See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/225587140

Quantitative trait loci analysis of morphological traits in Citrus

Article  in  Plant Biotechnology Reports · January 2011

DOI: 10.1007/s11816-011-0194-z

CITATIONS READS

12 125

2 authors, including:

Mehtap Sahin-Cevik
Applied Sciences University of Isparta
16 PUBLICATIONS   106 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Pomegranate View project

All content following this page was uploaded by Mehtap Sahin-Cevik on 19 May 2014.

The user has requested enhancement of the downloaded file.

Plant Biotechnol Rep (2012) 6:47–57
DOI 10.1007/s11816-011-0194-z

ORIGINAL ARTICLE

Quantitative trait loci analysis of morphological traits in Citrus

Mehtap Şahin-Çevik • Gloria A. Moore

Received: 3 December 2010 / Accepted: 3 August 2011 / Published online: 30 August 2011
Ó Korean Society for Plant Biotechnology and Springer 2011

Abstract The objectives of this study were to understand markers and putative QTLs associated with these traits
the genetic basis of morphological variation observed in the using interval mapping method. QTL analysis revealed 18
genus Citrus and its relatives and to identify genomic putative QTLs of LOD [ 3.0 associated with 13 of the
regions associated with certain morphological traits using morphological traits analyzed. The putative QTLs were
genetic linkage mapping and quantitative trait loci (QTLs) distributed in several different linkage groups, and QTLs
analysis with random amplified polymorphic DNA (RAPD) associated with similar traits were mostly mapped to the
markers. First, a genetic linkage map was constructed with same LG or similar locations in the linkage group, indi-
RAPD markers obtained by screening 98 progeny plants cating that the same genomic region is involved in the
from a {Citrus grandis 9 [C. paradisi 9 Poncirus trifoli- inheritance of some of the morphological traits.
ata]} 9 {[(C. paradisi 9 P. trifoliata) 9 C. reticula-
ta] 9 [(C. paradisi 9 Poncirus trifoliata) 9 C. sinensis]} Keywords Citrus
Morphological traits

intergeneric cross. The map contains 69 RAPD markers QTL mapping
Molecular markers
RAPD
distributed into nine linkage groups. Then, 17 different
morphological traits, including six tree and two leaf char-
acters of 98 progeny plants and six floral and three fruit Introduction
characters of about half of the same progeny plants were
evaluated for 2 years and statistically analyzed for varia- A significant amount of phenotypic variation can be
tion. Statistical analysis of individual traits indicated that observed within the genus Citrus for nearly every charac-
trunk diameter and growth, tree height, canopy width, tree ter, including growth habit, vigor, and vegetative mor-
vigor and growth, leaf length and width, petal and anther phology (Gmitter et al. 1992). The phenotypic variability is
numbers, petal length and width, length of pistil and style, even greater between Citrus and closely related interfertile
fruit length and diameter, and fruit segment number showed genera such as Poncirus (Soost and Cameron 1975).
normal or close to normal distribution, suggesting that these However, the genetic basis and the inheritance of most
traits may be inherited quantitatively. Quantitative data variation in Citrus are not understood due to a number of
from the morphological traits were analyzed to detect practical difficulties, such as large plant size, long juve-
nility periods, polyembryony, heterozygosity, sterility,
self- and cross-incompatibility, inbreeding depression and
M. Şahin-Çevik (&) particularly, the quantitative inheritance of most charac-
Department of Agricultural Biotechnology,
teristics. In Citrus, the genetics of morphological traits has
Faculty of Agriculture, Suleyman Demirel University,
32260 Isparta, Turkey not been explored in great detail, and no comprehensive
e-mail: [email protected] research has been conducted on genetic variations in the
morphological traits of this genus; in addition, few mor-
G. A. Moore
phological traits of Citrus and its relative Poncirus have
Department of Horticultural Sciences, Plant Molecular and
Cellular Biology Program, University of Florida, Gainesville, been studied individually. Cameron and Soost (1980)
FL 32611, USA studied leaf types and suggested that the trifoliate leaf type

