This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/authorsrights
Transcript
This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies areencouraged to visit:
Contents lists available at SciVerse ScienceDirect
Scientia Horticulturae
journa l h o me page: www.elsev ier .com/ locate /sc ihor t i
Agronomic traits and vegetation indices of two onion hybrids
S. Marinoa,∗, B. Bassob, A.P. Leonec, A. Alvinoa
a Department of Agricultural, Environmental and Food Sciences (DAEFS), University of Molise, Via De Sanctis, I-86100, Campobasso, Italyb Department of Geological Science and W.K. Kellogg Biological Station, Michigan State University, 307 Natural Science Building, East Lansing, MI48824-1115, USAc CNR—ISAFoM, via Patacca 85, Ercolano, 80056, Italy
a r t i c l e i n f o
Article history:Received 20 November 2012Received in revised form 5 February 2013Accepted 8 March 2013
Environmental, genotypic and agronomic factors have an effect on the yield value of an onion crop, whichis determined primarily by number, weight and size of bulbs. Spatial variability of soil properties affectscrop yield. Remote sensed hyperspectral vegetation indices (VIs), calculated using crop reflectance at fieldscale can be used either as an index of the plant biophysical status, or as a tool to estimate crop variability.The aim of the study was to evaluate the relationships among traditional agronomic measurements oftwo irrigated onion hybrids (Cometa and Red Mech) with spectroradiometric measurements taken atfield scale. The two hybrids differed significantly either for agronomic response (yield, yield componentsand distribution of yield classes) or for their spectral properties. Cometa showed significant higher yieldand biomass than Red Mech, as well as significant higher VI values, although no correlations were foundamong agronomic parameters and spectroradiometric indices. On the contrary, Red Mech showed sig-nificant correlation among biomass, yield and bulb weight with VIs. Differences between the two onionhybrids in the spectroradiometric readings and agronomic traits underlined the importance of groundtruth data verification when air-born images or satellite data are taken over onion crop field.
Onion, a member of Amaryllidaceae family, has been widely usedeven in ancient times as seasonings, foods and for medical uses. Incurrent times, onion is one of the most important vegetable cropsgrown in the world (Kumar et al., 2007b). It is valued for its distinc-tive pungent flavor and it is an essential ingredient of the cuisine inmany regions of the world (Sharma et al., 2005). The total surfacecultivated in the world is about 3.6 million ha with a productionof 72 Mt (FAOSTAT data, 2011). In Italy, onion is produced eitherby small farmers or commercial growers on 12.800 ha with a totalproduction of 0.385 Mt (FAOSTAT data, 2011).
Growth and development of onion relies on interactions amonggenotype, agronomic practices and pedo-climatic conditions (Hayand Walker, 1989). The crop value is determined by the yield andsize of bulbs and experiences high spatial variability. Genetic fac-tors affect the variability of morpho-physiological characteristics,photoperiod, harvesting date, bulb yield, size, color, bulb shape andflavonoid contents, onion pungery, sulfur content of bulbs, solublesolids content, maturity, storage ability and bulbing response under
different day length (Marotti and Piccaglia, 2002; Yoo et al., 2006;Magruder et al., 1941; Randle, 1992a,b).
Hybrids of onion in many cases have increased the uniformity ofbulb yield and yield security (Jones and Davis, 1944; Havey, 1991,1993; Currah and Proctor, 1990).
Environment affects the variability of onion pungency (Yooet al., 2006; Randle et al., 1998), the nitrogen/sulphur (N/S)requirement (Randle, 2000; Coolong and Randle, 2003; McCallumet al., 2005), weed competition (which is extremely sensitive inonion) (Bleasdale, 1959; Wicks et al., 1973; Menges and Tamez,1981; Glaze, 1987), and the water requirement (Kumar et al.,2007b).
Agronomic practices, such as planting dates and density, nitro-gen fertilization and irrigation can have a significant influence ononion plant characteristics (i.e., biomass, leaf area index, watercontent and chlorophyll concentration), yield and yield compo-nents and quality (Pasricha and Abrol, 2003; Dubuis and Mauch,2003; Haneklaus et al., 2001; Martin de Santa Olalla et al., 2004;Kumar et al., 2007a; Hay and Walker, 1989; Bleasdale, 1959;Hartridge-Esh and Bennet, 1980; McGeary, 1985; Jones and Davis,1944).
