Do Nascimento, 2015

8/18/2019 Do Nascimento, 2015

1/12

Correlations

between

subjective quality

and

physicochemicalattributes

of

fresh

fruits

and

vegetables

M. Cecilia do Nascimento Nunes*

University of South Florida, Department of Cell Biology, Microbiology and Molecular Biology, Tampa, FL 33602, USA

A

R

T

I

C

L

E

I

N

F

O

Article history:

Received

9

November

2014

Received in revised form 30 April 2015Accepted 3 May 2015

Available

online

19

May 2015

Keywords:

Quality rating charts

Color

Texture

Weight loss

Anthocyanins

Chlorophyll

A

B

S

T

R

A

C

T

Color charts and rating scales have been developed for several fresh fruits and vegetables (FFVs) but

limited information is available regarding the correlation between subjective evaluations and

physicochemical attributes. The objective of this work was to correlate subjective quality data with

quantitative analytical data collected for several fruits exposed to different environmental conditions.

Avocados, blueberries, peppers, strawberries and tomatoes were exposed to a range of different

temperaturesandhumidity conditions forvariedperiods of time, andquality evaluated using both rating

scales and physicochemical analysis. The strength of the relationship between variables wasmeasured

using thePearson correlation coef cient (r ) and the coef cient of determination(r 2) and, thesignicance

of the relationship was expressed by probability levels ( p=0.05). Overall, there was a signicant

correlationbetweenmostof the subjectivequality attributesevaluatedand the physicochemicalanalysis

performed. Subjectivecolorwas signicantlycorrelated with hue angle for all fruits evaluated except for

blueberries for which subjective color had a stronger correlation with L* values. Correlations between

subjective color and anthocyanins, ascorbic acid or chlorophyll contents were also signicant. Shriveling

or stem freshnesswas strongly correlatedwithweightlosswhereas subjectivermnesswassignicantly

correlated with instrumental texture. Results fromthis work showed that subjective quality evaluations

using ratingscalescan be a reliableandsimplemethodto estimate changes in color, softening, waterloss,

and ultimately changes in specic chemical components when FFVs are exposed to different

environmental conditions. A color chart is proposed for the visual evaluation of strawberry quality.ã 2015 Elsevier B.V. All rights reserved.

1. Introduction

Visual appearance of fresh fruits and vegetables (FFVs) has the

greatest impact on retailers buying decisions and on consumer

choices and purchases. Attributes such as appearance, freshness

and color are considered the foremost criteria used to evaluate the

immediate quality of FFVs (Clydesdale, 1991; Mitcham et al., 1996;

Barrett et al., 2010). As a result, they are used as quality indicators

throughout the supply chain, from the farm to the consumer, and

ultimately determine product acceptance or rejection. Texture,taste and aroma are also important sensory attributes but are

mostly related with subsequent purchases (Clydesdale, 1991;

Barrett et al., 2010). Nutrient content is not visible or touchable,

and therefore is often disregarded as a quality attribute when it

comes to food choices and purchase decisions. However, FFVs are

important contributors to a well-balanced diet and to human

wellbeing as they supply important macronutrients, such as

carbohydrates and ber, and micronutrients such as vitamins and

minerals as well as polyphenols.

Subjective quality evaluations are often used to rate the

appearance, texture and avor of FFVs and, unlike formal sensory

panels, these are usually performed by trained individuals but not

on a sensory panel setting (e.g., quality control, and to estimate

ripeness stage and maturity at harvest). Although these evalua-

tions are criticized by some as being inexact, in the absence of

formal analytical or affective sensory panels they are valuable toquality control, and to determine the ripeness stage or the end of

shelf life of FFVs. In addition, they are faster, easier and less

expensive than sensory panels or instrumental measurements

which may require extensive training and complex logistics, or

expensive equipment (Mitcham et al., 1996; Barrett et al., 2010).

Consequently, many researchers frequently use somewhat quanti-

tative scoring systems either alone or combined with drawings or

photographs to evaluate the sensory quality of FFVs. For example,

Kader and Cantwell (2006) developed several color charts along

with rating scales and descriptors for physical damage of produce.* Tel.: +1 813 974 9307; fax: +1 813 905 9919.

E-mail address: [email protected] (M. C.d.N. Nunes).

http://dx.doi.org/10.1016/j.postharvbio.2015.05.001

0925-5214/ã 2015 Elsevier B.V. All rights reserved.

Postharvest Biology and Technology 107 (2015) 43–54

Contents

lists

available

at

ScienceDirect

Postharvest

Biology

and

Technology

j ou rna l homepage : www.elsev ier .com/locate/postharvbio

mailto:[email protected]:[email protected]://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://www.sciencedirect.com/science/journal/09255214http://www.elsevier.com/locate/postharvbiohttp://www.elsevier.com/locate/postharvbiohttp://www.elsevier.com/locate/postharvbiohttp://www.elsevier.com/locate/postharvbiohttp://www.sciencedirect.com/science/journal/09255214http://dx.doi.org/10.1016/j.postharvbio.2015.05.001http://dx.doi.org/10.1016/j.postharvbio.2015.05.001mailto:[email protected]://crossmark.crossref.org/dialog/?doi=10.1016/j.postharvbio.2015.05.001&domain=pdf


2/12

White et al. (2005) developed a color chart to describe skin color

and avocado ripening. Our group also have developed a system that

uses scores and descriptors to rate individual sensory quality

attributes of various fruits and vegetables (Laurin et al., 2003;

Nunes et al., 2003a,b,c; Nunes et al., 2004, 2006, 2007, 2011, 2012,

2013; Proulx et al., 2005; Nunes and Emond, 2007; Nunes, 2008;

Proulx et al., 2010; Chilson et al., 2011). In the produce industry,

color charts are also frequently used to assess the stage of ripeness

(e.g., banana, tomato and avocado color charts) or to grade (e.g.,

size, color, and defects) or even to decide if a load of produce should

be accepted or rejected based on visual inspection.

