Professional Documents
Culture Documents
Consorcio Latinoamericano para Investigacin y Desarrollo de la Yuca (CLAYUCA), Apartado Areo 6713, Cali, Colombia
International Center for Tropical Agriculture (CIAT), Apartado Areo 6713, Cali, Colombia
a r t i c l e
i n f o
Article history:
Received 8 October 2012
Accepted 14 June 2013
Keywords:
Gene expression
Genetic tolerance
Post-harvest losses
Shelf life
a b s t r a c t
Cassava (Manihot esculenta Crantz) roots spoil 23 days after harvest due to post-harvest physiological
deterioration (PPD), which had remained an unsolved problem until recent reports of genetic variation
for tolerance to it. PPD is a genetically active, oxidative process triggered when the harvested roots are
separated from their mother plant. The short shelf life of harvested roots results in large losses and high
transport and marketing costs. Recent reports on positive genetic variation for tolerance to PPD will
facilitate breeding for extended shelf life of the roots and a better understanding of the biochemical and
genetic events leading to PPD. However, PPD scoring is difcult and prone to large experimental errors.
It is often the case that roots from the same plant can have 0 and 100% PPD score due to injuries during
the harvest process, variation in dry matter content and, most likely, other variables yet to be identied.
Therefore, sampling a root for biochemical or genetic studies and measuring PPD in a different root, is not
a reliable approach. A device has been developed and tested for the possibility of extracting a core of root
parenchyma, lling the space with melted parafn (to reduce oxygen availability), and then one or two
weeks later visually quantifying PPD in the same root. Sampling the roots did not have any signicant
effect on PPD suggesting that the protocol can be used for biochemical composition and gene expression
studies related to the causes of PPD and the possibilities of developing tolerance to it.
2013 Elsevier B.V. All rights reserved.
1. Introduction
Cassava is among the most important staple crops in tropical
and subtropical regions of the world. It shows a remarkable and
well recognized adaptation to marginal growing conditions due
to its perennial growth habit. When there are biotic and/or abiotic stresses the plant can assume a dormant status until favorable
growing conditions return. This characteristic provides exibility
and plasticity for the plant to adapt to changes in the environment
(Ceballos et al., 2011). Compared with other staple foods, cassava
also offers the advantage of a exible harvesting date, allowing
farmers to keep the roots in the ground until needed (Iglesias
et al., 1997). Although the starchy root is the primary product, fresh
leaves are also used for animal and/or human consumption in Africa
and Asia (Benesi et al., 2010; Howeler, 2011).
In addition to the important role cassava plays in food security,
there is a growing demand for cassava roots by the starch, food,
animal feed and ethanol industries (Balagopalan, 2002; Buitrago,
2011a,b,c; Chauynarong et al., 2009; Moorthy, 2004; Sriroth et al.,
Corresponding author. Tel.: +57 2 445 0125; fax: +57 2 445 0083.
E-mail address: h.ceballos@cgiar.org (H. Ceballos).
0925-5214/$ see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.postharvbio.2013.06.026
530
roots severely limits the marketing options by increasing the likelihood of losses and the overall marketing costs (Salcedo et al.,
2010). Extending shelf life of cassava roots to 45 days would result
in annual benets above 35 million US$ in Thailand alone (Vlaar
et al., 2007). Economic impact would be considerably higher in
other countries where road and transport infrastructure is not as
developed as in Thailand.
The processes involved in PPD is initiated as soon as the
root is detached from the mother plant and resemble typical changes associated with the plants response to wounding.
It triggers a cascade of biochemical reactions, in which reactive oxygen species (ROS) are central. Specic genes involved
in PPD have been identied and characterized, and their
expressions evaluated (Reilly et al., 2001). Several secondary
metabolites, particularly hydroxycoumarins, accumulate in the
process (Bayoumi et al., 2008a,b, 2010). The PPD begins 2448 h
after harvest (at 2030 C and 6580% relative humidity), but
handling and storage conditions of the roots greatly affect its
speed and magnitude. Keeping roots at 10 C and 80% relative humidity delays the onset of PPD by two weeks. The
PPD also has been linked with high dry matter content in
the roots (van Oirschot et al., 2000; Snchez et al., 2006),
an unfortunate association because high dry matter content
of the roots is a common objective of many cassava-breeding
projects.
There are early reports on genotypic variation to PPD (Booth,
1976; Ekanayake and Lyasse, 2003). However in no case the degree
of tolerance reached those recently reported by Morante et al.