An onion crop is sold well in advance of bulb maturity on thebasis of on-farm visual assessment and on the basis of good repu-tation of the farmers and their ability to manage spatial variabilityto obtain the highest marketable yield. Marketable yield is a com-promise of yield and size of bulbs (Shock et al., 2004; de Visser and
S. Marino et al. / Scientia Horticulturae 155 (2013) 56–64 57
van den Berg, 1998). Increasing agronomic yield (e.g., t/ha−1) doesnot necessarily lead to higher economic benefits.
For these reasons, yield spatial variability monitoring across thefield is important for site-specific management and to optimizeagronomic practices (Qarallah et al., 2008).
Remote and proximal sensing data can provide non-destructively and instantaneously quantitative informationabout the most different agricultural crops. Spectral reflectancemeasurements can be used to monitor plant conditions (Gibbonsand Freudenberger, 2006; Ji-Hua and Bing-Fang, 2008), estimateplant nutrient status (Nguyen and Byun-Woo, 2006; Zhao et al.,2005), detect abiotic and biotic stresses (Clay et al., 2006), estimateplant growth rate (Beck et al., 2007) and forecast crop yields(Galvao et al., 2009; Li et al., 2007). In order to make better useof spectral data, the vegetation indices are calculated using cropreflectance at differing wavelengths.
Changes in plant characteristics, affect crop spectral reflectancein the visible-near infrared (vis-NIR) range. Changes in chloro-phyll content affect the reflectance of green plants in the visible(vis) region (350–700 nm) (Salisbury and Ross, 1969). Chloro-phyll loss causes an increase of reflectance in the entire visibleregion and a narrowing of the main chlorophyll absorptionband centered at 680 nm (Hardisky et al., 1983; Wessman, 1994,1990; Penuelas et al., 1997a), due to electron transition pro-cesses. Narrowing in chlorophyll absorption results in a shift ofthe inflection point between the red and near infrared region(or “red-edge”) to shorter wavelength (“blue shift”). Thus, thered-edge position, which is easily identified by calculating thefirst derivative of the reflectance spectrum, may be considereda useful spectral indicator of leaf chlorophyll content and indi-rectly infer about nitrogen status of the plant (Cammarano et al.,2011).
Changes in biomass (and related LAI) affect the reflectance in thenear infrared region (NIR), between 700 nm and 1300 nm. Leaf pig-ments have limited absorption in the NIR region and reflectancein this region is high (Carter, 1991), with about 55% of the inci-dent light being reflected, 40% transmitted and only 5% absorbed(Belward, 1991). In the same wavelength range, reflectance is alsogreatly influenced by cellular structure and refractive index discon-tinuities within the leaves (Knipling, 1970; Gausman et al., 1978).Owing to the fact that soil reflectance is lower than green vegetationreflectance in the NIR and higher in the VIS, as the fractional vege-tation cover increases the reflectance of a soil–vegetation mixturealso increases in the NIR, whereas it decreases in the VIS. An inversetrend can be observed as vegetation cover decreases (Guyot, 1989).