Many color charts and rating scales have been developed for

several types of produce, and various studies have shown that

correlations exist between sensory and physical and/or chemical

attributes of FFVs (Ressureccion and Shewfelt, 1985; Maul et al.,

2000,b; Harker et al., 2002a,b; Safner et al., 2008; Gunness et al.,

2009; Pace et al., 2011; Corollaro et al., 2014). For example,

rmness and color of tomatoes were highly correlated with

sensory attributes (Ressureccion and Shewfelt, 1985) whereas

perceived sweetness or sourness was correlated with specic

volatile compounds (Maul et al., 2000). Pace et al., (2011) reported

a signicant correlation between b*, chroma, pH, titratable acidity

and appearance of fresh-cut nectarines. In apples, titratable acidity

was suggested to be a good predictor of acid taste (Harker et al.,2002a,b) and texture analysis correlated well with sensory

perception of apple texture (Harker et al., 2002a,b; Corollaro

et al., 2014). However, these studies used consumer or trained

sensory panels and to our knowledge there is a lack of published

studies showing that subjective quality data collected by trained

individuals (not in a sensory panel setting) can also be, in the

absence of formal trained sensory panel, a reliable way the

determined changes in the overall quality of FFVs. The objectives of

this work were to: (1) correlate subjective quality data, such as

color, rmness and shriveling, with quantitative analytical data

collected for different FFVs, and to show that in the absence of a

formal sensory panel the use of color charts and rating scales can

be used by trained individuals as an accurate way of determining

changes in overall quality of FFVs; and (2) give an example of aunique color chart designed for the evaluation of visual quality of

strawberry based on correlations between individual subjective

quality characteristics and physicochemical attributes, and that

was validated in research and commercial settings.

2. Material and methods

2.1.

Sampling

Fresh fruits and vegetables were harvested twice from

commercial operations in Florida and transported to the laboratory

within minimal delay after harvest (i.e., 1 to 6 h, depending on the

distance between the eld and the laboratory) (Tables 1 and 2). All

FFVs were harvested at the commercial maturity stage, and cluster

tomatoes were harvested from a greenhouse at the light-red stage.

Upon arrival to the laboratory, FFVs were visually selected for

uniformity of color/ripeness stage, size and freedom of defects.

Sample sizes were chosen based on the size and variability

within each commodity (i.e., the smaller the size of the fruit the

larger the number of fruits per replicates). Thus, three avocados,

three peppers, two clusters of three tomatoes each, and three

replicated samples of 15 or 20 strawberries and blueberries each,

respectively, were used for initial subjective quality evaluation, and

for instrumental color and texture analysis, and immediately

frozen to be later used for chemical compositional analysis. A total

of 20 avocados or 20 peppers (three fruit per RH), and 15 clusters of

three tomatoes each (3 clusters per RH), and a total of 15 clamshells

(3 clamshells per RH) containing 15 or 20 strawberries or

blueberries, respectively, were distributed among the ve RH-

controlled rooms and reused daily or every two days for non-

destructive quality evaluations (i.e., subjective quality evaluations

and weight loss). For destructive quality evaluations (i.e., texture

analysis and chemical analysis) and for non-destructive evalua-

tions that required manipulation of the fruit to an extent that could

cause minor bruising (i.e., instrumental color) 165 avocados or

peppers (33 fruits per RH), and 110clusters of three tomatoes each(22 clusters per RH), and 120 or 135 clamshells (24 clamshells per

RH; 27 clamshells per RH, respectively) containing 20 or 15 blue-

berries and strawberries each, respectively were distributed

among the ve RH-controlled rooms. However, every day three

clamshells of these strawberries or every two days, three of these

avocados or peppers, and two clusters or tomatoes, and three

clamshells of these blueberries were removed from their respec-

tive RH and immediately frozen, to be later used for chemical

compositional analysis. Avocados were stored for 20 d, blueberries

were stored for 16 d, and peppers and tomatoes were stored for 22

days and quality evaluated every two days. Strawberries were

stored for nine days and quality evaluated every day (Table 2). For

temperature treatments the experimental setup was similar to that

used for RH treatments except that only avocados, strawberriesand tomatoes were used in this part of the experiments (Table 1).

Since all non-destructive quality analysis (i.e., subjective quality

evaluations and weight) were assessed using always the same FFVs

samples, those were conducted within approximately 30 min after

the products were removed from storage, to minimize temperature

uctuations that could affect the quality.

2.2.