(2010). Research on PPD and the physiological, chemical, genetic
and/or environmental factors affecting it is difcult because of the
large experimental errors typically associated with the available
scoring protocols. Root handling during harvest and transport, and
root dry matter content are among the known factors inuencing PPD. It is often the case that roots from the same plant may
have scores from 0 to 100% PPD. A major problem, therefore, is that
the current approach of sampling one root for biochemical or gene
expression studies (usually soon after harvest) and measuring PPD
in a different root from the same genotype (typically 710 d after
harvest), is largely inefcient.
The availability of tolerance and advances in our understanding of the process and factors affecting it (Bayoumi et al., 2008a,b,
2010; Ndidi and Akeem, 2011; Reilly et al., 2007) have lead to
the need for a protocol that: (a) allows a non-destructive sampling of root tissue; (b) that can be performed days before PPD is
quantied; and (c) that will not induce (or prevent) the normal
PPD onset and development. This article describes an innovative,
non-destructive approach for sampling cassava roots soon after
harvest (for biochemical composition proles and/or gene expression) without noticeable changes in PPD development through
periods ranging from one to two weeks of storage. This method will
allow research for a better understanding of the factors leading to
PPD and the biochemical/genetic factors determining tolerance to
this problem.
Hour
Average
Minimum
Maximum
95.1
57.7
75.1
92
53
72
97
62
79
7:00 AM
1:00 PM
7:00 PM
96.7
62.6
75.1
95
58
70
98
71
81
Daily temperature ( C)
First
Second
24.7
24.4
19.2
19.4
30.8
30.0
quantication of PPD. Only the visual signs of deterioration (graybluish vascular streaking) were assessed in this study.
2.3. Dry matter content
A sample from these roots was taken for the quantication of
dry matter content after measuring PPD. To estimate it, 2030 g of
chopped and grated fresh roots were dried in an oven at 60 C for
24 h. Dry matter was expressed as the percentage of dry weight
relative to fresh weight.
2.4. Statistical analysis
The PPD values were expressed as percentages following the
scoring procedure originally developed by Wheatley et al. (1985).
The data were arcsine-square root transformed prior to analysis
(Steel and Torrie, 1960). Analysis of variance was conducted using
the PROC GLM from SAS (SAS, 2008). The experimental unit was
each individual root from different cassava cultivars and subjected
to one of four possible treatments (a1b1; a1b2; a2b1; a2b2).
2.5. Extraction of a cylindrical sample of root parenchyma
PPD is an oxidative process actively controlled by the expression of genes in the root which is triggered by the separation of
the root from the mother plant and is clearly linked to the presence of reactive oxygen species (ROS) (Reilly et al., 2003). Different
studies require sampling of the roots soon after harvest (to monitor chemical composition or gene expression) for a phenomenon
that will develop and become evident several days later. However,
this sampling requires that ROS is prevented from promoting or
accelerating PPD around the area injured during the extraction of
the sample. A device was therefore developed for extracting a core
of root tissue and then immediately pouring melted parafn in to
ll the space left by the extraction of the root sample (Fig. 1). Key
features of this device are: (a) A stainless steel heated container on
top containing the melted parafn; (b) a heated valve to manually
release the melted parafn immediately after the sample of the root
was taken; (c) a stainless steel cylinder with an internal diameter
of 8 mm, whose penetrating edge was slanted and sharpened; (d)
a lever connected to the sampling cylinder; and (e) a pusher rod
inside the cylinder that was used to push the core root sample out
of the cylinder.
3. Results
Table 1 presents a summary of key environmental parameters
that have been linked to the development of PPD. Temperature
ranged from 19 to 30 C both in the rst and second week of the
experiment. Data for the second week is only relevant for the roots
stored without their extremes cut. Air moisture was always above
531
90% at 7:00 AM, and ranged between 70 and 80% at 7:00 PM. At
1:00 PM moisture was 5362% in the rst week and 5871% in the
second.
At the beginning of the study 15 roots were assigned for each
combination of treatments. One of the common problems often
observed while assessing PPD in cassava roots is the rotting of roots
from microbial infections. The symptoms of rotted roots are very
distinctive and different from those of PPD. It is very easy to distinguish one from the other. However, PPD cannot be assessed in
rotten roots. Averages across the 15 roots representing each combination of treatments along with those for the three main effects
are presented in Table 2. This table also presents information on
the averages for dry matter content (DMC) and number of rotten
roots, both variables quantied at the end of the experiment.
A key nding presented in Table 2 is the fact that in no case
sampling the roots resulted in changes on the PPD score (either
increasing or decreasing it). Average PPD scores from sampled or
un-sampled roots were similar when analyzing roots from individual clones and after 7 or 14 d of storage. Therefore, PPD averages
across clones and duration of storage period were not statistically
different. This is important because data would suggest that, in no
case, sampling the roots resulted in a noticeable change in PPD evolution. Although the analysis of variance for PPD was made on the
arcsine-square root transformed data, to facilitate understanding
of results Table 2 presents the original values for PPD. The signicances of those means, however, relate to the transformed data.