This significant difference between red and NIR bands is aunique spectral property that makes vegetation different from allother ground objects and that can be usefully used to monitorvegetation from distance. Differences between red and NIR canbe very effectively enhanced by differencing and ratio operations(Liu and Mason, 2009). One of the most common spectral ratiosused in studies of vegetation status is the ratio of the NIR to theequivalent red band value. More complex ratios involve sums ofand differences between spectral bands. The most popular oneis the so-called Normalized Difference Vegetation Index (NDVI)(Liu and Mason, 2009; Mather and Koch, 2011). A Green Normal-ized Difference Vegetation Index (GNDVI) has been also proposed(Gitelson et al., 1996), where the green band is used instead ofthe red band. Both NDVI and GNDVI are usually well related tovegetation amount until saturation at full canopy cover and aretherefore also well related to the biophysical properties of plantcanopies, such as the absorbed photosyntetically active radiation,efficiencies, and productivity (Rondeaux et al., 1996). The intrinsicindices, however, such as NDVI and GNDVI, are extremely sensitiveto soil background brightness (Huete et al., 1985), and are difficultto interpret with low vegetation cover. In order to overcome this
problem, several vegetation indices have been developed whichtake into account the linear relationship between soil visible andnear-infrared reflectance (soil line). Such indices attempt to reducethe influence of the soil by assuming that most spectra follow thesame soil line. The perpendicular vegetation index (PVI) expressesthe distance between the canopy red and NIR reflectance and thesoil line. Although better than NDVI at low vegetation densities, PVIis still significantly affected by the soil (Rondeaux et al., 1996). Sig-nificant improvements are found with the soil-adjusted vegetationindex (SAVI) (Huete, 1988). Several modifications have been madeto SAVI, and the transformed SAVI (TSAVI) (Baret and Guyot, 1991;Baret et al., 1989), modified SAVI (MSAVI) (Qi et al., 1994), opti-mized SAVI (OSAVI) (Rondeaux and Baret, 1996), generalized SAVI(GSAVI) (Gilabert et al., 2002), and second soil-adjusted vegetationindex (Major et al., 1990) were subsequently proposed. Anothervegetation index has been proposed, called the weighted differ-ence vegetation index (WDVI) (Clevers, 1989), which minimize thesoil background effect.
Changes in canopy water content also affect the canopy spectralresponse (Penuelas et al., 1993). They can be monitored by look-ing to specific spectral absorption bands of the water molecule inthe NIR. High-resolution reflectance in this region has already beentested as a method for estimating plant water concentration (Carter1991; Danson et al., 1992). For single leaves, the water absorp-tion bands in the region between 1300 nm and 2500 nm showedthe highest sensitivity to leaf water concentration (Carter 1991).However, atmospheric water vapour also strongly absorbs in thisspectral region making it difficult to use them for leaf water diag-nosis. Secondary effects of light absorption of the leaf water may beobserved at shorter wavelengths in the NIR region between 780 nmand 1300 nm. Although light absorption of leaf water is relativelyweak in this region (Curcio and Petty, 1951), it is less sensitive tothe atmospheric water vapour than in the 1300–2500 nm regionand it is very useful to monitor plant water status. Specifically, areflectance at 970 nm corresponding to a weaker water absorptionband has been proved to be useful for that purpose (Penuelas et al.,1993, 1996). The ratio between the reflectance at 970 nm/R970and the reflectance at a reference wavelength, 900 nm (R900) washighly correlated to plant relative water concentration (Penuelaset al., 1997b; Leone et al., 2007). The ratio R900/R970 is defined asWater Index (WI), which decreases with increasing water contentin the plant leaves.
In a study on 30 species (Gausman and Allen, 1973), onionshave shown interesting distinctive features in optical parameters(adsorption coefficient, infinite reflectance and scattering coeffi-cient) at 550 nm (green reflectance peak), 650 nm (chlorophyllabsorption band) and 850 nm (on infrared reflectance plateau).
Onion vegetative growth is divided into three stages (Brewster,1990). The first one is a period of slow growth, followed by a secondperiod of rapid leaf growth, where successive leaves grow larger,and a third phase when the bulb grows.
The use of remote sensing on onion crop growth at the earlystages has not yet been successful (Haack and Jampoler, 1995;Schotten et al., 1995) probably because an onion crop grows inrows and its biomass is relatively small (Bosch Serra and Casanova,2000); on the other hand Bosch Serra and Casanova (2000) havefound from six leaves to onset of bulbing a good indirect estima-tion of LAI from NDVI and RVI (more clearly during the secondphase of vegetative growth), and have also shown as NDVI is a goodestimator of biomass (in particular before bulbing).
The aim of the study was to understand spatial variability ofyield and yield components of two widely used irrigated onionhybrids at bulb stages. A traditional agronomic approach was linkedto an innovative local scale technique intended to fill the gap in theliterature on the key factors affecting the yield variability responseof two onion hybrids.
Author's personal copy
58 S. Marino et al. / Scientia Horticulturae 155 (2013) 56–64
The results may contribute to understand the ways to assessthe agronomic value of an onion crop, in particular in the coun-tries which adopt contract farming (production sold by farmers todealers well before harvesting).