Storage

conditions

Storage conditions (i.e., temperature and relative humidity

treatments) and experimental setup used in this study were

similar to those previously described in detail by Nunes et al.

Table 1

Optimum storage conditions, cultivar, harvest location and date, and storage conditions during the temperature experiments.

Commodity Optimum temperature (C) Optimum RH (%) Cultivar Origin Harvest date Storage duration (d)a

Avocados 5–12b 85–95b ‘Choquette’ Homestead, Florida October 1, 2008 22

‘Choquette’ Homestead, Florida November 19, 2009 22

Strawberries 0c 90–95c ‘Albion’ Floral City, Florida December 12, 2008 10

‘Albion’ Floral City, Florida March 9, 2009 10

Tomatoes 7–13d 90–95d ‘Success’ Wellborn, Florida January 15, 2009 22

‘Success’ Welborn, Florida February 24, 2008 22

Storage conditions: (A) 1.8 0.8 C; (B) 5.2 0.2 C; (C) 10.6 0.6 C; (D) 15.2 0.4 C; (E) 20.2 0.2 C; 90% RH in all ve temperature-RH controlled chambers.a In some cases the experiments were terminated before the end of the storage period, at the time when at least one of the visual quality attributes evaluated reached the

maximum acceptable (rating of 3).b Woolf et al. (2014); evaluated every two days.c Mitcham (2014); evaluated every day.d Sargent

and

Moretti

(2014); light

red

greenhouse-grown

tomatoes;

evaluated

every

two

days.

44 M.C.N. Nunes / Postharvest Biology and Technology 107 (2015) 43–54


3/12

(2013). After sorting, FFVs designated for temperature experiments

(avocados, strawberries, tomatoes) were distributed into ve

temperature-humidity-controlled chambers (Forma Environmen-tal Chambers Model 3940 Series, Thermo Electron Corporation,

OH, USA) that were set at ve different temperatures with an

relative humidity (RH) of approximately 90% (Table 1). These

temperatures were chosen because they cover a wide range of

physiological temperatures, between optimum and ambient

temperature, with an optimum RH. For the humidity experiments,

temperature of the chambers containing blueberries and straw-

berries was maintained as close as possible to the optimum storage

temperature (1.5 C) whereas for avocados, peppers and tomatoes,

the temperature of the chambers was maintained at 15C in order

to reduce the risk of chilling injury (Table 2). The levels of humidity

used were chosen based on RH that may be encountered

throughout the supply chain (Nunes et al., 2009; Lai et al., 2011

;Pelletier et al., 2011).

2.3.

Temperature

and

relative

humidity

monitoring

The temperature inside the temperature and RH controlled

rooms was monitored throughout the study using Stow Away1

XTI02 temperature loggers from (5 C to +37 C) (Onset Computer

Corporation, Pocasset, MA). The RH was monitored with Stow

Away1 RH loggers (10 to 95% RH) (Onset Computer Corporation,

Pocasset, MA).

2.4. Subjective quality evaluation

Subjective quality evaluation of strawberries was performedevery day during nine to10 d, depending on the storage temperature

and RH (Tables 1 and 2). Subjective quality evaluations for avocados,

blueberries, peppers and tomatoes were performed every two days

with the length of storage depending on the commodity and storage

regime (Tables 1 and 2). A trained individual(s) conducted the

subjective quality evaluations using rating scales and descriptors

(Tables 3–7). Color, shriveling and stem freshness were determined

subjectively using a 1 to 5 visual rating scale and rmness was

determined based on the whole fruit resistance to slightly applied

nger pressure and recorded using a 1–5 tactile rating A score of

3 was considered the limit of acceptability for sale. Thus, when fruit

were visibly deteriorated or when at least one of the subjective

quality attributes had attained a ratingof 3 or lower, the treatments

were terminated. Some of the visual quality scores and descriptorsused (e.g., shriveling and cluster tomato quality scores and

descriptors; Table 7) weredeveloped byour team based on multiple

visual observations during postharvest storage of FFVs at different

temperatures (Nunes, 2008) whereas other scores were used based

on previously published works (Tables 3–6; see footnotes). The

individual(s) that performed the visual evaluations were trained

(using FFVs and photographs) to detect minimal changes in the

appearance and rmness of the FFVs and had extensive experience

using the rating scales.

Table 2

Optimum storage conditions, cultivar, harvest location and date, and storage conditions during the humidity (RH) experiments.

Commodity Optimum temperature (C) Optimum RH (%) Cultivar Origin Harvest date Storage duration (d)a

Avocados

5–12b 85–95b ‘Simmonds’

Homestead,

Florida

July

17,

2008

20

‘Simmonds’ Homestead, Florida August 2, 2008 20

Blueberries 0.5 to 0c >90c ‘ Jubilee’ Winter Haven, Florida April 5, 2012 16

‘ Jubilee’ Winter Haven, Florida April 26, 2012 16

Pepper 7–10d 90–95d ‘Revolution’ Immokalee, Florida November 24, 2009 22

‘Revolution’ Immokalee, Florida December 1, 2009 22

Strawberries 0e 90–95e ‘Strawberry Festival’ Floral City, Florida March 5, 2012 9

