is perceived minerality related to
ndash Acidity
ndash Relative absence of varietal flavour
ndash Reductive characteristics
bull What other wine characteristics associate with or drive perception of mineral character
How did we go about it(Parr et al FQAP 2015)
bull 32 French amp 31 NZ wine professionals evaluated 16 100 Sauvignon wines (8 French 8 NZ)ndash Via 3 modes of perception
bull Smelling only
bull Full lsquotastingrsquo
bull Palate only taste amp trigeminal (mouthfeel)
bull Wines were categorised amp describedndash Varietal characters
ndash Mineral characters (flinty smoky calcareous)
ndash Acidity freshness
ndash Reductive characters (graphite burnt rubber sulphide)
ndash Overall mineral intensity complexity familiarity liking
bull Physico-chemical analyses on the wines
What did we find
French and NZ wines were judged similarly in terms of overall intensity of mineral character
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFCFSHB
NZAVFB
NZAVV
NZLWB
NZLWWH
NZRS
NZRSL
NZSVANZSVCH
Mineral NC
Mineral OMineral G
-2
-1
0
1
2
3
-4 -3 -2 -1 0 1 2 3
F2
(24
83
)
F1 (6944 )
Biplot (axes F1 et F2 9427 )
How were the wines characterised French Free SortingColours represent MDS clusters
pyrazine
good balance
acidic
varietal character
boxwood
fruit
tropical
mineral
asparagus
expressive aromatics
reductive
cat pee
citrusfloral
herbaceous
bitter
faulty
terroir expressioncreamy
oxidised
unripe
oak
smokey
NZRSL
FSHB
NZAVFB FSFC
FBCLMFSBI
NZRS
FSBAGA
FLCH
FLPCNZSVA
NZLWWH
FBCBB
NZLWB
NZAVV
NZSVCH
-15
-1
-05
0
05
1
-1 -05 0 05 1 15 2 25 3
F2 (
20
50
)
F1 (3900 )
FR axes F1 et F2 5950
How were the wines characterised NZ Free SortingColours represent MDS clusters
FSHB
FBCBB
NZSVCH
FLCH
NZLWBNZLWWH
NZRSNZAVFB
NZAVV
FSBI
FSBAGA
FSFC
FLPC
NZRSL
NZSVA
FBCLM
tropical
herbaceous
acidic reductive
pyrazine
boxwood
ripe
agedcitrus
faulty
complex
unripe
good balance
simplethin oak
varietal character
gooseberry
expressive aromatics
good weight
mineral lack of varietal character
oxidised
grapefruit
lack acidity
-15
-1
-05
0
05
1
-15 -1 -05 0 05 1 15 2
F2 (
15
33
)
F1 (4813 )
NZ_axes F1 et F2 6345
Can mineral be smelled
bull Both cultures reported perceiving minerality via each mode of perception ie
ndash Mineral can be smelled and
ndash Is a palate experience (taste amp mouthfeel)
bull French lsquotastersrsquo relied more on their noses than NZers in their full-tasting judgments of minerality NZers relying equally on olfaction (smelling) and palate phenomena
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
ldquoAll in the mindrdquo
bull Are we sensing specific characters in wine that evoke geological metaphors
ndash Data-driven perception based on the phenomenological properties of the wine
bull Or is minerality primarily a mental construction
ndash Top-down or knowledge-based perception
bull And is our experience of perceived minerality in wine shared by others
ndash Within a culture
ndash Across cultures
bull Influence of wine composition
Qs arising from industry media amp anecdotal reports
bull Can minerality be smelled or is it a palate sensation only
bull How is perceived minerality related to
ndash Acidity
ndash Relative absence of varietal flavour
ndash Reductive characteristics
bull What other wine characteristics associate with or drive perception of mineral character
How did we go about it(Parr et al FQAP 2015)
bull 32 French amp 31 NZ wine professionals evaluated 16 100 Sauvignon wines (8 French 8 NZ)ndash Via 3 modes of perception