2. Materials and methods
2.1. Experimental conditions
The experiment was carried out during the crop-growing season2010 in Central Italy (latitude N 41◦52.855, longitude E 14◦55.728,30 m a.s.l.) on a plain dedicated to horticultural crops. The onionseedlings hybrids used in the study were: Cometa (white) and RedMach (Red). The Cometa hybrid (Nunhems, The Netherlands) bulbspresents a regular shape (conical medium long-tapered) to mediumgauges with white robes and high percentage of regular bulbs withdiameters between 40 mm and 80 mm, medium-high productiv-ity; resilience and low incidence to bud rot, high consistency ofbulbs in storage. The cultivar Red Mech (ISI Sementi, Italy) is a longday hybrid, suitable for spring sowing, is harvested in late July andAugust. The bulbs of good size, characterized by high homogeneity,have high-top form, full at the collar, the outer skins are thick andshow a intense and brilliant purplish red color. Red Mech can bestored all winter long and is particularly suitable for mechanicalharvesting. Both hybrids are used as control in all Italian exper-imental trials. Both hybrids were sown on 25th of January, at adensity of 75 seeds m−2; harvested lasted one week, starting firstweek of August.
A rectangular field of three hectares has been divided alongthe longitudinal axis into two adjacent plots, each of 1.5 ha, withdimensions of 30 × 500 m. Georeferred measurements were per-formed on the whole field, according to a rectangular regulargeoreferenced grid; Spectroradiometric measurements were inten-sified, following the same regular grid.
The texture of the soil was clay (sand 30%, silt 24%, clay 46%),with medium organic matter (1.83%), medium content in total N(1.16 g kg−1), moderately alkaline (pH 7.61).
Crop was sprinkler irrigated; irrigation pipes were 40 mm indiameter and the spray heads 50–22 PSI (NAAN, Israel). Weeds werecontrolled with recommended chemical herbicides. At harvest time(50 ± 60% of the leaves were dead), the diameter, weight and thenumber of bulbs were determined at 84 geo-referenced points of1 m2. At each sampling plots, plants were hand cut for whole plantdry mass accumulation (below and above ground) after oven dryingplant material at 75 ◦C until reaching constant weight. At 140 and147 days after sowing (DAS), leaf area index (LAI) was measuredby using a portable LI-COR LAI 2200 (LI-COR Biosciences, Lincoln,Nebraska, USA) on the same georeferred points used for destructivesamples.
Daily maximum and minimum temperatures and rainfall wererecorded through a standard agro-meteorological station (Skyeinstruments Ltd., Llandrindod Wells, UK) placed beside the exper-imental field.
2.2. Reflectance measurements and spectral data processing
The radiance over the canopy was measured with an ASD Field-Spec Hand-Held spectroradiometer (Analytical Spectral DevicesInc., Boulder, CO, USA), and converted to spectral reflectance bydividing the radiance reflected by the target by that reflected by astandard white reference plate (spectralon). Measurements weretaken simultaneously with the agronomic survey (i.e., 140 and 147DAS) at the same georeferred points. Spectral reflectance was mea-sured at nadir in a cloud-free day. The height of the optics was, onaverage 80 cm above the canopy. In order to minimize the effects of
Fig. 1. Monthly rainfall (mm), maximum (Tmax) and minimum (Tmin) temperaturesand evapotranspiration (ET), during crop seasons.
the sun’s position, reflectance measurements were taken in about2 h, near solar noon (between 11:00 and 13:00 solar time). A totalof 184 spectral measurements were taken at each date. To fur-ther improve the quality of the spectra, a 10-point Savitzky–Golayfilter (Savitzky and Golay, 1964) was applied to the spectra. TheSavitzky–Golay filter is based on least squares polynomial fittingacross a moving window within the data. Using the above proce-dure, the quality of spectra improved noticeably, without degradingthe shape of the original spectra. All the above spectral data pro-cessing was carried out with help of the SpecLab software (Calabròand Tosca, 2007). After filtering, the following vegetation indiceswere calculated (Table 1): NDVI, GNDVI, SAVI, TSAVI, SAVI2, MSAVI,OSAVI, WDVI, WI. The first derivative of the averaged filtered(Savitzky–Golay) original spectra was also calculated to identifythe position of the red-edge.
