вход по аккаунту



код для вставкиСкачать
Journal of the Science of Food and Agriculture
J Sci Food Agric 80:275±283 (2000)
Temperature sums experienced before harvest
partially determine the post-maturation juicing
quality of oranges grown in the Murrumbidgee
Irrigation Areas (MIA) of New South Wales
Ronald J Hutton1* and Joseph J Landsberg2
NSW Agriculture, Yanco Agricultural Institute, Yanco, NSW 2703 Australia
CSIRO Land and Water, Pye Laboratory, GPO Box 1666, Canberra, ACT 2601, Australia
Abstract: Fruit acid levels are important to maturity standards for both fresh fruit and juice
acceptability for processing oranges. Citrus fruit maturity standards are based on sweetness (Brix),
titratable acidity (% acid) and their ratio. Harvest of Navel oranges for fresh juice can extend for up to
3 months (August±October) and for as long as 7 months (November±June) for Valencia oranges after
they ®rst reach acceptable fresh market maturity. Juice processors encourage `late hanging' of fruit on
trees to secure a continuous supply of fresh juice. During this time, gradual changes occur in % acid,
Brix and juice content. This study tested the hypothesis that a signi®cant proportion of the variation in
fruit quality late in the harvest season could be accounted for by environmental conditions during the
period that the fruits were left on the trees. Fruit juice quality was assessed in terms of Brix and % acid
at the time of harvest for 19 000 fruit deliveries during a 9 year period from 1988 to 1996. The
relationship between quality (Brix and acid content) and temperature sums (effective heat unitsÐ
EHUs) for the period the fruits were held on the trees was tested. A temperature sum (`day degree')
model for predicting relative changes in acid and Brix content of late-harvested fruit was developed for
both Valencia and Navel oranges. The juice acid relationship with EHUs was stronger than the Brix
relationship with EHUs. In addition, the seasonal `behaviour' of % acid was more consistent than that
of Brix in both Valencia and Navel oranges. A linear reduction in % acid with increasing EHUs was
evident in both varieties. Estimates of fruit quality (Brix/acid ratio) at harvest time should be made
using the equations describing changes in % acid with EHUs and the relationship between Brix and %
acid to determine Brix. These equations have potential use in predicting the internal quality of citrus
fruit during extended harvest periods. EHU-based fruit quality estimates could determine harvest
schedules that enable growers to achieve the optimum factory price for their fruit whilst maximising
returns to processors through more ef®cient product utilisation.
# 2000 Society of Chemical Industry
Keywords: oranges; juice quality; Brix; titratable acidity; ratio; climatic modelling; temperature sums
Citrus is a major horticultural commodity produced in
the MIA. In the 1996/97 growing season, 203 188 tons
of sweet oranges were produced under irrigation in the
region. This constitutes approximately one-third of
the national production. The predominant variety
grown is Late Valencia (160 000 tons), of which about
75% is used for juice, with the balance being directed
to domestic and export fresh fruit markets. Navel
oranges (43 188 tons) are the other major sweet orange
variety and are speci®cally grown for the fresh fruit
markets. A small proportion of this crop is also used for
juicing to maintain the fresh character of juice
produced from either late-harvested Valencia oranges
from the previous season or from new-season Valencia
fruit that is still too acid to be used without blending.
Maturity standards for citrus fruit are legally
enforced in many countries, varying for different
species and cultivars,1 so the need to predict harvest
maturity for all market outlets is of critical concern in
meeting protocols, especially in export markets. For
both fresh citrus and processed products each growing
area tends to seek standards that will capitalise on its
strengths and minimise its weaknesses in the marketplace, but because climate overrides cultural management practices, it will always be dif®cult to specify
particular maturity standards.2 Nevertheless, maturity
standards are used to determine the earliest time that
fruit can be harvested.