‘Strawberry Festival’ Floral City, Florida February 21, 2013 9

Tomato 7–13f 90–95f ‘Success’ Wellborn, Florida June 11, 2009 22

‘Success’ Wellborn, Florida June 27, 2009 22

Storage conditions: (A) 40.1% 3.2% RH; (B) 62.0% 2.0% RH; (C) 81.0% 2.1% RH; (D) 87.9% 1.9% RH; (E) 91.6 1.4% RH; temperature of the chambers was set at 15C for

avocado,

pepper

and

tomato

and

1.5C

for

blueberry

and

strawberry.a In some cases the experiments were terminated before the end of the storage period, at the time when at least one of the visual quality attributes evaluated reached the

maximum acceptable (rating of 3).b Woolf et al. (2014); evaluated every two days.c Perkins-Veazie (2014); evaluated every two days.d González-Aguilar (2014); evaluated every two days.e Mitcham (2014); evaluated every day.f Sargent and Moretti (2014); evaluated every two days.

Table 3

Visual quality scores and descriptors for avocado.

Scores

and

description

1

2

3a 4

5

Very poor Poor Acceptable Good Excellent

Colorb Very dark skin with no traces of

green;

fruit

appears

completely

brown or black

Some green on brown or

black;

approximately

75% colored

Some yellow/black or brown

on

green;

approximately

25% colored

Darker lime green and not so

glossy;

beginning

of

color

changes;

some yellowing

Full dark lime green

color;

very

glossy;

freshly harvested

Firmnessb Extremely

soft

on

touch;

easily

yields to slight hand pressure

Soft;

whole

fruit

deforms

with slight hand pressure

Elastic;

slightly

yield

to

hand pressure

Hard

but

less

resistant

to

extreme

nger pressure

Extremely

hard;

does

not yield to extreme

nger pressure

Shriveling Skin appears extremely dry Serious shriveling Shriveling evident, but not

serious

Minor signs of shriveling Field-fresh, no signs of

shriveling

a Score of 3 was considered to be the minimum acceptable quality before avocados become unmarketable.b Modied from White et al. (2005).

M.C.N. Nunes / Postharvest Biology and Technology 107 (2015) 43–54 45


4/12

2.5. Instrumental surface color

A total of two color measurements per FFVs were taken on

opposite sides at the equatorial region. A hand-held tristimulus

reectance colorimeter (Model CR-300, Minolta Co., Ltd., Osaka,

Japan) equipped with a glass light-protection tube (CR-A33,

Minolta Co., Ltd., Osaka, Japan) was used. Color was recorded

using the CIE-L*a*b* uniform color space (CIE-Lab), L* (lightness),

a* (redness) and b* (yellowness) values. Numerical values of a* and

b* were converted into hue angle and chroma using the Minolta

Color Management Software (1996–1999CyberSoft SpectraMatch/

QC software version 3.3, CyberChrome, Inc., Stone Ridge, N.Y.).

2.6. Texture analysis

Before texture was analyzed, FFVs from each treatment were

conditioned at room temperature for approximately 1 h. Thereaf-

ter, rmness of each individual FFVs was measured using a TA.XT

plus Texture Analyzer (Texture Technologies Corp., NY, USA)

equipped with a 50 kg load cell.

Table 4

Visual quality scores and descriptors for blueberry.

Scores and description

1 2 3a 4 5


Colorb Extremely dark;

overripe or

senescent

Very dark blue/purplish Fully dark blue More blue, less

bright

Bright blue color

Firmnessc Berry rupture ontouch

Berry surface very depressed on touchbut

no

ruptureBerry surface depressed on touch,softer

than rmer

Slight depressionon

touchFirm berry, not yielding to touch

Shriveling Extremely wilted

and dry

Severe shriveling Shriveling evident but not serious Slight signs of

shriveling

Field-fresh, fruit appear very

fresh and turgid

a Score of 3 was considered to be the minimum acceptable quality before blueberries become unmarketable.b Modied from Sanford et al. (1991).c Modied from Beaudry et al. (1998).

Table 5

Visual quality scores and descriptors for pepper.


1

2

3a 4

5


Colorb Extremely dull

green or 25%

colored

Dull green or slight coloration Green but showing loss of glossiness Less bright dark green Completely bright

dark green; very

glossy

Firmnessb Extremely soft on

touch

Soft on touch, yields easily to nger

pressure; not crisp; 75% of the fruit is

soft

Fruit starts to soften and yield to

nger pressure; 50% of the fruit is

soft

Firm but less resistant to nger

pressure; 25% of the fruit is soft

Very rm, turgid and

crisp

Shriveling Extremely

shriveled

and

dry

Serious shriveling Shriveling evident, but not serious Slight signs of shriveling Field-fresh, no signs

of

shriveling

a Score of 3 was considered to be the minimum acceptable quality before peppers become unmarketable.b Modied from Lownds et al. (1994).