2.3. Statistical analysis
Data for each parameter were subjected to analysis of vari-ance. Treatment means were compared using Fisher’s protectedleast significant difference (LSD) test at P < 0.05. LSDs for differ-ent main effect and interaction comparisons were calculated usingthe appropriate standard error terms following Gómez and Gómez(1984). The Statistica (Version 6.1, StatSoft Inc. Tulsa, Oklahoma,USA, 2002) package was used for this purpose. Coefficients of deter-mination (R2) were calculated for relationships among yield andyield component and vegetation indices and LAI.
3. Results
3.1. Meteorological conditions
Decadal air temperature, rainfall and ETo during the studyperiod (January–August) are presented in Fig. 1. The daily aver-age temperature ranged from 6.5 ◦C to 25.1 ◦C, while the minimumand maximum temperatures reached −4.7 ◦C (in March) and 37.4 ◦C(in July), respectively. The total rainfall was about 370 mm, with afairly uniform distribution except for August which accounted forabout 1% of total rainfall of the growing season). The ETo was about830 mm during crop cycle, and consequently the total amount ofwater applied was 3500 m3 ha−1.
3.2. Onion yield
The average onion yield was 8.82 kg m−2 for Cometa and6.57 kg m−2 for Red Mech (Table 3). The minimum production persquare meter was 5.51 kg and 3 kg respectively for Cometa andRed Mech, while the maximum areic productions were 11.2 kg and9.40 kg, respectively. For Cometa hybrid (Fig. 2a) about the 60% ofyield has been included in the range 9–11 kg m−2, the 22% was in the
Author's personal copy
S. Marino et al. / Scientia Horticulturae 155 (2013) 56–64 59
Tab
le
1R
eflec
tan
ce
ind
ex, a
cron
ym, f
orm
ula
tion
and
refe
ren
ces
of
vege
tati
on
ind
ices
obta
ined
from
spec
tral
refl
ecta
nce
mea
sure
men
ts.
Refl
ecta
nce
ind
ex
Acr
onym
Form
ula
tion
Ref
eren
ces
Wat
er
Ind
exW
I
(R90
0/R
970)
Pen
uel
as
et
al. (
1993
, 199
7a,b
)R
ed
edge
�R
EW
avel
engt
h
of
max
imu
m
slop
e
in
the
incr
ease
of
refl
ecta
nce
from
red
to
nea
r
infr
ared
Fill
ella
and
Pen
uel
as
(199
4)W
eigh
ted
dif
fere
nce
vege
tati
on
ind
ex
WD
VI
R80
0
−
(1.2
344
* R67
0)
Cle
vers
(198
9)Pe
rpen
dic
ula
r
vege
tati
on
ind
ex
PVI
1/R
AD
Q
(1.2
3442
+
1
* (R
800
−
1.23
44
* R67
0
−
0.01
83))
Ric
har
dso
n
and
Wie
gan
d
(197
7)So
il-a
dju
sted
vege
tati
on
ind
ex
SAV
I
R80
0
−
R67
0)/(
R80
0
+
R67
0
+
0.5)
) *
(1
+
0.5)
Hu
ete
(198
8)Tr
ansf
orm
ed
soil
-ad
just
ed
vege
tati
on
ind
ex
TSA
VI
(1.2
344
* (R
800
−
1.23
44
* R67
0
−
0.01
83))
/(1.
2344
* R80
0
+
R67
0
−
1.23
44
* 0.0
183)
Bar
et
et
al. (
1989
)Se
con
d
soil
-ad
just
ed
vege
tati
on
ind
ex
SAV
I2
R80
0/(R
670
+
0.01
83/1
.234
4)
Maj
or
et
al. (
1990
)M
odifi
ed
soil
-ad
just
ed
vege
tati
on
ind
ex
MSA
VI
(R80
0
−
R67
0)/(
R80
0
+
R67
0
+
(1
+
(1
−
1.23
42
* R80
0
−
(1.2
344
* R67
0)
* (R
800
−
R67
0)/(
R80
0
+
R67
0)))
)
Qi e
t
al. (
1994
)O
pti
miz
ed
soil
-ad
just
ed
vege
tati
on
ind
exO
SAV
I
(1
+
0.16
) *
(R80
0
−
R67
0)/(
R80
0
+
R67
0
+
0.16
)
Ron
dea
ux
et
al. (
1996
)N
orm
aliz
ed
Dif
fere
nce
Veg
etat
ion
Ind
exN
DV
I
(R80
0
−
R67
0)/(
R80
0
+
R67
0)
Rou
se
et
al. (
1974
)G
reen
Nor
mal
ized
Dif
fere
nce
Veg
etat
ion
Ind
exG
ND
VI
(R80
0
−
R55
0)/(
R80
0
+
R67
0)
Git
elso
n
et
al. (
1996
)
range 7–9 kg m−2 and 16% between 5 kg m−2 and 7 kg m−2. The RedMech hybrid showed a negligible yield in the range 9–11 kg m−2,53% of yield was in the range 5–7 kg m−2, the 30% in the range7–9 kg m−2 (30%) and the 15% in the range 3–5 kg m−2.