The citrus fruit maturation process is characterised
by gradual changes in juice content and in some of its
constituents. There is a decline in titratable acidity
* Correspondence to: Ronald J Hutton, NSW Agriculture, Yanco Agricultural Institute, Yanco, NSW 2703, Australia
(Received 21 July 1999; accepted 27 September 1999)
# 2000 Society of Chemical Industry. J Sci Food Agric 0022±5142/2000/$17.50
RJ Hutton, JJ Landsberg
(% acid) brought about by decomposition of citric
acid, which is the principal organic acid of citrus
juice.3 Concurrently there is an increase in total
soluble sugars (Brix). The Brix/acid ratio is a sensitive
measure of maturation and is commonly agreed as a
`maturity index' for citrus fruit,4,5 although it actually
represents two biochemically independent metabolic
processes. The acid levels are important to maturity
standards for both fresh fruit and juice acceptability for
processing,6±8 as acid is a key determinant of fruit
The minimum maturity standard for fresh fruit
marketing in Australia is a Brix/acid ratio of 7:1;
processors seeking oranges for fresh juice are reluctant
to accept fruit deliveries before Brix/acid ratios of 11:1
are achieved. In the temperate inland production
regions of Australia such as the MIA, ¯owering in both
Navel and Late Valencia oranges occurs in late spring
(mid-October) and usually lasts for 3±4 weeks, with
fruit set being determined by mid-December.10,11
Following fruit set, the normal expected time course
for the development of acceptable maturity in oranges
depends on both variety and seasonal weather conditions. The 7:1 ratio is attained in Navel oranges
between May and July of the following year (7±
9 months after ¯owering), while for Valencia oranges
the required quality standard is not reached until
September±November (11±13 months after ¯owering). Very little of the Navel orange fruit is used for
processing because of the presence of limonin, which
makes the juice bitter,12 and it is only used for
blending. In contrast, most of the Valencia crop is
used for processing into fresh juice. However, Valencia
oranges can be left on the trees for periods of up to
7 months after ®rst reaching acceptable quality levels
for marketing as fresh fruit, 21 months after ¯owering.
The annual growing cycle for citrus is therefore spread
across 2 years in southern hemisphere citrus-producing areas, and time-based regressions of Brix and acid
content do not hold for late-harvested fruit, whereas in
the northern hemisphere the majority of phenological
development in citrus occurs during a single calendar
The ability to predict the time when acceptable
maturity standards might be reached, and to estimate
the Brix/acid quality attributes for a given season,
offers potential to maximise quality of both marketable
and processed product presented to consumers and
therefore increase the ef®ciency of crop utilisation.
Chandler and Nicol13 established a basal date/maturation model to describe maturation in oranges using
time-dependent equations based on the maturation
period to the point in time when fruits ®rst reach
harvest maturity. They also established a secondary
relationship between Brix and juice acidity which
allowed the development of direct separate relations
between Brix and acid for the period of fruit growth to
harvest maturity. However, after this time the above
time-based relationships could not be applied.
Similar investigations to model time-dependent
changes in fruit quality with the progress of maturation
have been reported for early-season citrus grown in
Florida and California.4 This work demonstrated that
simple linear models were capable of correlating the
change in Brix and Brix/acid ratio for early-season
citrus varieties in a given year, but that the relationship
broke down for long-season (Valencia) oranges which
developed over a period in excess of 12 months.
Correlating changes in citrus fruit quality with climatic
data14 to predict time to reach harvest maturity has
also been attempted for California Navel and Valencia
oranges with some degree of success.
This paper presents the results of an analysis of fruit
juice acidity and sugar content of Valencia and Navel
oranges after they reach maturity, based on 19 000
fruit deliveries made to National Foods Juice Limited
in Leeton, NSW during a 9 year period from 1988 to
Our analysis was based on the hypothesis that a
signi®cant proportion of the variation in fruit quality
could be accounted for by environmental conditions
experienced during the period that the fruits were on
the trees. Temperature sums over time, expressed in
terms of effective heat units (EHUsÐBower J,
personal communication), were back calculated from
harvest date for the time (months) fruits were on the
trees. These values (abbreviated to EHUs in the text
and Tsum for algebraic purposes) were regressed (all
years) against mean monthly Brix and % acid content
to establish workable relationships that could predict
the decline in fruit juice quality of late-hung fruit, after
having reached optimum quality levels.
From the relationships developed, it should be
possible to calculate how long orange harvest could
be extended after reaching a minimum acceptable
maturity standard before the onset of inferior juice
quality. This could be tested and re®ned on the basis
of results from experiments deliberately set up to
analyse fruit quality at a speci®c location in relation to
EHUs generated from temperature data collected at
the site.
A quality prediction model based on acid decline in
relation to temperature sums for the full fruit-growing
period can be used to select growing locations where
acceptable fruit acid and Brix contents might be
sourced to extend the harvest season. This has the
additional bene®t, from a practical point of view, of
allowing forward projection to predict the effects of
leaving fruits on the trees for varying lengths of time
and to estimate probable Brix/acid ratio (see `A fruit
quality prediction model').