Table 6

Visual quality scores and descriptors for strawberry.a


1 2 3b 4 5


Color Very dark purplish-red;

extremely overripe or

senescent

Overripe; very dark

red

Fully red Fully light red Three-quarter to fully light red

Firmness Extremely soft and

deteriorated

Soft and leaky Minor signs of softness Firm but less turgid Very rm and turgid

Shriveling

Extremely

wilted

and

dry

Severe

shriveling,

fruit

is shriveled and

calyx is wilted and dry

Shriveling

evident,

fruit

and

calyx

show

evident signs of moisture loss

Minor

signs

of

shriveling,

calyx slightly wilted

Field-fresh,

fruit

and

calyx

appear very fresh and turgid

a Nunes et al. (2003c).b Score of 3 was considered to be the minimum acceptable quality before strawberries become unmarketable.



5/12

For avocado, the instrument was tted with a 76.2 mm diameter

stainless compression plate and rmness was measured on two

cheeks of each non-pealed fruit. The probe was then driven with a

crosshead speed of 2 mm s1, and the compression force was

recorded at 3.0 mm deformation ( Jeong et al., 2002). For blueberry,

three replicated samples of 30 g each were placed into three-

100 mL plastic beakers and the fruit was compressed to a depth of

30 mm using a 38 mm diameter and 20 mm high acrylic cylinder

probe. The probe was then driven with a crosshead speed of

1 mm s1, and the compression force was recorded after the probe

had compressed the fruit by 30 mm (Sanford et al., 1991). Peppers

were rst cut transversely, approximately three-quarter way from

the shoulder end and then placed, with the cut end down, on the

at

platform

of

the

texture

analyzer.

A

7.95

mm

stainless

convexprobe was centered on a shoulder lobe with two measurements

being taken on opposite lobes. The probe was then driven with a


recorded at 5.0 mm deformation (Nunes et al., 2012). Strawberryrmness was measured at the equatorial part of the fruit using a

7.95 mm stainless convex probe. The probe was then driven with a


recorded at 3.0 mm deformation (Whitaker et al., 2012). Tomato

was placed on the at surface of the texture analyzer with stem-

end down, so the pressure was applied on the blossom-end part of

the fruit. The instrument was tted with a 76.2 mm diameter

stainless compression plate and the probe was then driven with a


recorded

at

10.0

mm

deformation

(Chilson

et

al.,

2011).

2.7.

Weight

loss

Weight loss of each individual avocado, pepper or tomato and of

each individual triplicated clamshell containing 20 blueberries or

15 strawberries each was calculated from the initial weight and

after every day or every twodays during ve to 22 d, depending on

the fruit and on the temperature or RH regimes. Concentrations of

chemical constituents were expressed in terms of dry weight in

order to show the differences between temperatures or RHs that

might be obscured by differences in water content. The following

formula was used for water loss corrections: [chemical compo-

nents (fresh weight) 100 g/average dry weight (avocado = 12.9 g;

blueberry = 13.2 g; pepper = 6.1 g; strawberry = 9.4 g; tomato = 5.5

g) + weight loss during storage (g)]. The dry weight was deter-

mined by drying three weighed aliquots of homogenized fruit

tissue at 80 C, until weight stabilized.

2.8. Total anthocyanins content

Three replicated samples of blueberry or strawberry fruit were

homogenized in a laboratory blender (Waring Products Div.

Dynamics Corp. of America, New Hartford, CO) at high speed for

2 min. The homogenate (2 g) was mixed with 28 mL of 0.5% (v/v)

HCl in methanol and held for 1 h at 4 C for pigment extraction. The

occulate was removed by ltering the extract through a single

layer of facial tissue, and absorbance of the resulting liquid

containing

the

pigments

was

measured

at

520

nm

(maximumabsorbance for anthocyanins) using a BioTek microplate reader

(BioTek Instruments, Inc., Highland Park, Vermont, USA). Pigment

content was calculated using the following formula: A520

dilution factor [molecular weight (MW) of PGN/molar extinction

coef cient] where MW of PGN = 433.2 and the molar extinction

coef cient = 2.908 10,000. The amount of total anthocyanins was

expressed in terms of dry weight (g kg1) to compensate for water

loss during storage.

2.9. Total chlorophylls content

Chlorophyll content of avocado puree was extracted in the dark

with N ,N -dimethylformamide and absorbance of the ltrate

measured

at

625,

647

and

664nm

using

a

BioTek

microplatereader (BioTek Instruments, Inc., Highland Park, Vermont, USA).

Total chlorophylls content was calculated according to Moran

(1982) and the results expressed in terms of dry weight (g kg1) to

compensate for water loss during storage.

2.10.

Total

ascorbic

acid

Total ascorbic acid was quantied by mixing 2 g of strawberry,

pepper or tomato homogenates with 20 mL metaphosphoric acid

mixture (6% HPO3 containing 2 N acetic acid). Samples were then

ltered (0.22 mm) prior to HPLC analysis. Ascorbic acid analysis

was conducted using a Hitachi LaChromUltra UHPLC system with a

diode array detector and a LaChromUltra C18 2 mm column

(2 50 mm) (Hitachi, Ltd., Tokyo, Japan). The analysis was

Table 7

Visual quality scores and descriptors for greenhouse-grown cluster tomato.