3.3. Number of bulbs per square meter
The average areic bulb number (Table 2) of Cometa and RedMech was 65 and 60 respectively. The minimum (39 m−2) and max-imum values (75 m−2) was the same for both varieties. In Cometa,the percentual areic number of bulbs in each size class (Fig. 2b)was as follows: 43% in the 60–70 class, about 32% in the 70–80class, 19% for the 50–60 class and 6% in the 40–50 class. In the RedMech 36% of bulbs was in the classes 50–60 and 60–70; the 19%were represented in the 40–50 class and about 9% in 70–80 class.
3.4. Weight of bulbs
The average bulb weight (Table 2) was 136 g and 110 g respec-tively for Cometa and Red Mech, while the extreme values were100 g and 184 g for Cometa and 64 g and 147 g for Red Mech.
In Cometa (Fig. 2c) 41% of weight bulbs were in the rangebetween 120 g and 140 g, while only a small percentage (3%) is clas-sified in the class 80–100 g and in the class 160–180 g; the 28% ofbulb weight were in 140–160 g class and 25% in the class 100–120 g.
In Red Mech hybrid the 47% of the bulbs showed a weightbetween 100 g and 120 g, 28% in the 80–100 g class; 15% in the class120–140, 8% in the class 60–80 g class and only 2% of bulbs weightwas in the 140–160 g class.
3.5. Size of the bulbs
The average size of the bulbs (Table 2) was 56 mm and 51.4 mmfor the Cometa and Red Mech respectively, while the extreme val-ues were 33 and 79 for Cometa and 37 and 76 for Red Mech.
In Cometa hybrid (Fig. 2d) the 60% of the bulbs showed diam-eter included between 40 mm and 60 mm, 34% of the bulbs wereincluded in 60–80 mm class, about 6% of bulbs had a diameter lowerthan 40 mm. In the Red Mech hybrid 88% of the bulbs belonged to40–60 mm class, the 12% of bulbs shoved a lover or higher size.
3.6. Biomass and plant height
Biomass ranged between 6.3 kg m−2 and 12.8 kg m−2 for Cometaand from 3.4 kg m−2 to 10.7 kg m−2 for Red Mech. The Cometahybrid produced an average of about 10 kg m−2 vs. 7.45 kg m−2 forthe red hybrid. The average plants height was 48.8 cm for Cometaand 41.8 cm for Red Mech. The height ranged, for the Cometa hybrid,from 24 cm to 70 cm, and for the Red hybrid from 17 cm to 60 cm(Table 2).
3.7. Leaf area index
Medium LAI values (Table 3) detected on two different dates(140 DAS and 147 DAS) were 3.01 and 2.99 for Cometa, while theminimum LAIs were 1.85 and 1.25 and the maximum ones were4.43 and 6.76, for the first and second date respectively. In RedMech the mean and minimum values were 2.55 and 1.06 at bothdates, while the maximum LAI was 5.7 and 4.83 for the first andsecond date, respectively.
3.8. Onion yield vs. yield components
Fig. 3 shows onion yield plotted against biomass and yield com-ponents. Yield of both hybrids were highly linearly correlated withbiomass (R2 of Cometa = 0.79 and of Red Mech = 0.87). Different
Author's personal copy
60 S. Marino et al. / Scientia Horticulturae 155 (2013) 56–64
Table 2Mean value of onion yield and yield component. For each mean effect, values in a column followed by different letter are significantly different at P ≤ 0.05, values in a columnfollowed by no letter are not significative at P ≤ 0.05, as determined by the LSD test. One-way analysis of variance (ANOVA) was applied.