Fruit data
National Foods, which processes approximately 35%
of the citrus crop produced in the MIA region, allowed
access to its long-term juice quality data records for
fruit receivals of both Navel and Late Valencia
oranges. This database provided the basis for develJ Sci Food Agric 80:275±283 (2000)
Effect of temperature sums on juicing quality of oranges
Table 1. An example of fruit quality data extracted
from the National Foods fruit delivery database
Locality Variety
% acid
oping relationships between climatic conditions and
the maturity characteristics of Brix and acidity of
orange fruit.
All fruit was harvested using standard commercial
picking and transport procedures.15 Bulk fruit delivery
by truck from the farms to the juicing plant always
occurred within 1 week of harvest.
A 20 kg subsample was taken from each fruit
delivery to the processor for internal juice quality
determination using standard laboratory procedures in
its testing laboratory.
The processor pays growers for their fruit based on
mean sugar, acid and juice content of each delivery to
the factory. An example of the fruit quality database
generated by this procedure is given in Table 1. Other
information of no relevance to this analysis, such as
grower and delivery identi®cation, sample size, %
juice, etc, was also provided. There were thousands of
such records, which in many cases amounted to
hundreds per month.
To reduce the data set to manageable numbers, we
used computer spreadsheet procedures to sort them
and select, initially, results for Valencia oranges
delivered from 31 farms in one growing sub-location
within the MIA (site 1). The Brix and % acid data were
averaged by months. A typical example of the variance
of these average values is given in Table 2. Analysis of
the Valencia fruit quality data at this location showed
that the approach adopted was worth pursuing, so we
applied the same procedures to all the Valencia and
Navel orange data available to us.
The analysis was in terms of effective heat unitsÐ
temperature sums over time (see next subsection)Ðso
we needed to estimate the period that fruits were on
the trees. The harvest period for Valencia oranges
grown in the MIA extends from 10 to 21 months after
®rst reaching a minimum maturity standard (Brix/acid
ratio) of 7:1. The median fruit age at harvest is about
16 months. Fruit age of samples was calculated from
the time of ¯owering and fruit set in October11 up to
(and inclusive of) the month of harvest (time of fruit
delivery). Fruit delivered for processing in October
would therefore have been assumed to be 13 months
old and fruit delivered in April the following year
would therefore be 19 months old. The oldest fruits are
those harvested and delivered in June (21 months) and
the youngest are those harvested in about August/
September (10±11 months), so there is no danger of
This procedure will, on average, slightly overestimate the period the fruits were on the trees, and
there are several other sources of error associated with
the approach. For example, fruit may be held on
growers' properties for several days before being
transported to the processing plant, where fruit
samples for quality testing may again be stored for
several days before being analysed. Such delays may
cause recorded Brix levels to be slightly higher than
straight-picked fruit, and grower deliveries could also
to be recorded in month n ‡ 1 instead of month n in the
database. In addition, we used temperature records
from a single regional centre rather than from the
actual fruit production locations. However, we concluded that the considerable amount of extra work that
would be involved in attempting to account for the age
of each load of fruit to the nearest day, as well as
temperature records at each location, would not be
Meteorological data and analysis
We tested the idea that fruit quality might be strongly
affected by short-wave (solar) radiation received
during growth, since radiation determines photosynthesis, which would be expected to in¯uence the amount
of sugars available to the fruit. There were indeed
strong and signi®cant relationships between measurements of fruit quality and total incoming radiation
during growing periods, but these were no better than
the relationships between temperature sums and fruit
quality. Furthermore, there was a typically strong
positive correlation between average monthly solar
radiation and average monthly maximum temperature
for the study period. Therefore, because temperature
data are more readily available and more widely
understood than radiation data, we concentrated on
the relationships between temperature and fruit
The meteorological records used were obtained
from a registered Bureau of Meteorology site located
in the centre of the citrus production area at the
CSIRO Land and Water Research Station, Grif®th
(34.18S, 146.04E). We assumed that these (daily)
Table 2. Typical monthly means and standard
errors (SEs) for the juice quality attributes Brix and
% acid in Valencia and Navel oranges. The valuesa
are means of between 20 and 80 deliveries
J Sci Food Agric 80:275±283 (2000)
% acid
% acid
October 1993 11.09 0.19 1.31 0.04
July 1988
12.36 0.15 1.11 0.02
April 1994
10.26 0.09 0.85 0.01 August 1991 13.21 0.11 0.99 0.02
Mean 2 SE 95% con®dence limits.