123
48 Plant Biotechnol Rep (2012) 6:47–57

of Poncirus is dominant over the unifoliate leaf of Citrus. RAPD markers have been used for linkage mapping,
However, segregation patterns in some crosses included marker-assisted selection, genotype identification, taxon-
bifoliate leaves, suggesting that leaf type does not result omy, and other types of genetic studies (Mohan et al. 1997)
solely from a single gene. The results of a study on poly- on cereals (González et al. 2002; Schoenenberger et al.
embryony indicated that most Citrus species are polyem- 2005; Orr and Molnar 2008), vegetables (Cheng et al.
bryonic and the trait is possibly controlled by one or two 2009; Finger et al. 2010), and fruit crops (Raina et al. 2001;
dominant genes (Soost and Roose 1996). Possibly because James et al. 2004; Venkatachalam et al. 2008; Cerqueira-
of the economical implications, tree size has been the most Silva et al. 2010).
extensively studied among the morphological traits in The loci involved in the inheritance of quantitative traits
Citrus. In the late 1960s, Cameron and Frost (1968) theo- are termed QTLs. The approximate location of QTLs in the
rized that a single major gene might be involved in genome can be identified by QTL mapping (Gelderman
dwarfness in Citrus. Later, when the inheritance of 1975), in which the association between a quantitative trait
dwarfing was studied in a population of open-pollinated and markers segregating in a population is detected. Since
seedlings of Poncirus trifoliata ‘Flying Dragon’, which is the first published reports of QTL mapping with DNA
used as dwarfing rootstock for Citrus, dwarfing was found markers in plants involving the analysis of traits such as
to segregate in a 3:1 ratio, suggesting that it is inherited by pH, size, and soluble solids in tomato fruit (Paterson et al.
a single dominant gene for which ‘Flying Dragon’ is het- 1988), QTL mapping has widely been used in many plants
erozygous. Two morphological traits, curved thorns and to study a number of economically important traits, such as
twisted trunk growth, and three random amplified poly- environmental stresses (Foolad and Chen 1998; Tondelli
morphic DNA (RAPD) markers closely linked to the et al. 2006; Khowaja et al. 2009; Genc et al. 2010), disease
dwarfing trait have also been identified in P. trifoliata resistance (Caranta et al. 1997; Zwonitzer et al. 2009;
(Cheng and Roose 1995). Zhang et al. 2011), fruit quality (Kennard and Havey 1995;
Although variation in fruit characteristics of Citrus Moreno et al. 2008; Pérez-Vega et al. 2010), and physio-
cultivars is obvious and Citrus scions and rootstocks have logical (Mian et al. 1998; Villalta et al. 2008; Abdel-
been bred for fruit quality, the genetics of fruit character- Haleem et al. 2011) and morphological characteristics
istics in Citrus has not been studied in great detail. Studies (Paran et al. 1997; Gyenis et al. 2007; Xue et al. 2008).
on genetic parameters for fruit morphology in orange, QTL mapping has also been used in Citrus to understand
grapefruit, and mandarin suggest that most fruit characters the inheritance of complex economically important traits,
demonstrate intermediate to high broad sense heritability. including salt tolerance (Tozlu et al. 1999a, b), cold tol-
The presence of large variability for most fruit character- erance (Weber et al. 2003), resistance to Phytophthora
istics of different mandarin cultivars was demonstrated and gummosis (Siviero et al. 2006), Citrus tristeza virus (Asins
documented by Russo and Fanizza (1992). et al. 2004), nematode (Ling et al. 2000), and leaf minor
The application of modern genetics and molecular (Bernet et al. 2005), and yield and seed number (Garcia
biology tools, such as molecular markers, linkage mapping, et al. 2000).
and quantitative trait locus (QTL) analysis has facilitated The results of these and QTL studies in other plants
the genetic studies of quantitative traits in many crops and suggest that QTL analysis with molecular markers would
may provide a better understanding of the genetics of provide insight into the genetics of important morpholog-
morphological traits in Citrus. RAPD markers have been ical traits, including tree, leaf, flower and fruit character-
extensively used in genetic studies of many organisms, istics, of Citrus. Study of the genetics of morphological
ranging from humans (Benter et al. 1995; Lee et al. 2007; traits in Citrus is important not only for understanding the
Vega et al. 2010) to microorganisms (Fani et al. 1993; genetic basis of inheritance of these traits but also for the
Douhan et al. 2008; Souza et al. 2010) as well as in Citrus manipulation of morphological traits for desirable horti-
(Cai et al. 1994; Fang et al. 1997; Hao et al. 2004) because cultural characteristics.
of the several advantages of RAPDs over other markers In the study reported here, we studied the variation in
(Welsh and McClelland 1990; Williams et al. 1990). morphological traits in Citrus by evaluating and analyzing
RAPD markers are not only more useful for generating measured values of 17 different morphological traits,
polymorphisms between closely related genotypes, but including tree, flower, and fruit characters. In addition, a
they also enable the screening of large populations of genetic linkage map was constructed using RAPD mark-
breeding materials. In addition, the methodology does not ers. The linkage map and the morphological data were
require any sequence information or large amount of DNA, used to identify QTLs associated with the measured
and the production and subsequent analysis of RAPD morphological traits and to determine genomic regions
markers are simple, quick, and inexpensive and require less involved in inheritance of these traits using a QTL
labor (Deng et al. 1995; Gostimsky et al. 2005). In plants, mapping approach.

123
Plant Biotechnol Rep (2012) 6:47–57 49

Materials and methods acetic acid)] at a constant power of 60 V for about 2 h.

Gels were stained with ethidium bromide and visualized
Plant materials and photographed under UV light using the IS-1000 digital
gel imaging system (Alpha Innotech, San Leandro, CA).
A progeny population of 98 mature plants from the complex
intergeneric cross {Citrus grandis Obs. 9 [C. paradisi Macf. Scoring of segregating markers and linkage analysis
cv. Duncan 9 Poncirus trifoliata Raf.]} 9 {[(C. paradisi
cv. Duncan 9 P. trifoliata) 9 C. reticulata Blanco] 9 All primers were initially screened using the genomic DNA
[(C. paradisi cv. Duncan 9 Poncirus trifoliata) 9 C. sinen- of all parents and five individuals from the progeny pop-
sis Osb. cv. Succary]} (Fig. 1) was randomly selected from a ulation to detect polymorphisms. The primers generating
larger population for establishing the linkage map and for polymorphic DNA bands were later used for screening the
measuring morphological traits. The plants were 5 years old genomic DNA of 98 progeny plants from the population.
and maintained in the experimental orchard of the Horticul- To identify polymorphic bands, we scored stably and
tural Sciences Department at the University of Florida, consistently amplified DNA bands from PCR products
Gainesville, Florida. as ? for presence and - for absence of a specific band.
Those markers showing a 3:1 segregation ratio were then
DNA extraction, amplification, and primer screening scored as b- for presence and aa for absence of the bands
for each progeny. The markers showing a 1:1 segregation
Genomic DNA was extracted from leaf tissue using the ratio were scored for each progeny as ab for presence and
method described by Edwards et al. (1991). A total of 115 aa for absence of the bands. Each specific polymorphic
random 10-mer oligonucleotide primers from the A, B, C, DNA band was named first with the primer number fol-
E, F, G, O, Q, R, and T kits (Operon Technologies, Ala- lowed by the approximate size of the amplified fragments.
meda, CA) were tested for RAPD marker production. PCR JoinMap version 2.0 (Stam and Van Ooijen 1995) was used
amplification was conducted in a 50-ll reaction mixture for the linkage analysis. JoinMap is the only mapping
containing PCR buffer (50 mM KCl, 10 mM Tris–HCl pH program available that allows the construction of an inte-
9.0, 1% Triton X-100), 2 mM MgCl2, 200 lM of each grated map from what is defined as a cross-pollinator (CP)
dNTP, 0.4 lM primer, 2 units of Taq polymerase (Pro- population, which is a population resulting from a cross
mega, Madison, WI), and 30 ng genomic DNA. The between two heterogeneously heterozygous and homozy-
amplification reaction was carried out in a PTC-100 ther- gous diploid parents (2–4 alleles per locus, linkage phases
mocycler with 96 or 60 wells (MJ Research, Waltham, unknown). The linkage groups were established using a
MA) using a cycling profile of 93°C for 1 min (initial logarithm of the odds ratio (LOD) score of 4.0. Map dis-
denaturation), 42 cycles of denaturation at 92°C for 1 min, tances were determined as centiMorgans (cM) based on
primer annealing at 35°C for 1 min, and primer extension recombination frequencies using Kosambi’s mapping
at 72°C for 2 min, followed by final primer extension at function. After establishing a core map with a LOD level of
72°C for 10 min. The PCR products were separated in 4.0, the map was reconstructed using the lower LOD
1.5% agarose gels by electrophoresis in 19 TAE buffer scores, between 3.5 and 3.9, to distribute some of the
[509 (2 M Tris-base, 0.1 M EDTA pH 8.0, 5.7% glacial unlinked markers to the core map of each linkage group.
Citrus paradisi X Poncirus trifoliata
Morphological traits
80-9
X A total of 17 morphological traits related to tree, leaf, floral,
and fruit characteristics were measured and evaluated during
Citrus sinensis Citrus grandis
a 2-year period (Table 1). Measurement data for individual
Citrus reticulta
traits were initially analyzed using the Quattro Pro program,
and then the data were tested and analyzed for normal dis-
1169 X 2-6-10 tribution using the Statistical Analysis Software (SAS
Institute, Cary, NC) program. Correlations among the traits
were also determined using SAS.
20-23 X 124-10
Tree characteristics
Progeny Population
The height of each individual from the progeny population
Fig. 1 Genetic background of the progeny population was measured using an extension pole at the beginning and