1 2 3a 4 5


Color Very dark red, overripe Dark red; dull color Vivid red; loss of glossiness Light red but less glossy Light red with no trace of

green; very glossy

Firmness Extra-soft, overripe, fruit yields

very readily to slight pressure

Soft, fruit yields readily

to slight pressure

Firm, fruit yields slightly to

moderate pressure

Hard, fruit yields only slightly

to considerable pressure

Extra hard, fruit does not yield

to considerable pressure

Shriveling Extremely shriveled and dry; fruit

appears old and deteriorated

Severe shriveling Shriveling evident, but not

serious

Slight signs of shriveling Field-fresh, no signs of

shriveling

Stem freshness Stem is completely dry

and dark brownish-

green

Stem is dry, wilted and

brownish-green

Signs of dryness are evident

but not objectionable

Stem appears slightly less

green and less turgid

Stem is

very

bright

green and

turgid

a Score of 3 was considered to be the minimum acceptable quality before cluster tomatoes become unmarketable.



6/12

performed under isocratic mode at a ow rate of 0.5 mL/min with a

detection of 254 nm. Sample injection volume was 5 mL, each with

duplicate HPLC injections. Mobile phase was buffered potassium

phosphate monobasic (KH2PO4, 0.5%, w/v) at pH 2.5 with

metaphosphoric acid (HPO3, 0.1%, w/v). The retention time of

the ascorbic acid peak was 2.5 min. After comparison of retention

time with the ascorbic acid standard, the peak was identied. The

amount of total ascorbic acid content in strawberry was quantied

using calibration curves obtained from different concentrations

(0.01 g L 1, 0.02 g L 1, 0.03 g L 1, 0.05 g L 1, 0.10g L 1, 0.15 g L 1,

0.20 g L 1 and 0.30 g L 1) of ascorbic acid standards. Total ascorbic

acid content was expressed in terms of dry weight (g kg1) to

compensate for water loss during storage.

2.11. Statistical analysis

The SigmaPlot Version 12.0 (Systat Sotware, San Jose CA) was

used for the analysis of the data. Since there were no signicant

differences between harvests in respect to the strength and

signicance of the relationships, data from the two harvests was

combined. The strength of the relationship between variables (i.e.,

subjective versus quantitative) was measured using the Pearson

correlation coef cient (r ). The coef cient of determination (r 2) and

the signicance of the relationship between variables was

expressed by probability levels ( p = 0.05). A linear regression

model was applied to the data collected and was used to describe

the relationship between variables.

3. Results and discussion

3.1. Correlation between subjective color and L *, hue angle,

anthocyanins, chlorophyll and ascorbic acid contents

Subjective color was signicantly correlated to L* values for all

the FFVs evaluated, regardless of the storage conditions, except for

strawberry stored under different RH regimes (Table 8; Fig. 1). In

general, as color ratings decreased, L* values also decreased,

meaning that the surface color of the FFVs became darker.

Similarly, with the exception of blueberry, there was a signicant

correlation between color ratings and hue angle for all the fruit

evaluated (Table 8; Fig. 1). As color ratings decreased, hue angle

values also decreased; for avocado and pepper, decreased in hue

angle corresponded to changes in color from bright dark lime green

to a brownish-green color (Tables 3 and 5; Fig. 1). In pepper,

decreased in hue angle corresponded to changes in color from

bright dark-green to dull green (Table 3).In strawberry and tomato,

decreased in hue angles corresponded to changes in the surface

color from bright light-red to a deep dull red (Tables 6 and 7; Fig.1).

Table 8

Pearson correlation coef cient (r ), coef cient of determination (r 2), linear regression equation, and signicance of the relationship ( p) between subjective color and:

instrumental color coordinates (L* value and hue angle), anthocyanins, chlorophyll and ascorbic acid content contents for avocado, blueberry, pepper, strawberry and tomato

stored at different temperature or humidity regimes.

Commodity (treatment) Color vs L* value Color vs hue angle Color

vs anthocyanins

(g kg1)

Color

vs chlorophyll

(g kg1)

Color vs ascorbic acid

(g kg1)

Avocado

(RH) r = 0.602 r = 0.906 NM p > 0.05 (NS) NM

r 2 = 0.363 r 2= 0.821

y = 4.409 x + 68.321 y =13.870 x + 56.882

p


7/12

Nunes et al. (2013) also showed that there was a highly positive

correlation between banana visual color and hue angle values and

suggested that color can be a reliable way to estimate visual color

changes. Visual evaluation of browning in fresh-cut nectarines was

also found to be highly correlated to color parameters, particularly

with b* and chroma (Pace et al., 2011). In addition, color changes

have also been shown to be strongly correlated with sugar content

in banana (Nunes et al., 2013). Therefore, the use of color charts

(e.g., banana industry) when used by trained individual(s) can be

an easy and fast way to estimate FFVs quality attributes that are

associated with changes in surface color.

Color ratings were also signicantly correlated with anthocya-

nin contents, particularly in strawberry (Table 8; Fig. 2). As color

ratings decreased, anthocyanin content also decreased, meaning

that fruit with a deep purplish-blue or purplish-red color had

lower anthocyanin contents than fruit showing a bright light blue

or red color. Although in avocado the relationship between color

and chlorophyll content was not very strong, there was a

signicant correlation between color ratings and chlorophyll

content in avocado stored at different temperatures (Table 8).