RJ Hutton, JJ Landsberg
records, covering the complete period of the study
from 1987 to 1996, provided a good approximation of
the prevailing weather conditions throughout the
region. This assumption is supported by comparison
of meteorological data recorded at the Yanco Agricultural Institute (34.35S, 146.24E) which is located
at the eastern end of the citrus-growing area. Total
radiation and temperature sums over the growth
periods of the fruit deliveries received at the processing
plant in Leeton were calculated from the CSIRO data.
Temperature determines the rate of biochemical
processes in plant growth. The rate of fresh weight and
dry weight gain (accumulation of sugars) in citrus16 is
closely associated with the summed product of
(temperature time). These `day degrees' provide a
surrogate for the integral of the biochemical processes
operating at a rate determined by temperature, and
can therefore be expected to be related to the state or
condition of the fruit at harvest maturity.17 For
oranges it is accepted that root, shoot and fruit growth
and development slow considerably below 13 °C, so
this is treated as a threshold18 and temperatures below
13 °C were discounted when calculating `day degrees',
which are commonly referred to as effective heat units.
This term, abbreviated to EHUs in the text and Tsum
for algebraic purposes, will be used throughout this
The procedure outlined above is described by the
‰sŠ…t ˆ harvest† ˆ f
…T ÿ 13†t
where [s] is the sugar concentration or % acid, T
denotes the daily average temperature, t is the time
(days) and d[s]/dt = f(T). We calculated the EHUs
…T ÿ 13†t) over the period the fruits
(Tsum ˆ f
Figure 1. Average Brix (% soluble solids) for all Valencia orange fruit
deliveries processed during the years 1988–1996, plotted against effective
heat units (EHUs) for the fruit-growing periods. The two sets of data
represent site 1 (rectangles) and all other sites (diamonds).
The best fit linear equation is Brix = 12 ÿ 0.0006Tsum (r2 = 0.24).
Dt = 9 months, the advantage over using the whole
period for which the fruits were on the trees was
statistically marginal, so we have focused on the full
fruit-growing period. The bene®t of this, from a
practical point of view, is to project estimates of
probable fruit maturity standard (see `A fruit quality
prediction model').
The data in Fig 1 show that% total soluble solids
(Brix) falls slightly as EHUs increase from 1500 to
3500. Each point represents the mean Brix value for all
the fruit deliveries processed in any one month, ie fruit
of the same age, plotted against EHUs for the
appropriate fruit-growing periods. The data are for
the fruit-growing seasons from 1988 to 1996. Site 1
data (rectangles) were analysed separately from the
data from all other sites (diamonds), but there were no
signi®cant differences between the data sets and they
can be considered as groups of the same set. The
equation of the regression line is
were on the trees and analysed fruit quality in relation
to the results.
Valencia oranges
The dominant process in the early stages of citrus fruit
growth is cell division. In Valencias, cell enlargement
may be completed only 6±7 months after ¯owering;10
then the ripening and maturation phase follows for the
next 8±12 months. Therefore the early stages of fruit
growth might be expected to have little in¯uence on
®nal fruit quality, so we calculated EHUs backwards
from harvest.
We examined the possibility that EHUs over
intervals (Dt) less than the full period from ¯owering
to harvest may have greater in¯uence on fruit quality
than EHUs over the full fruit-growing period. The
relationships improved as the interval increased, and
the best results were obtained using an interval (Dt) of
either 9 months or the whole fruit-growing period.
Although the best relationships were obtained using
Brix ˆ 12:00 ÿ 0:0006Tsum
…r 2 ˆ 0:24†
The linear relationship in Fig 1 between Brix and
EHUs is statistically highly signi®cant, but it accounts
for only 24% of the variation in Brix. Therefore, using
this relationship to predict the Brix value of a
particular batch of oranges grown in a particular
season would not necessarily lead to an accurate
prediction of the Brix value measured at the processing
plant. The variation in the general relationship re¯ects
the considerable year-to-year variation in Brix values
in relation to EHUs. This is illustrated in Fig 2, where
Brix values for each growing season (site 1 data),
`expanded' to better illustrate the range, have been
plotted against EHUs. (The Brix values were expanded by subtracting 9 (the highest integer smaller
than the lowest Brix value) and multiplying by 10, so a
real value of 11.2 `expands' to (11.2-9) 10 = 22.) The
lines drawn through the points for the November
1989±May 1990 season and the October 1993±May
1994 season have slopes of 0.0061 and ÿ0.0117
respectively. The positive slope of the 1989/90 data is
J Sci Food Agric 80:275±283 (2000)
Effect of temperature sums on juicing quality of oranges
Figure 2. Mean Brix (% soluble solids) values for the months of each
growing season, plotted on a scale expanded by (Brix ÿ 9)/10 to show the
scatter, against effective heat units (EHUs). The lines are drawn through
the data for the November 1989–May 1990 season (upper line) and the
October 1993–May 1994 season (lower line). The graph shows that there is
high inter-seasonal variation in the general relationship between Brix and
EHUs, so it would not provide accurate predictions of Brix values for the
processor in any particular season.