123
50 Plant Biotechnol Rep (2012) 6:47–57

Table 1 Characterization of
Morphological trait n Mean Standard deviation Min Max
morphological traits analyzed
Tree characteristics
Tree height, TH (cm) 98 348.1 90.3 68.0 507.0
Tree growth, TG (cm) 98 64.4 35.3 4.0 145.0
Tree canopy width, TCW (cm) 98 220.9 86.0 32.0 460.0
Tree vigor, TV 98 0.69 0.21 0.30 1.38
Trunk diameter, TD (cm) 98 6.7 2.8 1.0 13.0
Trunk diameter growth, TDG (cm) 98 1.6 0.8 0.2 3.4
Leaf characteristics
Leaf length, LL (cm) 98 6.31 1.09 3.00 8.36
Leaf width, LW (cm) 98 3.48 0.66 1.66 5.04
Flower characteristics
Petal length, PL (cm) 43 1.78 0.32 1.16 2.55
Petal width, PW (cm) 43 0.88 0.17 0.47 1.22
Length of pistil, LP (mm) 43 9.4 2.8 1.0 15.0
Length of style, LS (mm) 43 5.5 2.1 0.5 10.4
Anther number, AN 43 23.5 3.3 18.0 32.0
Petal number, PN 43 5.1 0.7 4.0 8.0
Fruit characteristics
n, Number of progeny analyzed; Fruit length, FL (cm) 48 7.01 1.30 4.45 11.06
Min, minimum value for a Fruit width, FW (cm) 48 6.91 1.04 4.94 9.78
specific trait, Max, maximum Fruit segment number, FSN 48 10.4 1.2 8.0 13.0
value for a specific trait

end of the growth period in the spring and fall, respec- flowers from each flowering progeny were collected near
tively. For each individual progeny, the fall measurement the full bloom period, and numbers of petals (PN) and
was used for tree height (TH), and the difference between anthers (AN) were counted for each flower. The length
the spring and fall measurements was used to determine the (PL) and width (PW) of petals, length of pistil (LP), and
periodic tree growth (TG). The tree canopy width (TCW) length of style (LS) from each flower were also measured
of each individual progeny tree was measured at the widest with a vernier caliper, and the average of the measurements
point of the canopy using an extension pole. The vigor was used for the analysis of each trait.
(TV) of each tree was determined by dividing the measured
canopy width by the measured TH. The trunk diameter Fruit characteristics
(TD) of each progeny was measured 10 cm above the
ground using tree calipers. The measurement was repeated Five fruit samples were collected from 48 fruiting plants
twice in the spring and fall, and the TD differences out of 98 progeny plants and evaluated. The length (FL)
between the two measurements were evaluated as trunk and width (FW) of fruits were measured using a vernier
diameter growth (TDG). caliper, and the average of five measurements for the
length and width of each fruit was used for analysis. Fruits
Leaf characteristics were horizontally cut at their widest point and the number
of segments for each fruit counted; the average was used
Five leaves were collected from each progeny and their for the analysis of fruit segment number (FSN).
length (LL) and width (LW) measured using a ruler. In the
progeny plants having bifoliate and trifoliate leaf types, QTL analysis
only the terminal leaves were measured. The averages of
the measurements for LL and LW from each tree were used The association of markers with QTLs for each trait was
for the analysis. tested using the MAPQTL 3.0 program (Van Ooijen and
Maliepaard 1996). This program assesses the degree of
Floral characteristics association between measured traits and marker genotypes
using the interval mapping method, which searches for
Floral characteristics were measured in 43 flowering plants effects of QTLs by using sets of markers in a linkage map
out of the 98 individuals in the progeny population. Ten and a selected LOD score. The associations of markers and

123
Plant Biotechnol Rep (2012) 6:47–57 51

QTLs for each trait were analyzed at every 0.5 cM in the linkage map is 530.1 cM, with an average map distance of
linkage map. Putative QTLs for each trait were detected 7.68 cM between markers. The number of markers mapped
when a LOD score of C3.0 was observed. LOD peaks for to individual linkage groups ranged between four (LG VI)
each detected QTL were used to locate the QTLs on the and 15 (LG V), with LG lengths varying from 11.0 (LG IV)
linkage map. to 83.8 cM (LG IX). More than 50% of the total markers
were distributed into linkage groups III, V, and VII
(Fig. 3). It has been estimated that the Citrus genome
Results ranges from 1500 to 1700 cM (Liuo 1990; Jarrell et al.
1992). Thus, we estimated that the map generated in this
Morphological trait analysis study covers at least 31–35% of the genome.