Therefore, as color ratings decreased (color changed from green to

brown) chlorophyll content also decreased. Proulx et al. (2010) also

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

L * v a

l u e

30

35

40

45

50

55

60

65

70

Avocado (RH)

Tomato (RH)

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

H u e a n g

l e

60

80

100

120

140

160

Avocado (RH)

Avocado (T)

Pepper (RH)

Color rating (1-5)

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

H u e a n g

l e

0

20

40

60

80

100

120

Strawberry (RH)

Strawberry (T)

Fig. 1. Scatter plots and linear regression lines showing the relationship between

subjective color (1 = very poor; 5 = excellent) and L* value for avocado and cluster

tomato and the relationship between subjective color and hue angle for avocado,

pepper and strawberry stored at different temperature and RH regimes. RH =

relative humidity; T = temperature.

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

T o

t a l a n

t h o c y a n

i n s

( g k g

- 1 )

0

1

2

3

4

5

6

7

8

Strawberry (RH)

Strawberry (T)

Blueberry (RH)

Color rating (1-5)

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

T o

t a l a s c o

r b i c a c

i d ( g k g

- 1 )

0

2

4

6

8

10

12

14

Strawberry (RH)

Strawberry (T)

Pepper (RH)

Fig. 2. Scatter plots and linear regression lines showing the relationship between

subjective color (1 = very poor; 5 = excellent) and total anthocyanins content for

strawberry and blueberry and the relatioship between subjective color and total

ascorbic acid content for strawberry and pepper stored at different temperature and

RH

regimes.

RH

=

relative

humidity;

T

=

temperature.



8/12

showed that in snap beans there was a signicant correlation

between visual color scores and chlorophyll content, that is,

chlorophyll content decreased as the color turned from bright

green to yellowish-green. In avocado, the reason for the weak

relationship between visual color and chlorophyll content might

be related to the fact that color ratings were based on the color of

the avocado skin whereas chlorophyll content was measured in the

pulp of the fruit.

Ascorbic acid content was signicantly correlated with color

ratings particularly for pepper and strawberry (Table 8; Fig. 2).

Therefore, as color ratings decreased, ascorbic acid content also

decreased. After harvest, decline in ascorbic acid content of FFVs

usually occurs rapidly and it normally parallels the increase in

water loss, particularly when FFVs are exposed to adverse

environmental conditions. Excessive loss of water causes tissue

damage and results in deterioration of the overall quality; with cell

wall disruption promoting the release of ascorbate oxidase and

subsequent oxidation of ascorbic acid (Proulx et al., 2010).

3.2. Correlation between subjective shriveling and weight loss

For avocado, blueberry, pepper and strawberry, shriveling was

negatively correlated with weight loss whereas in tomato weight

loss was negatively correlated with shriveling and particularly

with stem freshness, regardless of the storage conditions (Table 9;

Figs. 3 and 4). Therefore, as ratings for shriveling or stem freshness

decreased weight loss signicantly increased. Nunes and Emond

(2007) evaluated several fruits and vegetables and also found a

highly signicant correlation between weight loss and subjective

color, rmness and shriveling/wilting. Thus, as weight loss

increased during storage, rmness decreased, and wilting,

shriveling or browning increased. Lownds et al. (1994) also

reported that accidity in peppers appeared to be directly

associated with water loss whereas color ratings paralleled

differences in water loss rates suggesting a direct relationship

between these variables.

3.3.

Correlation

between

subjective

rmness

and

instrumental

texture

Subjective rmness was signicantly correlated with instru-

mental texture, regardless of the fruit and the storage regimes used

(Table 9; Fig. 5). As subjective rmness ratings decreased, values

for quantitative texture also decreased. In a previous study, a

signicant correlation was found between banana subjective

rmness (1–5 rating scale) and analytical texture, suggesting that

rmness on touch can be used to estimate softening (Nunes et al.,

2013). White et al. (2005) produced softening curves for avocado

based on hand rmness scores versus quantitative texture

Table 9

Pearson correlation coef cient (r ), coef cient of determination (r 2), linear regression equation, and signicance of the relationship ( p) between

subjective shriveling or steam freshness and weight loss and between subjective rmness and instrumental texture for avocado, blueberry, pepper,

strawberry and tomato stored at different temperature or humidity regimes.

Commodity (treatment) Shriveling/stem freshnessavs weight loss (%) Firmness vs texture (N )

Avocado (RH) r = 0.938 r = 0.725

r 2= 0.880 r 2= 0.523

y= 3.807 x + 21.751 y = 28.002 x 48.454

p


9/12

measured with a penetrometer. The curves showed that as hand

rmness decreases from very hard (rating of 0) to very soft (rating

of 7) the force required to puncture the fruit also decreased.

3.4. Data tting

A linear regression model was applied to all datasets, and thus

the following equation was used to summarize the relationship

between variables: [ y = ax + b]; where y = quantitative attribute and

x = subjective quality attribute score (1–5) (Tables 8 and 9). Graphic

representations were shown only for those relationships in which r

or r 2 0.6 and p


10/12

Fig. 6. Color chart for strawberry showing photographs of visual quality deterioration with subjective quality ratings (1 = very poor; 3 = acceptable; 5 = excellent) and

correspondent descriptors.