As reported in the literature PPD seems to be correlated with
DMC (van Oirschot et al., 2000; Snchez et al., 2006). It is important,
therefore, to report DMC values as a reference point when analyzing PPD in cassava roots. There was a signicant difference for the
average DMC of roots from AM 206-5 sampled or not sampled, after
both 7 and 14 d of storage. Also sampled roots from HMC1 stored
for 7 d had higher DMC than un-sampled roots (41.6 vs. 34.4%). It
is unlikely that these differences are just random sampling variation. Therefore as expected, the overall average of sampled roots
also showed signicantly higher levels of DMC compared with roots
that had not been sampled (41.2 vs. 38.6%). It is not clear why these
differences occurred and only in roots from AM 206-5 and HMC-1. A
feasible explanation would be that in some cases sampling the root
allowed some water loss through the injured tissue which would
explain the observed increases in DMC. It is important to emphasize that the changes observed in DMC for some clones did not have
an impact on PPD scores.
A summary of the analyses of variance is presented in Table 3.
Four variables were analyzed: weight loss expressed in grams or as
percentage of the initial weight, PPD (expressed as percentage and
transformed) and DMC. Highly signicant effects (P > 0.01) were
found for the clone source of variation for the four variables. The
length of storage period also had highly signicant effects on both
ways of measuring weight loss, but not for PPD or DMC. Sampling a
cylinder of root parenchyma had a signicant effect (P > 0.05) only
for weight loss expressed as percentage of the initial weight and for
DMC. No interaction achieved statistical signicance for any variable except for changes in DMC. It is important to highlight that
extracting a cylinder of root parenchyma did not have any noticeable effect on PPD development. This variable, as expected, was
highly affected by the genetic differences.
As indicated by the analysis of variance sampling the roots and
lling the space with parafn did not result in signicant changes
in PPD (11.9 versus 12.5%, Table 2). When individual treatments
(clones and duration of the storage period) are considered, then
larger variation between averages of sampled and un-sampled
roots can be observed. For example, in the case of AM 206-5, roots
stored for 14 days showed 9.1% PPD when they had been sampled but average PPD of un-sampled roots was only 3.9%. However,
these differences were not statistically signicant. The standard
532
Fig. 1. (A) Illustration of the system to extract core samples of root parenchyma and ll the space with melted parafn; (B) example of core of the root parenchyma extracted;
(C) appearance of the root after parafn had been applied and solidied; (D) cooled down and solidied cylinder of parafn extracted from the root for illustration.
Table 2
Least square means and standard deviations (within parenthesis) of the experiment to assess PPD in roots from four cassava genotypes. At the bottom of the table the
averages across genotypes, duration of storage period and sampling versus not sampling the roots is provided for comparison.1 Roots stored for 7 d were subjected to
Wheatles method, whereas roots stored for 14 d were not cut at the extremes.
Clone
AM206-5 (Tolerant)
CM523-7 (Suscept.)
HMC1 (Suscept.)
MPER183 (Tolerant)
AM206-5
CM523-7
HMC1
MPER183
Roots stored for 7 days
Roots stored for 14 days
Roots sampled
Roots not sampled
1
2
Sample taken
Storage (d)
7
7
14
14
7
7
14
14
7
7
14
14
7
7
14
14
Rotten roots
0
1
3
1
0
1
4
1
1
0
0
0
3
1
5
3
5
6
1
12
7
17
16
8
Weight loss2
PPD2
DMC2
(g)
(%)
(%)
(%)
24.3(5.4)
21.8(5.3)
41.9(15.2)
28.6(9.8)*
35.0(12.4)
25.3(11.2)*
35.3(19.4)
36.4(20.5)
33.0(8.8)
36.7(14.4)
58.3(25.1)
50.8(28.1)
41.0(17.1)
45.3(14.7)
54.5(25.9)
47.7(32.0)
29.2b
33.0b
44.7a
47.1a
32.8b
44.2a
40.4a
36.6a
7.3(1.6)
6.2(1.1)*
8.9(1.9)
7.3(1.7)
9.0(1.6)
8.5(1.4)
9.7(2.1)
10.1(2.4)
8.6(1.3)
8.4(1.4)
10.0(2.6)
9.5(2.3)
10.1(2.0)
10.8(1.7)
12.5(2.5)
9.7(2.4)*
7.4c
9.3b
9.1b
10.8a
8.6b
9.7a
9.5a
8.8b
3.0(4.6)
0.6(1.9)
9.1(9.3)
3.9(8.4)
24.8(11.2)
20.3(10.9)
18.2(18.3)
20.6(16.1)
13.6(10.9)
17.4(12.5)
17.2(11.3)
27.1(19.0)
4.3(4.6)
5.8(10.6)
4.7(6.8)
3.8(7.7)
4.2b
21.0a
18.8a
4.6a
11.2a
13.1a
11.9a
12.5a
43.2(1.9)
33.4(3.9)**
35.0(2.3)
31.5(4.3)*
43.9(1.7)
44.3(1.6)
43.0(1.7)
43.5(2.0)
41.6(3.0)
34.4(5.1)**
41.7(2.9)
41.7(1.9)
39.5(2.0)
39.0(4.5)
41.8(2.8)
41.4(1.9)
35.8c
43.7a
39.8b
40.5b
39.9a
40.0a
41.2a
38.6b
Treatments followed by the same superscript letters are not statistically different.