Measurable variation was observed in all of the morpho- QTL analysis

logical traits listed in Table 1. To the best of our knowledge,
our study is the first to examine this type of morphological The analysis of 17 different morphological traits related to
data in a segregating Citrus population. For most of the leaf, tree, leaf, floral, and fruit characteristics revealed 18
fruit, and flower characteristics, the range in variation was potential QTLs with an LOD C 3 (Table 2). These QTLs
approximately two- to threefold. However, the range in LP were associated with six tree characters (TD, TDG, TH,
and LS values was 15- and 20-fold, respectively. There was TG, TCW, and TV), two leaf characters (LL and LW), four
also a great variation in tree characteristics during the floral characters (PN, PL, PW, and AN), and one fruit
experiment: a sevenfold difference in TH, a 14-fold differ- character (FSN). No QTLs were detected for the LP, LS,
ence in TCW, a 13-fold difference in TD, and a 36-fold FW, and FL traits. LOD scores of the detected QTLs were
difference in TDG. Statistical analysis of the measured data between 3.08 (TD8.1.2) and 5.03 (TG1.1.1), and interval
for these traits revealed that most of them showed normal or lengths of the QTLs ranged from 0.5 cM (FS4.1.1) to
close to normal distribution (Fig. 2a), thereby demonstrating 21.3 cM (PN3.1.1) (Table 2; Fig. 2).
the presence of continuous variation. This result indicates QTLs associated with tree characters, including TD, TH,
that the traits are possibly controlled by several or many TCW, and TDG, were mapped in the same linkage groups
genes and inherited quantitatively. (LG VII and LG VIII) and located in the same regions of
these linkage groups (Fig. 2). These four traits showed
Generation of RAPD markers and construction significant positive correlation r values ranging from 0.58
of the genetic linkage map to 0.82. These findings suggest that at least two genes
involved in determining all of these tree characteristics are
A total of 115 random primers were used for the initial located in these linkage groups. That these different but
screening of the parents and five progeny. Among these, 22 correlated characters mapped to the same place also gives
primers failed to amplify any DNA bands, and 57 primers us confidence in our map. Morphological data for flower
either produced no polymorphism or were inconsistent. and fruit characteristics were obtained from a smaller
These primers were not used for further analysis. The number of progeny plants because fewer than 50% of the
remaining 36 random primers (31% of all primers tested) progeny plants were flowering and fruiting at the time of
produced consistent and obvious polymorphic DNA frag- data collection. Nevertheless, some QTLs with significant
ments and were subsequently used for screening the LOD scores were detected. QTLs associated with flower
genomic DNA of 98 plants from the progeny population and fruit characters were all mapped to the same linkage
for the segregation and genetic linkage analysis. A total of groups (LG III, LG IV, LG VII, and LG VIII) in different
347 bands were produced, and a single primer amplified regions of the map from where QTLs associated with tree
2–16 DNA bands ranging from 200 bp to 3 kb. Of these and leaf traits were detected.
amplified bands, 83 showed reproducible polymorphism
and were scored as RAPD markers for linkage mapping.
Linkage analysis of the 83 available markers with Discussion
JoinMap generated a linkage map containing 69 RAPD
markers distributed into nine linkage groups (LGs) possibly Most morphological traits in Citrus are suspected to dis-
corresponding to the nine haploid chromosomes of Citrus play quantitative inheritance; however, the genetic basis of
(Fig. 3). The stability and reproducibility of RAPD mark- inheritance of these traits has not been studied extensively
ers were confirmed in this study because more than 85% of and genetic information on these traits is very limited. We
the polymorphic bands showed linkage and mapped to a studied the genetics of morphological traits in Citrus using
specific linkage group. The total maximum length of the linkage mapping with molecular markers and QTL analysis

123
52 Plant Biotechnol Rep (2012) 6:47–57

24 24

24 20 20
20 16
16
16
12 12
12
8 8
8
4 4 4

0 0 0
60 150 240 330 420 510 0 86 172 258 344 430 0 0.2 0.4 0.6 0.8 1 1.2

24
20 20
21
16 18
16
15
12 12
12
8 9 8
6
4 4
3
0 0
0 30 60 90 120 150 0 3.6 7.2 10.8 0
0 0.99 1.98 2.97

20 12
25
10
20 16
8
15 12
6
10 8
4
5 4 2

0 0 0
2.4 3.6 4.8 6 7.2 8.4 1.36 2.26 3.16 4.06 4.96 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75

10 12
20

8 10
16
8
6 12
6
4 8
4
2 4 2

0 0 0
0.3 0.5 0.7 0.9 1.1 1.3 1.5 3.6 4.4 5.2 6.0 6.8 7.6 8.4 0 2 4 6 8 10 12

10
14 9
8 12 8
7
10
6 6
8 5
4 6 4
4 3
2 2
2 1
0 0 0
4.5 5.5 6.5 7.5 8.5 9.5 10.5 4 5.4 6.8 8.2 9.6 11 12.4 7.5 8.5 9.5 10.5 11.5 12.5 13.5

Fig. 2 Distribution of mean values of morphological traits from 98 individual progeny plants