unique in this 9 year data set and can be attributed to
the presence of both a very light mature crop (off-crop
year) and a heavy set in the developing crop. The
seasonal conditions were different from the other years
of the study, and the Brix levels continued to increase
slightly throughout the delivery period despite the fruit
being between 14 and 19 months old. The occurrence
of abnormally warm weather conditions in the spring
and early summer of 1989 followed by a wetter late
summer and warmer-than-average autumn during
1990 also contributed to this anomalous trend in Brix
during this season.
The slopes of the data sets for the other growing
seasons also varied considerably, but in all cases the
slope was negative for day degree accumulations from
1100 to 3000 units.
Plotting the Brix values against the time (months)
that the fruits were on the trees gave a relationship
which was statistically equivalent to that between Brix
and EHUs.
The relationship between % acid and EHUs was
better than that between Brix and EHUs, but only
after about 1700 EHUs (Fig 3). Here again the data
for site 1 and all other sites are shown as the analogue
of Fig 1. The two data sets are not signi®cantly
different and can be ®tted by a single line:
% acid ˆ 1:62 ÿ 0:000275Tsum
…r 2 ˆ 0:5†
Variation in EHUs accounts for more than 40% of
the variance in % acid. Analysis of inter-seasonal
variation in % acid in relation to EHUs indicated that
this was not a signi®cant factor in the way fruit acids
change. The general relationship across all sites and
seasons gives a reasonable prediction of the % acid
values likely to be measured in the processing plant at
the time of fruit delivery.
The relationship between % acid and time conJ Sci Food Agric 80:275±283 (2000)
Figure 3. Average % acid for all Valencia orange fruit deliveries processed
during the years 1988–1996, plotted against effective heat units (EHUs) for
the fruit-growing periods. The two sets of data represent site 1(rectangle)
and all other sites (decimals). The (overall) best fit linear equation is %
acid = 1.62 ÿ 0.000275 Tsum (r2 = 0.5), although the graph indicates the
possibility of a threshold at about 1700 EHUs.
tained much more scatter and less information than
that between % acid and EHUs. This is not surprising,
since the rate at which acid levels decline from their
typical peak during the middle of fruit development
involves both dilution (rate of water accumulation in
the juice sacs) and metabolism,19 both of which are
Navel oranges
The analytical procedure used for Valencias was also
applied to Navel oranges from a number of sites
harvested between 1988 and 1996 and delivered to the
processing plant at Leeton.
Plotting Brix against EHUs gave a scatter of points
with no clear trend. A line ®tted through the data for
all growing locations over the fruit-growing period
from ¯owering to harvest (processing) had a slight
negative slope (ÿ0.0001), but it accounted for only
5% of the variance. Therefore we conclude that the
development of total soluble sugars in Navels is not
directly associated with temperature. This is consistent
with other reports16 that the rate of sucrose entry into
developing citrus fruit proceeds slowly over an
extended period against an ascending gradient in train
with fresh weight gain, but is later reversed during the
®nal period of fruit maturation when water in¯ow to
juice sacs stops.
The relationships between % acid and EHUs in
Navels were the same as in Valencias over the whole
period from fruit set to harvest, but not over shorter
periods. Over the full period the relationship in Navels
was similar to that in Valencias (Fig 4); the equation
describing it is
% acid ˆ 2:17 ÿ 0:0006Tsum
…r 2 ˆ 0:23†
Note that the slope of the line is greater than that
established for Valencias, so the rate of acid reduction
per EHU is, on average, greater in Navels than in
Valencias. There was also evidence of potentially
RJ Hutton, JJ Landsberg
Figure 4. % levels in Navel oranges plotted against temperature sums
calculated for the whole period the fruits were on the trees. As in the case of
Valencias, the graph indicates that there may be a threshold value of EHUs
below which they exert little influence on acid levels in the fruits. The line
drawn through the points is described by the equation %
acid = 2.17 ÿ 0.0006 Tsum (r2 = 0.23).
drawn through the falling pointsÐsimilar to the slope
of the line describing the relationship between EHU3
and per cent acid for the whole fruit growth period (Fig
4). This analysis indicates that in the case of Navels the
temperatures in the winter and early spring period
immediately before harvest are of paramount importance in determining the % acid in the fruits.