11/12

http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0135http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0135http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0135http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0130http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0130http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0130http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0125http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0125http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0125http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0120http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0120http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0120http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0115http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0115http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0115http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0110http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0110http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0110http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0105http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0105http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0100http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0100http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0100http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0095http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0095http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0090http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0090http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0085http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0085http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0085http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0085http://ucce.ucdavis.edu/files/datastore/234-49.pdfhttp://ucce.ucdavis.edu/files/datastore/234-49.pdfhttp://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0075http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0075http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0075http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0070http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0070http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0065http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0065http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0060http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0060http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0060http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0055http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0055http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0055http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0050http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0050http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0050http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0040http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0040http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0040http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0035http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0035http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0035http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0030http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0030http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0030http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0030http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0025http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0025http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0025http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0025http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0020http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0015http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0015http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0015http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0010http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0010http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0010http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0005http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0005http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0005http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0005


12/12

Nunes, M.C.N., Delgado, A., Emond, J.P., 2012. Quality curves for green bell pepper(Capsicum annum L.) stored at low and recommended relative humidity levels.Acta Hort. 945, 71–78.

Nunes, M.C.N., Yagiz, Y., Emond, J.P., 2013. Inuence of environmental conditions onthe quality attributes and shelf life of ‘Goldnger’ bananas. Postharvest Biol.Technol. 86, 309–320.

Nunes, M.C.N., 2008. Color Atlas of Postharvest Quality of Fruits and Vegetables.Blackwell Publishing, Iowa, USA.

Pace,

B.,

Cefola,

M., Renna,

F.,

Attolico,

G.,

2011.

Relationship

between

visualappearance and browning as evaluated by image analysis and chemical traits infresh-cut nectarines. Postharvest Biol. Technol. 61, 178–183.

Pelletier,

W.,

Brecht,

J.K.,

Nunes,

M.C.N.,

Emond,

J.P.,

2011.

Quality

of

strawberriesshipped by truck from California to Florida as inuenced by postharvesttemperature

management

practices.

HortTechnology

21

(4),

482–493.Perkins-Veazie, P., 2014. Blueberry. In: Gross, K.C., Wang, C.Y., Saltveit, M. (Eds.), The

Commercial Storage of Fruits, Vegetables, and Florist and Nursery Stocks.Agriculture Handbook 66. USDA-ARS http://www.ba.ars.usda.gov/hb66/blueberry.pdf (accessed 01.29.15.).

Proulx,

E.,

Nunes,

M.C.N.,

Emond,

J.P.,

Brecht,

J.K.,

2005.

Quality

attributes

limitingpapaya postharvest life at chilling and non-chilling temperatures. Proc. Fla.State Hort. Soc. 118, 389–395.

Proulx, E., Yagiz, Y., Nunes, M.C.N., Emond, J.P., 2010. Quality attributes limiting snapbean (Phaseolus vulgaris L.) postharvest life at chilling and non-chillingtemperatures. HortScience 45, 1238–1249.

Ressureccion, A.V.A., Shewfelt, R.L., 1985. Relationships between sensory attributesand objective measurements of postharvest quality of tomatoes. J. Food Sci. 50,1242–1245.

Safner, R., Polashock, J., Ehlenfeldt, M., Vinyard, B., 2008. Instrumental and sensoryquality characteristics of blueberry fruit from twelve cultivars. Postharvest Biol.Technol. 49, 19–26.

Sanford, K.A., Lidster, P.D., McRae, K.B., Jackson, E.D., Lawrence, R.A., Stark, R.,Prange, R.K., 1991. Lowbush blueberry quality changes in response tomechanical

damage

and

storage

temperature.

J.

Am. Soc.

Hort.

Sci.

116,

47–51.Sargent, S.A., Moretti, C.L., 2014. Tomato. In: Gross, K.C., Wang, C.Y., Saltveit, M.

(Eds.), The Commercial Storage of Fruits, Vegetables, and Florist and Nursery

Stocks.

Agriculture

Handbook

66.

USDA-ARS

http://www.ba.ars.usda.gov/hb66/tomato.pdf (accessed 01.29.15.).

Whitaker,

V.M.,

Chandler,

C.K.,

Santos,

B.M.,

Peres,

N.,

Nunes,

M.C.N.,

Plotto,

A.,

Sims,C.A., 2012. Florida MedallionTM (‘FL 05-1070) Strawberry. HortScience 47,296–298.

White, A., Woolf, A., Ofman, P., Arpaia, M.L., 2005. The International Avocado QualityManual, HortResearch, The Horticultural and Food Research Institute of NewZealand

Limited.

Auckland,

New

Zealand.Woolf, A.B., White, A., Arpaia, M.L., Gross, K.C., 2014. Avocado. In: Gross, K.C., Wang,

C.Y., Saltveit, M. (Eds.), The Commercial Storage of Fruits, Vegetables, and Floristand Nursery Stocks. Agriculture Handbook 66. USDA-ARS http://www.ba.ars.usda.gov/hb66/avocado.pdf (accessed 01.29.15.).

http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0140http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0145http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0150http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0155http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0160http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0165http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0170http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0175http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0180http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0180http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0180http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0180http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0180http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0180http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0180http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0180http://refhub.elsevier.com/S0925-5214(15)30011-9/sbref0180http://refhub.elsevier.com/S

Documents

Do Nascimento, 2015