123
Plant Biotechnol Rep (2012) 6:47–57 53

LG I LG II LG III LG IV LG V

0 E15043 0 E07065 0 B05037 0 Q18026 0 E04090

2 C02072

FS4.1.1
8 C08043 8 E07094 8 R12041
11 G10117
16 E04093 15 A04077
18 E06127 17 C02063

AN3.1.1

LW5.1.2
27 E04082 27 C02092

PN3.1.1
31 Q01040
32 A18117 34 Q01078

TV5.1.1
35 E18099 36 C13166 36 B01091
39 B01042 39 C04044 E16104
42

LW5.2.2
46 G05045 A02106
45 Q18041 45 R09140
47 C05145 B18141
48 E15163
48 E14068
50 R08043
C12039 50 Q18096
53
TG1.1.1

57 C13060
60 B10099 60 B05152
62 A18050

68 E14026 67 C19060

LG VI LG VII LG VIII LG IX
TCW8.1.2
TDG7.1.2

TD8.1.2
TH8.1.2
TH7.1.2
TCW7.1.2

0 C19115 0 A18034 0 E17064 0 C08054

TD7.1.2

5 C02038 7 B05096

16 A18062
19 C11112
22 E14034 22 R02154
24 R12056
27 C08111
29 Q01117
31 E15050 32 Q01102
PL7.1.1

TDG8.1.2

44 A02133
47 Q18118
48 Q18105
PW8.1.1

50 O13052 51 C13070 52 B17069

55 B04054
58 E15091 59 C04096
LL8.1.1

64 B17088

76 C11097

84 G11059

Fig. 3 Genetic linkage map showing the locations of quantitative trait the random amplified polymorphic DNA (RAPD) markers. Bars to the
loci (QTLs) associated with tree, leaf, flower, and fruit characteristics as left of the LGs correspond to the location of the QTL. Specific traits are
detected by MAPQTL and based on a LOD score of [3.0. Roman indicated by their name code (as listed in Tables 1 and 2) followed
numerals at the top of each group indicate the linkage group (LG) sequentially by the LG number, order of the QTL in that LG, and by the
number. Numbers on the left of each LG show map distances in number of total QTLs for the same trait. Horizontal lines from bars
centiMorgans (cM), codes on the right of each LG indicate the names of show the most likely position of the QTL

to improve our knowledge and understanding of the intergeneric cross. Seventeen different morphological traits
genetics of these traits. A linkage map was constructed related to tree, leaf, flower, and fruit characters were
with RAPD markers using progeny from a complex evaluated and analyzed for QTLs using the linkage map to

123
54 Plant Biotechnol Rep (2012) 6:47–57

Table 2 Quantitative trait loci associated with measured traits detected using the interval mapping method (LOD [ 3)
Trait QTLa Markersb LGc Interval QTL LODf % QTL
lengthd (cM) positione (cM) effectg

Trunk diameter (TD) TD7.1.2 A18034–A18062 VII 12.5 4.5 4.65 22.7
TD8.1.2 E17064–B05096 VIII 2 3.5 3.08 46.8
Trunk diameter growth (TDG) TDG7.1.2 A18034–A18062 VII 6.5 4.5 3.25 16
TDG8.1.2 Q01117–C13070 VIII 18.5 50.9 3.99 62.4
Tree height (TH) TH7.1.2 A18034–A18062 VII 9 4.5 3.59 19.0
TH8.1.2 E17064–B05096 VIII 2.5 3 3.28 48.2
Tree growth (TG) TG1.1.1 C12039–E14026 I 5 58.5 5.03 68.3
Tree canopy width (TCW) TCW7.1.2 A18034–A18062 VII 12 5.5 3.99 20.9
TCW8.1.2 E17064–B05096 VIII 4 2 3.93 38.7
Tree vigor (TV) TV5.1.1 Q01078–G05045 V 6.4 40.2 3.27 30.7
Leaf length (LL) LL8.1.1 B04054–C11097 VIII 8.2 64.2 3.36 33.1
Leaf width (LW) LW5.1.2 C02092–Q01078 V 1 26.7 3.14 32.8
LW5.2.2 E16104–G05045 V 1 45.1 3.59 28.8
Petal number (PN) PN3.1.1 E06127–E15163 III 21.3 44.9 3.81 60.1
Petal width (PW) PW8.1.1 B04054–B17088 VIII 1 55 3.16 62.7
Petal length (PL) PL7.1.1 C08111–Q18118 VII 3 39.7 3.12 71.4
Anther number (AN) AN3.1.1 E06127–Q01040 III 5 25.1 3.84 77.8
Length of pistil (PL) NS
Length of style (LS) NS
Fruit width (FW) NS
Fruit length (FL) NS
Fruit segment (FS) FS4.1.1 C02072-E07094 IV 0.5 7.8 3.20 26.5
a
Quantitative trait loci (QTLs) are named according to trait abbreviations followed by the linkage group (LG) number, order of the QTL in that
LG, and by the number of total QTLs for the same trait
b
Names of the random amplified polymorphic DNA (RAPD) markers between which the QTLs are located
c
Linkage group number on which the QTLs are located
d
Length in centiMorgans where the QTL associations with a specific trait are significant (LOD C 3.0)
e
Location of QTL on the LG where the LOD score reaches the highest value
f
Logarithm of the odds ratio (LOD) values
g
Percentage of the variation explained by the QTL