Difference between means of sampled versus not sampled roots (for individual clones and specic duration of storage period) signicant at P < 0.05 (*) or P < 0.01 (**).
533
Table 3
Mean squares from the analyses of variance for weight losses (expressed in g or in % related to initial weight) and post-harvest physiological deterioration (PPD) and dry
matter content (DMC) expressed in %.
Source of variation
df
Weight loss
(g)
Clone (C)
Root/clone
Length of storage (S)
Root sample taken (R)
CS
CR
CSR
Error
*
**
3
56
1
1
3
3
4
144
(%)
**
**
3886
214
6668**
739
522
119
386
90.14
2.90
64.44**
24.84*
1.04
4.73
9.31
365
3.94
Arcsin ( PPD)
DMC
(%)
(%)
**
34.83
0.73
0.62
0.41
2.60
2.03
0.52
0.04
516.5**
10.1
0.9
295.6**
198.6**
141.0**
312.3**
8.3
Fig. 2. Photographs of roots from (A) tolerant genotype AM 206-5 and (B) susceptible genotype CM523-7. For each genotype roots on the left had not been sampled and
those of the right had been sampled and the parafn cylinder is clearly visible. In some cases, a chalky healing tissue could be observed around the parafn cylinder (arrows
on right).
4. Discussion
Sampling the roots increased the frequency of rotten roots (16
versus 8 in Table 2). This makes sense, as the injury can serve
as an entry point for different microbes and fungi. MPER183, as
observed in the past, had a good tolerance to PPD but its roots tend
to rot considerably more than those from other genotypes (particularly HMC1). MPER183 also lost a lot of weight. Both MPER183 and
AM206-5 where clearly tolerant to PPD (4.6 and 4.2%, respectively),
however AM 206-5 lost considerably less weight than the former
(both in grams or in %). It can be concluded, therefore, that weight
loss is probably unrelated to tolerance to PPD.
Also in agreement with the analysis of variance information,
there was a clear effect of the duration of the storage period on
weight losses (both expressed in grams or as percentage). Weight
loss is higher in roots stored for 14 d (44.2 g or 9.7%) than for 7
d (32.8 g or 8.6%). However, there was no statistical difference in
the PPD levels when roots were stored for 7 or 14 d (11.2 versus
13.1%). This last nding reinforces the long standing perception that
cutting the extremes of the roots and covering the distal one with
cling lm contributes to accelerating PPD. Since no genotype by
treatment interaction was signicant, it can be concluded that both
methods of storage provide similar information on PPD.
PPD is an enzymatically mediated oxidative process which
parallels plant wound, senescence and defense responses. It is a
very active and complex process in which as many as 72 nonredundant expressed sequence tags were either induced or down
534
535
quality characteristics. Journal of the Science of Food and Agriculture 80, 1866
1873.
Vitrac, O., Dufour, D., Trystram, G., Raoult-Wack, A.L., 2002. Characterization of heat
and mass transfer during deep-fat frying and its effect on cassava chip quality.
Journal of Food Engineering 53, 161176.
Vlaar, P.W.L., van Beek, P., Visser, R.G.F., 2007. Genetic modication and its impact
on industry structure and performance: post-harvest deterioration of cassava
in Thailand. Journal on Chain and Network Science 7, 133 142.
Wheatley, C., Lozano, C., Gomez, G., 1985. Post-harvest deterioration of cassava roots.
In: Cock, J.H., Reyes, J.A. (Eds.), Cassava: Research, Production and Utilization.
UNDP-CIAT, Cali, Colombia, pp. 655671.