bull Smelling only
bull Full lsquotastingrsquo
bull Palate only taste amp trigeminal (mouthfeel)
bull Wines were categorised amp describedndash Varietal characters
ndash Mineral characters (flinty smoky calcareous)
ndash Acidity freshness
ndash Reductive characters (graphite burnt rubber sulphide)
ndash Overall mineral intensity complexity familiarity liking
bull Physico-chemical analyses on the wines
What did we find
French and NZ wines were judged similarly in terms of overall intensity of mineral character
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFCFSHB
NZAVFB
NZAVV
NZLWB
NZLWWH
NZRS
NZRSL
NZSVANZSVCH
Mineral NC
Mineral OMineral G
-2
-1
0
1
2
3
-4 -3 -2 -1 0 1 2 3
F2
(24
83
)
F1 (6944 )
Biplot (axes F1 et F2 9427 )
How were the wines characterised French Free SortingColours represent MDS clusters
pyrazine
good balance
acidic
varietal character
boxwood
fruit
tropical
mineral
asparagus
expressive aromatics
reductive
cat pee
citrusfloral
herbaceous
bitter
faulty
terroir expressioncreamy
oxidised
unripe
oak
smokey
NZRSL
FSHB
NZAVFB FSFC
FBCLMFSBI
NZRS
FSBAGA
FLCH
FLPCNZSVA
NZLWWH
FBCBB
NZLWB
NZAVV
NZSVCH
-15
-1
-05
0
05
1
-1 -05 0 05 1 15 2 25 3
F2 (
20
50
)
F1 (3900 )
FR axes F1 et F2 5950
How were the wines characterised NZ Free SortingColours represent MDS clusters
FSHB
FBCBB
NZSVCH
FLCH
NZLWBNZLWWH
NZRSNZAVFB
NZAVV
FSBI
FSBAGA
FSFC
FLPC
NZRSL
NZSVA
FBCLM
tropical
herbaceous
acidic reductive
pyrazine
boxwood
ripe
agedcitrus
faulty
complex
unripe
good balance
simplethin oak
varietal character
gooseberry
expressive aromatics
good weight
mineral lack of varietal character
oxidised
grapefruit
lack acidity
-15
-1
-05
0
05
1
-15 -1 -05 0 05 1 15 2
F2 (
15
33
)
F1 (4813 )
NZ_axes F1 et F2 6345
Can mineral be smelled
bull Both cultures reported perceiving minerality via each mode of perception ie
ndash Mineral can be smelled and
ndash Is a palate experience (taste amp mouthfeel)
bull French lsquotastersrsquo relied more on their noses than NZers in their full-tasting judgments of minerality NZers relying equally on olfaction (smelling) and palate phenomena
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Qs arising from industry media amp anecdotal reports
bull Can minerality be smelled or is it a palate sensation only
bull How is perceived minerality related to
ndash Acidity
ndash Relative absence of varietal flavour
ndash Reductive characteristics
bull What other wine characteristics associate with or drive perception of mineral character
How did we go about it(Parr et al FQAP 2015)
bull 32 French amp 31 NZ wine professionals evaluated 16 100 Sauvignon wines (8 French 8 NZ)ndash Via 3 modes of perception
bull Smelling only
bull Full lsquotastingrsquo
bull Palate only taste amp trigeminal (mouthfeel)
bull Wines were categorised amp describedndash Varietal characters
ndash Mineral characters (flinty smoky calcareous)
ndash Acidity freshness
ndash Reductive characters (graphite burnt rubber sulphide)
ndash Overall mineral intensity complexity familiarity liking
bull Physico-chemical analyses on the wines
What did we find
French and NZ wines were judged similarly in terms of overall intensity of mineral character
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFCFSHB
NZAVFB
NZAVV
NZLWB
NZLWWH
NZRS
NZRSL
NZSVANZSVCH
Mineral NC
Mineral OMineral G
-2
-1
0
1
2
3
-4 -3 -2 -1 0 1 2 3
F2
(24
83
)
F1 (6944 )
Biplot (axes F1 et F2 9427 )