determine genomic regions associated with the inheritance phenotypes because our population appeared more vari-
of these traits. able—at least morphologically—than the BC1 population,
The percentage of random primers that produced poly- which is why it was selected for analysis.
morphism in the progeny population used in this study Seventy-one of the markers were hypothesized to be
(31%) was disappointingly low—and considerably lower segregating in a 1:1 ratio, while 12 markers were tested for
than that reported in a previous study with BC1 progeny 3:1 segregation. Chi-square analysis demonstrated that all
from Citrus grandis 9 [Citrus grandis 9 Poncirus trifoli- of the loci expected to segregate 3:1 and 52 of the 71 loci
ata] in which 49% of the primers revealed polymorphism expected to segregate 1:1 fit the expected segregation ratio.
(Cai et al. 1994). This lower polymorphism might be due to Distorted segregations (P B 0.05) were observed for 19
differences in the characteristics of the mapping popula- markers (approx. 23%), with 12 of these skewed towards
tions; it would appear that there is less polymorphism for the 124-10 parental genotype and seven skewed towards
marker alleles within Citrus than there is between Citrus the 20–23 parental genotype. Relatively low distorted
and Poncirus, and there was a higher frequency of Citrus segregation was observed in our study compared to that
alleles in this population than in the BC1 population eval- reported in two previous studies on Citrus, which showed
uated by Cai et al. (1994). Our result indicates that poly- 37 and 40% distorted segregation with RFLP markers
morphism for molecular markers is not always highly (Durham et al. 1992) and RAPD markers (Cai et al. 1994),
correlated with the range of segregating morphological respectively. On the other hand, the level of distorted

123
Plant Biotechnol Rep (2012) 6:47–57 55

segregation observed in our study was similar to that References

reported using microsatellite (Kijas et al. 1997) and iso-
zyme (Jarrell et al. 1992) markers in Citrus. Abdel-Haleem H, Lee GJ, Boerma RH (2011) Identification of QTL
for increased fibrous roots in soybean. Theor Appl Genet
With the available software packages, there are only two 122(5):935–946
ways to perform linkage analysis of a CP population such Asins MJ, Bernet GP, Ruiz C, Cambra M, Guerri J, Carbonell EA
as this. One method is to construct separate maternal and (2004) QTL analysis of Citrus tristeza virus-citradia interaction.
paternal maps using only co-dominant markers and mark- Theor Appl Genet 108:603–611
Benter T, Papadopoulos S, Pape M, Manns M, Poliwoda H (1995)
ers in the same phase; this has been designated a pseudo- Optimization and reproducibility of random amplified polymor-
testcross strategy (Grattapaglia and Sederoff 1994). The phic DNA in humans. Anal Biochem 230:92–100
alternative is to use JoinMap 2.0 and the CP option to Bernet GP, Margaix C, Jacas J, Carbonell EA, Asins MJ (2005)
construct a linkage map by integrating the maternal and Genetic analysis of citrus leafminer susceptibility. Theor Appl
Genet 110:1393–1400
paternal segregation data (Stam and Van Ooijen 1995). Cai Q, Guy CL, Moore GA (1994) Extension of the linkage map in
Among the 83 polymorphic loci tested for linkage map- Citrus using random amplified polymorphic DNA (RAPD)
ping, 42 showed an aa 9 ab genotype in which a band was markers and RFLP mapping of cold-acclimation responsive loci.
absent in parent 20-23 and present in parent 124-10, and 29 Theor Appl Genet 89:606–614
Cameron JW, Frost HB (1968) Genetics, breeding, and nucellar
were identified as ab 9 aa where a band was present in embryony. In: Reuther W, Batchelor LD, Webber HJ (eds) The
parent 20-23 and absent in parent 124-10. These patterns citrus industry. Division of Agricultural Sciences, University of
were expected to segregate in a 1:1 ratio. The remaining 12 California, Riverside, pp 325–370
markers were classified as ab 9 ab in which a band was Cameron JW, Soost RK (1980) Leaf types of F1 hybrids and
backcrosses involving unifoliate Citrus and trifoliate Poncirus.
present in both parents and was expected to segregate in a J Am Soc Hortic Sci 105:517–519
3:1 ratio. These markers could be mapped in both parents if Caranta C, Palloix A, Lefebvre V, Daubeze AM (1997) QTLs for a
individual parental maps were constructed. However, this component of partial resistance to cucumber mosaic virus in
approach would allow only 54 markers to be analyzed for pepper: restriction of virus installation in host cells. Theor Appl
Genet 94:431–438
parent 124-10 and 41 markers to be analyzed for parent Cerqueira-Silva CB, Cardoso-Silva CB, Santos ES, Conceição LD,
20-23. Attempts were made to generate co-dominant Pereira AS, Oliveira AC, Corrêa RX (2010) Genetic diversity in
markers by evaluating the microsatellite markers devel- wild species of passion fruit (Passiflora trintae) based on
oped by Kijas et al. (1995, 1997) (data not shown). The molecular markers. Genet Mol Res 26 9(4):2123–2130
Cheng FC, Roose ML (1995) Origin and inheritance of dwarfing by
goal was to both add more markers to the parental maps the citrus rootstock Poncirus trifoliata cv. Flying Dragon. J Am
and to anchor these to other previously established maps. Soc Hortic Sci 120:286–291
The primer sequences used to detect polymorphisms in Cheng Y, Geng J, Zhang J, Wang Q, Ban Q, Hou X (2009) The
other populations either did not amplify bands or did not construction of a genetic linkage map of non-heading Chinese
cabbage (Brassica campestris ssp. chinensis Makino). J Genet
yield polymorphism. Because of the relative paucity of Genomics 36(8):501–508
markers and their relatively even distribution for mapping Deng ZN, Gentile A, Nicolosi E, Domina F, Vardi A, Tribulato E
in the two parents, we opted to use JoinMap 2.0 and the CP (1995) Identification of in vivo and in vitro lemon mutants by
option to produce an integrated map, recognizing that the RAPD markers. J Hortic Sci 70:117–125
Douhan GW, Smith ME, Huyrn KL, Westbrook A, Beerli P, Fisher
map derived would be a statistical compromise. AJ (2008) Multigene analysis suggests ecological speciation in
In summary, we have made a first start on the charac- the fungal pathogen Claviceps purpurea. Mol Ecol 17(9):2276–
terization of a number of morphological traits in Citrus. It 2286
was unexpectedly difficult to identify RAPD polymor- Durham RE, Liou PC, Gmitter FG Jr, Moore GA (1992) Linkage of
restriction fragment length polymorphisms and isozymes in
phisms in the selected population, which segregated widely citrus. Theor Appl Genet 84:39–48
for morphological traits. The linkage map reported here Edwards KJ, Johnstone C, Thompson C (1991) A simple and rapid
needs to be extended and integrated with other linkage method for the preparation of plant genomic DNA for PCR
maps of Citrus by developing co-dominant markers that analysis. Nucleic Acids Res 19:1349–1355
Fang DQ, Federici CT, Roose ML (1997) Development of molecular
segregate in the different mapping populations available. markers linked to a gene controlling fruit acidity in citrus.
However, the derived map could be used to identify some Genome 40(6):841–849
putative QTLs for morphological markers that can be more Fani R, Damiani G, Di SC, Gallori E et al (1993) Use of random
thoroughly characterized and used in marker-assisted amplified polymorphic DNA (RAPD) for generating specific
DNA probes for microorganisms. Mol Ecol 2:243–250
selection in breeding programs. Finger FL, Lannes SD, Schuelter AR, Doege J, Comerlato AP,
Gonçalves LS, Ferreira FR, Clovis LR, Scapim CA (2010)
Acknowledgments The authors thank Courtney Weber and Genetic diversity of Capsicum chinensis (Solanaceae) accessions
Antoinette Sankar for their technical assistance and F.G. Gmitter, Jr. based on molecular markers and morphological and agronomic
for assistance with the software. This research project was funded in traits. Genet Mol Res 21 9(3):1852–1864
part by the Florida Citrus Production Research Advisory Council.