With these data it was not possible to examine the
effects of temperatures during the early phase of fruit
growth on their quality at harvest; our assumption that
all fruits are set in October leads to only a single Tsum
value for the ®rst 3 months of each growing season.
Plotting the variable values of fruit acid content that
result from different periods on the trees against
temperature sums for the ®rst 3 months of fruit growth
provides no useful information.
A fruit quality prediction model
It has long been recognised,13,20 (also Harrison M,
personal communication) that Brix and per cent acid
are correlated. This is supported by the data analysed
here, although the correlations are not strong. For
Valencia oranges the relationship is described by the
higher amounts of acid accumulation in early stages of
fruit development in Navel oranges.
When % acid was plotted against EHUs for the
8 months before harvest (EHU8), an inverse relationship was obtained (Fig 5). This coincides with the
onset of rapid fresh weight gain and water ingress to
juice sacs of developing Navel oranges. The higher %
acid values at higher EHU values come from fruits
harvested early (July/August), which had high juice
acidity despite relatively high values of EHU8,
accumulated in the late summer. The fruits harvested
later (October), with lower % acid, had lower values of
EHU8, because temperature sums were calculated
using the lower temperatures which prevail during
winter months. To assess the in¯uence of temperatures during the last period before harvest, we plotted
% acid against EHUs for the last 3 months (EHU3).
The results are shown in Fig 6, which indicates that,
after a threshold EHU3 value of about 100 units over
that period (equivalent to average daily temperatures
of at least 14 °C), % acid falls steadily with increasing
EHUs. A line with a slope of about ÿ0.0005 can be
Because the relationships between EHUs and per
cent acid are better than those between EHUs and
Brix, the best option for predicting fruit quality at the
juicing plant (in terms of temperature during the
growing season) is to use eqns (3) and (4) describing
the relationships between EHUs and % acid, and then
estimate Brix from the correlation with acidity. This
will allow estimates of possible regional or seasonal
differences in fruit quality (caused by temperature
conditions) as well as estimates of the effects of leaving
fruits on the trees for varying lengths of time. Some
results from this procedure are given in Table 3.
Figure 5. % acid in Navel oranges at processing plotted against effective
heat units (EHUs) for the 8 month period before harvest (EHU8). The data
indicate that later-harvested fruits (September/October), have low acid
levels, despite low EHU values. The high acid values are associated with
winter-harvested fruits.
Figure 6. % acid in Navel oranges at processing plotted against effective
heat units (EHUs) calculated for the 3 months before harvest (EHU3). The
data indicate a threshold value of about 100 EHUs, below which
temperature in this period appeared to have little influence on fruit quality.
The line has a slope of about ÿ0.0005 (cf Fig 4).
Brix(V) ˆ 8:8 ‡ 2:06…% acid† …r 2 ˆ 0:4†
and for Navel oranges the best ®t line was
Brix(N) ˆ 9:17 ‡ 3:47…% acid†
…r 2 ˆ 0:3†
J Sci Food Agric 80:275±283 (2000)
Effect of temperature sums on juicing quality of oranges
Table 3. Sample calculations of fruit quality based on the use of eqns (3) and (4) to calculate per cent acid for Valencia (V) and
Navel (N) oranges respectively and eqns (5) and (6) to calculate Brix from the predicted per cent acid values. Reference values of
EHUs for the MIA are provided in Table 4
% acid(V)
% acid(N)
Note that the average annual total EHUs is about
1600. There are indications (Figs 3 and 4) that about
1700 EHUs is the critical (threshold) value, after
which % acid starts to fall. This indicates that the
spring±early-summer temperatures (October±December) are important in determining fruit quality; if
temperatures are high, % acid will fall rapidly, but if
they are low, fruit quality will be slow to reach its
optimum. This is consistent with the analysis of EHU3
for Navels (see Figs 5 and 6 and associated discussion).
We must emphasise that the value of the empirical
equations presented here lies in predictions of fruit
quality over a number of seasons in the MIA region of
New South Wales. They will not provide accurate
results for particular production sub-locations (sites)
in particular years. To develop equations for this
purpose, it will be necessary to study the fruit quality
coming from the sites of interest. If this is to be done,
more detailed and accurate data about time of fruit set,
local temperatures during fruit growth, and exact
picking dates would be required. Fruit quality should
also be determined immediately after picking.