How were the wines characterised French Free SortingColours represent MDS clusters
pyrazine
good balance
acidic
varietal character
boxwood
fruit
tropical
mineral
asparagus
expressive aromatics
reductive
cat pee
citrusfloral
herbaceous
bitter
faulty
terroir expressioncreamy
oxidised
unripe
oak
smokey
NZRSL
FSHB
NZAVFB FSFC
FBCLMFSBI
NZRS
FSBAGA
FLCH
FLPCNZSVA
NZLWWH
FBCBB
NZLWB
NZAVV
NZSVCH
-15
-1
-05
0
05
1
-1 -05 0 05 1 15 2 25 3
F2 (
20
50
)
F1 (3900 )
FR axes F1 et F2 5950
How were the wines characterised NZ Free SortingColours represent MDS clusters
FSHB
FBCBB
NZSVCH
FLCH
NZLWBNZLWWH
NZRSNZAVFB
NZAVV
FSBI
FSBAGA
FSFC
FLPC
NZRSL
NZSVA
FBCLM
tropical
herbaceous
acidic reductive
pyrazine
boxwood
ripe
agedcitrus
faulty
complex
unripe
good balance
simplethin oak
varietal character
gooseberry
expressive aromatics
good weight
mineral lack of varietal character
oxidised
grapefruit
lack acidity
-15
-1
-05
0
05
1
-15 -1 -05 0 05 1 15 2
F2 (
15
33
)
F1 (4813 )
NZ_axes F1 et F2 6345
Can mineral be smelled
bull Both cultures reported perceiving minerality via each mode of perception ie
ndash Mineral can be smelled and
ndash Is a palate experience (taste amp mouthfeel)
bull French lsquotastersrsquo relied more on their noses than NZers in their full-tasting judgments of minerality NZers relying equally on olfaction (smelling) and palate phenomena
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
How did we go about it(Parr et al FQAP 2015)
bull 32 French amp 31 NZ wine professionals evaluated 16 100 Sauvignon wines (8 French 8 NZ)ndash Via 3 modes of perception
bull Smelling only
bull Full lsquotastingrsquo
bull Palate only taste amp trigeminal (mouthfeel)
bull Wines were categorised amp describedndash Varietal characters
ndash Mineral characters (flinty smoky calcareous)
ndash Acidity freshness
ndash Reductive characters (graphite burnt rubber sulphide)
ndash Overall mineral intensity complexity familiarity liking
bull Physico-chemical analyses on the wines
What did we find
French and NZ wines were judged similarly in terms of overall intensity of mineral character
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFCFSHB
NZAVFB
NZAVV
NZLWB
NZLWWH
NZRS
NZRSL
NZSVANZSVCH
Mineral NC
Mineral OMineral G
-2
-1
0
1
2
3
-4 -3 -2 -1 0 1 2 3
F2
(24
83
)
F1 (6944 )
Biplot (axes F1 et F2 9427 )
How were the wines characterised French Free SortingColours represent MDS clusters
pyrazine
good balance
acidic
varietal character
boxwood
fruit
tropical
mineral
asparagus
expressive aromatics
reductive
cat pee
citrusfloral
herbaceous
bitter
faulty
terroir expressioncreamy
oxidised
unripe
oak
smokey
NZRSL
FSHB
NZAVFB FSFC
FBCLMFSBI
NZRS
FSBAGA
FLCH
FLPCNZSVA
NZLWWH
FBCBB
NZLWB
NZAVV
NZSVCH
-15
-1
-05
0
05
1
-1 -05 0 05 1 15 2 25 3
F2 (
20
50
)
F1 (3900 )
FR axes F1 et F2 5950
How were the wines characterised NZ Free SortingColours represent MDS clusters
FSHB
FBCBB
NZSVCH
FLCH
NZLWBNZLWWH
NZRSNZAVFB
NZAVV
FSBI
FSBAGA
FSFC
FLPC
NZRSL
NZSVA
FBCLM
tropical
herbaceous
acidic reductive
pyrazine
boxwood
ripe
agedcitrus
faulty
complex
unripe
good balance
simplethin oak
varietal character
gooseberry
expressive aromatics
good weight
mineral lack of varietal character
oxidised
grapefruit
lack acidity
-15
-1
-05
0
05
1
-15 -1 -05 0 05 1 15 2
F2 (
15
33
)
F1 (4813 )
NZ_axes F1 et F2 6345
Can mineral be smelled
bull Both cultures reported perceiving minerality via each mode of perception ie