123
56 Plant Biotechnol Rep (2012) 6:47–57

Foolad MR, Chen FQ (1998) RAPD markers associated with salt Moreno E, Obando JM, Dos-Santos N, Fernández-Trujillo JP,
tolerance in an interspecific crosses of tomato (Lycopersicon Monforte AJ, Garcia-Mas J (2008) Candidate genes and QTLs
esculentum x L pennellii). Plant Cell Rep 17:306–312 for fruit ripening and softening in melon. Theor Appl Genet
Garcia MR, Asins MC, Carbonell EA (2000) QTL analysis of yield 116:589–602
and seed number of Citrus. Theor Appl Genet 101:487–493 Orr W, Molnar SJ (2008) Development of PCR-based SCAR and
Gelderman H (1975) Investigations on inheritance of quantitative CAPS markers linked to beta-glucan and protein content QTL
characters in animals by gene markers. Theor Appl Genet regions in oat. Genome 51(6):421–425
46:319–330 Paran I, Goldman I, Zamir D (1997) QTL analysis of morphological
Genc Y, Oldach K, Verbyla AP, Lott G, Hassan M, Tester M, traits in a tomato recombinant inbred line population. Genome
Wallwork H, McDonald GK (2010) Sodium exclusion QTL 40:242–248
associated with improved seedling growth in bread wheat under Paterson AH, Lander ES, Hewitt JD, Peterson S, Lincoln SE,
salinity stress. Theor Appl Genet 121(5):877–894 Tanksley SD (1988) Resolution of quantitative traits into
Gmitter FG Jr, Grosser JW, Moore GA (1992) Citrus. In: Hammers- Mendelian factors using a complete linkage map of restriction
chlag FA, Litz RE (eds) Biotechnology of perennial fruit crops. fragment length polymorphisms. Nature 335:721–726
CAB International, Wallingford, pp 335–369 Pérez-Vega E, Pañeda A, Rodrı́guez-Suárez C, Campa A, Giraldez R,
González C, Camacho MV, Benito C (2002) Chromosomal location Ferreira JJ (2010) Mapping of QTLs for morpho-agronomic and
of 46 new RAPD markers in rye (Secale cereale L.). Genetica seed quality traits in a RIL population of common bean
115(2):205–211 (Phaseolus vulgaris L.). Theor Appl Genet 120(7):1367–1380
Gostimsky SA, Kokaeva ZG, Konovalov FA (2005) Studying plant Raina RM, Rani SN, Kojima V, Ogihara T, Singh Y, Devarumath KP
genome variation using molecular markers. Russ J Genet (2001) RAPD and ISSR fingerprints as useful genetic markers
41:378–388 for analysis of genetic diversity, varietal identification, and
Grattapaglia D, Sederoff R (1994) Genetic linkage maps of Eucalyp- phylogenetic relationships in peanut (Arachis hypogaea) culti-
tus grandis and Eucalyptus urophylla using a pseudo-testcross: vars and wild species. Genome 44:763–772
mapping strategy and RAPD markers. Genetics 137:1121–1137 Russo G, Fanizza G (1992) Genotypic variability and interrelation-
Gyenis L, Yun SJ, Smith KP, Steffenson BJ, Bossolini E, Sanguineti ship among morphological and biochemical fruit characters in
MC, Muehlbauer GJ (2007) Genetic architecture of quantitative mandarins. Proc Int Soc Citric 1:96–97
trait loci associated with morphological and agronomic trait Schoenenberger N, Felber F, Savova-Bianchi D, Guadagnuolo R
differences in a wild by cultivated barley cross. Genome (2005) Introgression of wheat DNA markers from A, B and D
50(8):714–723 genomes in early generation progeny of Aegilops cylindrica
Hao YJ, Wen XP, Deng XX (2004) Genetic and epigenetic Host 9 Triticum aestivum L. hybrids. Theor Appl Genet 111(7):
evaluations of citrus calluses recovered from slow-growth 1338–1346
culture. J Plant Physiol 161(4):479–484 Siviero A, Cristofani M, Furtado EL, Garcia AA, Coelho AS,
James CM, Clarke JB, Evans KM (2004) Identification of molecular Machado MA (2006) Identification of QTLs associated with
markers linked to the mildew resistance gene Pl-d in apple. citrus resistance to Phytophthora gummosis. J Appl Genet
Theor Appl Genet 110:175–181 47:23–28
Jarrell DC, Roose ML, Traugh SN, Kupper RS (1992) A genetic map Soost RK, Cameron JW (1975) Citrus. In: Janick JM, Moore JN (eds)
of citrus based on the segregation of isozymes and RFLPs in an Advances in fruit breeding. Purdue University Press, West
intergeneric cross. Theor Appl Genet 84:49–56 Lafayette, pp 507–540
Kennard WC, Havey MJ (1995) Quantitative trait analysis of fruit Soost RK, Roose ML (1996) Citrus. In: Janick JM, Moore JN (eds)
quality in cucumber: QTL detection, confirmation, and compar- Fruit breeding. Wiley, New York, pp 257–323
ison with mating-design variation. Theor Appl Genet 91:53–61 Souza EA, Camargo OA Jr, Pinto JM (2010) Sexual recombination in
Khowaja FS, Norton GJ, Courtois B, Price AH (2009) Improved Colletotrichum lindemuthianum occurs on a fine scale. Genet
resolution in the position of drought-related QTLs in a single Mol Res 9(3):1759–1769
mapping population of rice by meta-analysis. BMC Genomics Stam P, Van Ooijen JW (1995) JoinMap (TM) version 20: software
10:276 for the calculation of genetic linkage maps. Version 2.0 ed
Kijas JMH, Fowler JCS, Garbett CA, Thomas MR (1995) An CPRO-DLO, Wageningen
evaluation of sequence tagged microsatellite markers for genetic Tondelli AFE, Barabaschi D, April A, Skinner JS, Stockinger EJ,
analysis within Citrus and related species. Genome 38:349–355 Stanca AM, Pecchioni N (2006) Mapping regulatory genes as
Kijas JMH, Thomas MR, Fowler JCS, Roose ML (1997) Integration candidates for cold and drought stress tolerance in barley. Theor
of trinucleotide microsatellites into a linkage map of Citrus. Appl Genet 112:445–454
Theor Appl Genet 94:701–706 Tozlu I, Guy CL, Moore GA (1999a) QTL analysis of Na and Cl
Lee YC, Yang VC, Wang TS (2007) Use of RAPD to detect sodium accumulation related traits in an intergeneric BC1 progeny of
arsenite-induced DNA damage in human lymphoblastoid cells. Citrus and Poncirus under saline and nonsaline environments.
Toxicology 239(1–2):108–115 Genome 42:692–705
Ling P, Duncan LW, Deng ZB, Dunn D, Hu X, Huang S, Gmitter FG Tozlu I, Guy CL, Moore GA (1999b) QTL analysis of morphological
Jr (2000) Inheritance of citrus nematode resistance and its traits in an intergeneric BC1 progeny of Citrus and Poncirus under
linkage with molecular markers. Theor Appl Genet 100:1010– saline and nonsaline environments. Genome 42:1020–1029
1017 Van Ooijen JW, Maliepaard C (1996) MapQTL (TM) version 30
Liou PC (1990) A molecular study of the Citrus genome through software for the calculation of QTL positions on genetic maps.
analysis of restriction fragment length polymorphism and Version 3.0 ed CPRO-DLO, Wageningen
isozyme mapping. PhD thesis. University of Florida, Gainesville Vega AE, Cortiñas TI, Puig ON, Silva HJ (2010) Molecular
Mian MAR, Ashley DA, Boerma HR (1998) An additional QTL for characterization and susceptibility testing of Helicobacter pylori
water use efficiency in soybean. Crop Sci 38:390–393 strains isolated in western Argentina. Int J Infect Dis 14[Suppl 3]:
Mohan M, Nair S, Bhagwat A, Krishna TG, Yano M, Bhatia CR, e85–e92
Sasaki T (1997) Genome mapping, molecular markers and Venkatachalam L, Sreedhar RV, Bhagyalakshmi N (2008) The use of
marker-assisted selection in crop plants. Mol Breed 3:87–103 genetic markers for detecting DNA polymorphism, genotype