Given the wide range of Brix and acid contents
recorded at any one point in time for fruit deliveries
to the juice processor from across the MIA growing
region, the fact that workable relationships between %
acid and EHUs, between % acid and Brix and between
ratio Brix/acid and EHUs were found is evidence to
support the original hypothesis that variation in fruit
quality late in the harvest season could be accounted
for by temperature conditions during the period that
fruits were on the trees.
We present a relatively crude but practically useful
model for predicting changes in fruit quality after
reaching harvest maturity. The analysis also provided
some insights into the relationship between Brix,
which is a measure of sweetness, and % acid, the
way they vary relative to one another over time, and
the possibility of predicting optimum fruit quality.
The trends and results illustrated are statistically
signi®cant and can be accepted as re¯ecting real
effects. They show that gross changes in fruit internal
quality can be predicted in terms of temperature, using
effective heat units (EHUs) to calculate changes in
fruit Brix and acid contents. The equations will be
usable by processors to provide indications of the
probable quality of deliveries at any time in particular
seasons; the results will not be exact, but should be
useful in planning harvest schedules.
The relationships between temperature sums and %
acid were stronger and there was less variation about
the predicted values than there was for Brix. The
relationship between temperature and Brix (Fig 2)
varied considerably from one season to the next, so the
general relationship between Brix and EHUs would
not be a reliable predictor of fruit sweetness. There is
nothing in the climatic or laboratory data that gives
any indication of the reasons for this variation, but
alternate light and heavy cropping cycles21 would
largely explain the variation in recorded Brix levels for
similar EHU values across seasons in the same
locations. The results indicate that the processes
causing changes in total soluble sugars in oranges are
not strongly affected by temperatureÐan indication
that should be investigated in much more detailed
studies. To investigate the possible effects of alternating crop load, and its consequent effect on fruit size
and the rate of Brix and acid formation, would also
require investigation in separate studies.
Because of the uncertainty associated with predic-
Table 4. Effective heat units at Griffith in the MIA. The data are means for each month for 10 years from 1986 to 1995. Highest and
lowest values may have come from any year
J Sci Food Agric 80:275±283 (2000)
RJ Hutton, JJ Landsberg
tion of Brix using EHUs, estimates of fruit quality at
harvest time should be made using the equations
describing changes in % acid with EHUs in conjunction with the equations describing the relationship
between Brix and % acid and their ratio. These
equations have potential use in predicting the internal
quality of citrus fruit during extended harvest periods
and could also be used in comparing varietal performance at harvest.
Earlier studies13 demonstrated a linear increase in
Brix/acid ratio with time (months) until harvest
maturity, and the authors proposed the use of this
relationship to derive a basal date by reverse extrapolation to a time when this ratio was theoretically
zero. From this time, Brix was assumed to increase in
the developing juice segments and juice acidity was
assumed to decline owing to translocation of water to
the fruit. A daily increase in ratio was derived and the
time (absolute value in days) when fruit would reach a
particular quality level (Brix/acid ratio) could be
predicted. However, these relationships, describable
by kinetic equations which alter the Brix and acid
contents of maturing fruit, do not hold after the fruit
reaches its optimum maturity, as has been demonstrated in the current study.
From our work it is of interest to note the highly
non-linear nature of the Brix/acid ratio in relation to
EHUs after fruits reach a minimum acceptable
maturity standard for marketing. If we take eqns (3)±
(6) and calculate the Brix/acid ratio for the range of
EHU values used in Table 3, we obtain the results
shown in the curves (Fig 7) for Valencia, ratio V(T),
and Navel, ratio N(T), oranges.
The more negative slope for the decline in juice
acidity in Navels (ÿ0.0006) as opposed to that for
Valencias (ÿ0.000275) causes the upper Brix/acid
ratio curve for Navels (Fig 7) to rise more rapidly. As
juice acidity falls below 1% in both Navels and
Valencias, Brix/acid ratio rises rapidly, especially when
EHU3 (acid) > 120 and EHUTsum (acid) > 1700 are
exceeded. The curves shown in Fig 7 provide a very
good estimate of the observed data for the period of the
Figure 7. Changes in the Brix/acid ratio with increasing temperature sums
(T) as the harvest season is extended. The ratio curves (V, N) are based on
the prediction of % acid using EHU eqns (3) and (4) and the equations
relating Brix to % acid, eqns (5) and (6), for Valencia and Navel orange
varieties respectively.