ndash Mineral can be smelled and
ndash Is a palate experience (taste amp mouthfeel)
bull French lsquotastersrsquo relied more on their noses than NZers in their full-tasting judgments of minerality NZers relying equally on olfaction (smelling) and palate phenomena
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
What did we find
French and NZ wines were judged similarly in terms of overall intensity of mineral character
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFCFSHB
NZAVFB
NZAVV
NZLWB
NZLWWH
NZRS
NZRSL
NZSVANZSVCH
Mineral NC
Mineral OMineral G
-2
-1
0
1
2
3
-4 -3 -2 -1 0 1 2 3
F2
(24
83
)
F1 (6944 )
Biplot (axes F1 et F2 9427 )
How were the wines characterised French Free SortingColours represent MDS clusters
pyrazine
good balance
acidic
varietal character
boxwood
fruit
tropical
mineral
asparagus
expressive aromatics
reductive
cat pee
citrusfloral
herbaceous
bitter
faulty
terroir expressioncreamy
oxidised
unripe
oak
smokey
NZRSL
FSHB
NZAVFB FSFC
FBCLMFSBI
NZRS
FSBAGA
FLCH
FLPCNZSVA
NZLWWH
FBCBB
NZLWB
NZAVV
NZSVCH
-15
-1
-05
0
05
1
-1 -05 0 05 1 15 2 25 3
F2 (
20
50
)
F1 (3900 )
FR axes F1 et F2 5950
How were the wines characterised NZ Free SortingColours represent MDS clusters
FSHB
FBCBB
NZSVCH
FLCH
NZLWBNZLWWH
NZRSNZAVFB
NZAVV
FSBI
FSBAGA
FSFC
FLPC
NZRSL
NZSVA
FBCLM
tropical
herbaceous
acidic reductive
pyrazine
boxwood
ripe
agedcitrus
faulty
complex
unripe
good balance
simplethin oak
varietal character
gooseberry
expressive aromatics
good weight
mineral lack of varietal character
oxidised
grapefruit
lack acidity
-15
-1
-05
0
05
1
-15 -1 -05 0 05 1 15 2
F2 (
15
33
)
F1 (4813 )
NZ_axes F1 et F2 6345
Can mineral be smelled
bull Both cultures reported perceiving minerality via each mode of perception ie
ndash Mineral can be smelled and
ndash Is a palate experience (taste amp mouthfeel)
bull French lsquotastersrsquo relied more on their noses than NZers in their full-tasting judgments of minerality NZers relying equally on olfaction (smelling) and palate phenomena
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
How were the wines characterised French Free SortingColours represent MDS clusters
pyrazine
good balance
acidic
varietal character
boxwood
fruit
tropical
mineral
asparagus
expressive aromatics
reductive
cat pee
citrusfloral
herbaceous
bitter
faulty
terroir expressioncreamy
oxidised
unripe
oak
smokey
NZRSL
FSHB
NZAVFB FSFC
FBCLMFSBI
NZRS
FSBAGA
FLCH
FLPCNZSVA
NZLWWH
FBCBB
NZLWB
NZAVV
NZSVCH
-15
-1
-05
0
05
1
-1 -05 0 05 1 15 2 25 3
F2 (
20
50
)
F1 (3900 )
FR axes F1 et F2 5950
How were the wines characterised NZ Free SortingColours represent MDS clusters
FSHB
FBCBB
NZSVCH
FLCH
NZLWBNZLWWH
NZRSNZAVFB
NZAVV
FSBI
FSBAGA
FSFC
FLPC
NZRSL
NZSVA
FBCLM
tropical
herbaceous
acidic reductive
pyrazine
boxwood
ripe
agedcitrus
faulty
complex
unripe
good balance
simplethin oak
varietal character
gooseberry
expressive aromatics
good weight
mineral lack of varietal character
oxidised
grapefruit
lack acidity
-15
-1
-05
0
05
1
-15 -1 -05 0 05 1 15 2
F2 (
15
33
)
F1 (4813 )
NZ_axes F1 et F2 6345
Can mineral be smelled
bull Both cultures reported perceiving minerality via each mode of perception ie
ndash Mineral can be smelled and
ndash Is a palate experience (taste amp mouthfeel)
bull French lsquotastersrsquo relied more on their noses than NZers in their full-tasting judgments of minerality NZers relying equally on olfaction (smelling) and palate phenomena
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