123
Plant Biotechnol Rep (2012) 6:47–57 57

identification and phylogenetic relationships among banana Xue DW, Chen MC, Zhou Chen S, Mao Y, Zhang H (2008) QTL
cultivars. Mol Phylogenet Evol 47:974–985 analysis of flag MX, leaf in barley (Hordeum vulgare L) for
Villalta I, Reina-Sánchez A, Boları́n MC, Cuartero J, Belver A, morphological traits and chlorophyll content. J Zhejiang Univ
Venema K, Carbonell EA, Asins MJ (2008) Genetic analysis of Sci B 9:938–943
Na(?) and K (?) concentrations in leaf and stem as physiolog- Zhang YX, Wang Q, Jiang L, Liu LL, Wang BX, Shen YY, Cheng
ical components of salt tolerance in Tomato. Theor Appl Genet XN, Wan JM (2011) Fine mapping of qSTV11 (KAS), a major
16:869–880 QTL for rice stripe disease resistance. Theor Appl Genet
Weber CA, Moore GA, Deng Z, Gmitter FG Jr (2003) Mapping 122(8):1591–1604
freeze tolerance quantitative trait loci in a Citrus grandis 9 Pon- Zwonitzer JC, Bubeck DM, Bhattramakki D, Goodman MM,
cirus trifoliata F1 pseudo-testcross using molecular markers. Arellano C, Balint-Kurti PJ (2009) Use of selection with
J Am Soc Hortic Sci 128:508–514 recurrent backcrossing and QTL mapping to identify loci
Welsh JM, McClelland M (1990) Fingerprinting genomes using PCR contributing to southern leaf blight resistance in a highly
with arbitrary primers. Nucleic Acids Res 18:7213–7218 resistant maize line. Theor Appl Genet 118:911–925
Williams JGK, Kubelic AR, Livak KJ, Rafalski JA, Tingey SV (1990)
Polymorphisms amplified by arbitrary primers are useful as
genetic markers. Nucleic Acids Res 18:6531–6535

123

View publication stats