This strategy for determining total soluble solids
(Brix), acid content and Brix/acid ratios could be
transposed for prediction purposes to identify the
duration of the harvest period beyond which the juice
quality characteristics would be unsuitable for use by
processors. From this, harvesting and processing
schedules could be arranged to give growers the
optimum factory price for their crop and processors
the best quality for their products. The study indicates
that it will be a worthwhile procedure for growers to
keep good temperature records and records of the
values of % acid and Brix determined by the
processors for their fruit. This will enable them to
produce relationships speci®c to their trees and
growing conditions to maximise the price received
for their fruit based on juice quality at the time of fruit
1 Soule J, Grierson W and Blair JG, Quality tests for citrus fruits.
Florida Agriculture Extension Circular 315; (1967).
2 Grierson W and Ting SV, Quality standards for citrus fruits,
juices and beverages. Proc Int Soc Citric 21±27 (1978).
3 Monselise SP, Citrus. In CRC Handbook of Fruit Set and
Development, Ed by Monselise SP, CRC Press, Boca Raton.
FL, pp 87±108 (1986).
4 Chen CS, Models for seasonal changesin °Brix and ratio of citrus
fruit juice. Proc Fla State Hort Soc 103:251±255, (1990).
5 Spiegel-Roy P and Goldschmidt EE, Biology of citrus. In Biology
of Horticultural Crops, Cambridge University Press, Cambridge,
pp 230 (1996).
6 Erickson LC, The general physiology of citrus. In The Citrus
Industry, 2nd rev edn, Ed by Reuther W, Batchelor LD and
Webber HJ, University of California, Berkeley, CA, pp 86±126
7 Sinclair WB and Bartholomew ET, Compositional factors
in¯uencing the edible quality of oranges. Proc Am Soc Hort
Sci 50:177±186 (1947).
8 Sinclair WB, The Biochemistry and Physiology of the Lemon and
Other Citrus Fruits, University of California Division of
Agriculture and Natural Resources, Oakland, CA (1984).
9 Marloth RH, Fruit-growth and internal quality changes. In Citrus
Growth Studies II, South African Association for the Advancement of Science, Kimberley, pp 235±248 (1949).
10 Bain JM, Morphological, anatomical, and physiological changes
in the developing fruit of the Valencia orange, Citrus Sinensis
(L.) Osbeck. Aust J Bot 6:1±28 (1958).
11 Bevington KB, Hutton RJ, Papasidero S and Gee S, Development of phenological based crop management model for high
density citrus. HRDC Final Project Report CT216 (1996).
12 Chandler BV and Kefford JF, The chemical assay of limonin,
the bitter principle of oranges. J Sci Food Agric 17:193±197
13 Chandler BV and Nicol KJ, Quality changes in maturing
oranges. Proc Int Soc Citric 27±30 (1978).
14 Kimball DA, Brix/acid ratio correlation with climatic data for
California Navel and Valencia oranges. Citrus Ind Mag
(August): 26±28 (1984).
15 Al Soboh G and Moore G, Citrus harvest notesÐways to
improve orange harvesting and handling. HRDC Final Project
Report CT520 (1996).
16 Koch KE and Avigne WT, Postphloem, nonvascular transfer in
citrusÐkinetics, metabolism, and sugar gradients. Plant Physiol
93:405±1416 (1990).
17 Reuther W, Climate and citrus behaviour. In The Citrus
Industry, Vol 3, University of California, Berkeley, CA, pp
280±337 (1973).
J Sci Food Agric 80:275±283 (2000)
Effect of temperature sums on juicing quality of oranges
18 Bevington KB and Castle WS, Annual root growth pattern of
young citrus trees in relation to shoot growth, soil temperature,
and soil water content. J Am Soc Hort Sci 110:840±845 (1986).
19 Yen C and Koch KE, Developmental changes in translocation
and localization of 14C-labeled assimilates in grapefruit: light
and dark CO2 ®xation by leaves and fruit. J Am Soc Hort Sci
115:815±819 (1990).
J Sci Food Agric 80:275±283 (2000)
20 Tucker DPH and Reuther W, Seasonal trends in composition of
processed `Valencia' and Navel oranges from major climatic
zones of California and Arizona. Proc Am Soc Hort Sci 90:529±
540 (1967).
21 Hutton RJ, Improving fruit size and packout of Late Valencia
oranges with ethephon fruit-thinning sprays. Aust J Exp Agric
32:753±758 (1992).
Без категории
Размер файла
137 Кб
Пожаловаться на содержимое документа