How were the wines characterised NZ Free SortingColours represent MDS clusters
FSHB
FBCBB
NZSVCH
FLCH
NZLWBNZLWWH
NZRSNZAVFB
NZAVV
FSBI
FSBAGA
FSFC
FLPC
NZRSL
NZSVA
FBCLM
tropical
herbaceous
acidic reductive
pyrazine
boxwood
ripe
agedcitrus
faulty
complex
unripe
good balance
simplethin oak
varietal character
gooseberry
expressive aromatics
good weight
mineral lack of varietal character
oxidised
grapefruit
lack acidity
-15
-1
-05
0
05
1
-15 -1 -05 0 05 1 15 2
F2 (
15
33
)
F1 (4813 )
NZ_axes F1 et F2 6345
Can mineral be smelled
bull Both cultures reported perceiving minerality via each mode of perception ie
ndash Mineral can be smelled and
ndash Is a palate experience (taste amp mouthfeel)
bull French lsquotastersrsquo relied more on their noses than NZers in their full-tasting judgments of minerality NZers relying equally on olfaction (smelling) and palate phenomena
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Can mineral be smelled
bull Both cultures reported perceiving minerality via each mode of perception ie
ndash Mineral can be smelled and
ndash Is a palate experience (taste amp mouthfeel)
bull French lsquotastersrsquo relied more on their noses than NZers in their full-tasting judgments of minerality NZers relying equally on olfaction (smelling) and palate phenomena
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
PCA Mineral intensity judgements in the 3 conditions
O = OrthonasalN = NoseclipG = Global
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
So which wine attributes predicted perceived mineralityin Sauvignon wine
And was there much consensus across cultures
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Regression of Mineral by other descriptors Nose only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 529 lt 00001 Citrus 378 lt 0001
Passionfruit -464 lt 00001 Passionfruit -337 lt 0001
Chalkycalc 675 lt 00001 Chalkycalc 594 lt 00001
Flintsmo 975 lt 00001 Flintsmo 723 lt 00001
Leadgraph 262 lt01 Leadgraph 451 lt 00001
Liking 310 lt 001 Liking 301 lt 001
Freshzingy 251 lt 005 Concentrat -205 lt 005
Iodioyster 209 lt 005
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Regression of Mineral by other descriptors Full tasting Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Citrus 422 lt 00001 Citrus 209 lt 005
Passfruit -212 lt 005 Green -242 lt 005
Bitter 208 lt 005 Bitter 234 lt 005
Chalkycalc 205 lt 005 Chalkycalc 451 lt 00001
Flintsmo 461 lt 00001 Astringent -217 lt 005
Sweet -657 lt 00001 Freshzingy 220 lt 005
Leadgraphite 281 lt01 Herbaceous 223 lt 005
Liking 198 lt 005 ConcWght 248 lt 005
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Regression of Mineral by other descriptors Palate only Multiple Linear Regression alpha lt 005
Predictors
French
lsquotastersrsquo
t p Predictors
N Z lsquotastersrsquo
t p
Freshzingy 522 lt 00001 Freshzingy 400 lt 00001
Flintsmo 853 lt 00001 Flintsmo 444 lt 00001
Chalkycalc 666 lt 00001 Chalkycalc 693 lt 00001
Souracid -226 lt 005 Souracid -226 lt 005
Sweet -433 lt 005 Leadgraph 202 lt 005
Bitter 397 lt 00001
PalateWght 537 lt 00001
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Influence of wine composition(Parr et al FRIN 2016)
bull Standard parameters + an acidity index
bull pH TA ethanol RS total extract free amp bound SO2
bull Fermentation-derived volatile aroma compounds volatile
organic acids amp IBMP
bull Non-volatile organic acids
bull Tartaric lactic malic shikimic malonic
bull Wine elemental composition (ICP-OES)
bull Mg K Ca Na Rb Sr Al Fe Mn Si La P S Ba Cu Mo Ni Zn
bull Varietal thiols amp volatile sulphur compounds associated
with pungent aromas
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Sulphur compounds
bull 3-mercaptohexan-1-ol (3MH) citrusgrapefruit
bull 3-mercaptohexyl acetate (3MHA) passion fruit
bull 4-mercapto-4-methylpentan-2-one (4MMP)
sweatyboxwood
bull Ethyl-2-sulfanylacetate (E2SA) ldquobaked beansrdquo note
bull Benzenemethanethiol (BMT) flinty note
bull Dimethyl sulfide (quince truffle)
bull Hydrogen sulfide (rotten eggs)
bull Meacutethanethiol (stagnant water halitosis)
bull Ethanethiol (onion rubber)
bull Dieacutethyl sulfide
bull Dimeacutethyl disulfide (quince asparagus)
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
PLSR plot projection of Y variables (M perceived minerality scores) on X variables (chemical compounds in red) and Wines (in green)
Figure 2
MeSH
DMSDMDS
BMTEthyl Acetate
Isoamyl Acetate
Octanoic Acid
Tartaric acid
DL-Lactic acid
Ba
MoZn
Titratable AcidityAcidityIndex
Density
Free SO2
Bound SO2
DMS(2)
BMT (2)Ethyl Acetate (2)
Isoamyl Acetate (2)Octanoic Acid (2)
Tartaric acid (2)
Malonic acid (2)
Ca (2)
S (2)
Ba (2)
Mo (2)Zn (2)
Titratable Acidity (2)AcidityIndex (2)
Density (2)
Free SO2 (2)
Bound SO2 (2)
3MH
Isobutanol
Isoamyl Alcohol
Diethyl succinate
Hexanoic Acid
L-Malic acid
Ca
Na
Mn
Isobutanol (2)
Isoamyl Alcohol (2)
Diethyl succinate (2)
Hexanoic Acid (2)
L-Malic acid (2)
DL-Lactic acid (2)
Na (2)
FR_Minerality
NZ_Minerality
FBCBB
FBCLM
FLCH
FLPC
FSBAGA
FSBI
FSFC
FSHB
NZAVFB
NZAVV
NZLWB
NZLWWHNZRS
NZRSL
NZSVA
NZSVCH
-1
-075
-05
-025
0
025
05
075
1
-1 -075 -05 -025 0 025 05 075 1
t2
t1
t1 amp t2
X
Y
Obs
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Aspects of wine composition associated statistically with perception of mineral character in S blanc wines
(Parr et al FRIN 2016)
Positive predictors Negative predictors
French N Z French N Z
Free SO2 Bound SO2 Tartaric acid Isoamyl alcohol (fusel solvent)
Isoamyl
acetate(fruity)
Hexanoic acid (fatty waxy barnyard)
Titratable
acidity
Diethyl succinate (fruity wine-like)
Malic acid Isobutanol(fusel solvent)
Na
Ca
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Implications
Not just smart marketingbull Perceived minerality in Sauvignon wine appears
based on data-driven input (wine attributes) bull Characters driving perceived minerality in Sauvignon
ndash Citrus notes and freshzingy character (acid)ndash Absence of passion fruit amp green characters
Caution needed in how the term mineral is usedbull Much between-judge variability in judging intensity
of mineral character in winebull Wine varietal differences may influence how
minerality is perceived in any particular wine bull Wine composition aspects appear complex amp
require further research
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
Acknowledgement
bull Our tasters in France amp NZ amp wine producers who provided their wines for the study
bull Funding
ndash Grape and Wine Research Programme PFR NZ
ndash New Zealand Winegrowers
ndash Pernod Ricard NZ amp Pernod Ricard Research Centre Paris France
ndash Regional Council of Burgundy
ndash Bureau Interprofessionnel des Vins de Bourgogne and Conseil Interprofessionel des Vins de Bordeaux
ndash Royal Society of New Zealand ISAT Linkages Fund
And thank you for your attention
And thank you for your attention