Hybrid Yield (kg m−2) No bulb (m−2) Average bulbweight (g)
Table 3Leaf area index and vegetation indices as affected by onion hybrid and date of measurements. For each mean effect, values in a column followed by different letter aresignificantly different at P ≤ 0.05, values in a column followed by no letter are not significative at P ≤ 0.05, as determined by the LSD test. Two-way analysis of variance(ANOVA) was applied.
LAI �RE WI WDVI PVI SAVI TSAVI SAVI2 MSAVI OSAVI NDVI GNDVI
Fig. 2. Relative frequency of onion yield (a), number of bulbs per square meter (b), bulbs weight (c) and bulbs size (d) on Cometa and Red Mech hybrid.
Fig. 3. Linear regression and coefficient of determination (R2) in Cometa and Red Mech among yield and yield components.
Author's personal copy
S. Marino et al. / Scientia Horticulturae 155 (2013) 56–64 61
Fig. 4. Wavelength of Cometa and Red Mech at 140 and 147 days after sowing. Ingray wavelengths relating to the VIs which are set in table.
slope was evident for the linear significant regression between yieldand the number of bulbs (R2 = 0.79 and 0.67 for the white and redhybrid, respectively). Same slope for the weight of bulb, highly cor-related to onion yield (R2 = 0.88 for Cometa and R2 = 0.6 for RedMech). For the size of the bulbs the R2 values were 0.57 and 0.68for Cometa and Red Mech, respectively; the lines were parallel andwith the same slope.
3.9. Crop spectral response
Mean reflectance spectra of the two investigate anion hybridsfrom reflectance measurements carried out at 140 and 147 DASare shown in Fig. 4. Reflectance of both hybrids increased from thefirst date to the second one. The lowest values were shown by RedMech at the first date (140 DAS), at wavelength from 710 nm on;vice versa, Red Mech at 147 DAS showed the highest values at fre-quencies lower than 710 nm. Over 710 nm Cometa at 140 and at147 were well above Red Mech at the first date. The comparison ofthe four reflectance spectra have been performed at specific wave-lengths: 670 nm and 800 nm for biomass VIs, 900 and 970 for plantwater status vegetation index.
In the red region (670 nm), Cometa at 140 and 147 DAS showedvalues by 25% and 19% lower than Red Mech at the first date. In
the near-infra red region (800 nm), Red Mech at 147 DAS increasedits reflectance by 36% respect to the first date; Cometa at 140 and147 DAS increased their reflectance by 48% and 75% respect to RedMech at 140 DAS. Referring to wavelengths used for computing WI(900 nm and 970 nm), Red Mech and Cometa at both dates showedWIs value in the range that indicate a good plant water status.
Red edge of Cometa was computed as 725 nm and 717 nm at140 and at 147 DAS, respectively; Red Mech was positioned at of719 nm and 718 nm, respectively. The red edge values were higherin Cometa with respect to Red Mech at both dates: 0.0095 vs. 0.0055at 140 DAS; 0.013 vs. 0.0085 at 147 DAS.
The vegetation indices (Table 3) have shown significant differ-ences between Cometa and Red Mech. Cometa showed VIs valuehigher respect to Red Mech, ranging by over 17% for GNDVI to morethan 50% for SAVI2; PVI behaved differently, as expected, so as RedMech had higher values than Cometa.
Pooling of the two hybrids, PVI (−2%), WDVI (+19%), MSAVI(+19%) and SAVI (+8%) showed significant differences between thetwo dates, while the other VIs did not differ. The Water Index washigher in Cometa (1.32–1.34, at 140 and 147 DAS) compared to RedMech (1.23 and 1.16, respectively).
Table 4 reports the coefficients of determination of linear regres-sions (VIs vs. LAI, yield and yield components) performed for bothhybrids at the two sampling dates. For the second sampling date forboth hybrids, coefficients of determination were not significant. Atthe first sampling date, several coefficients of determination weresignificant for Red Mech: all VIs were correlated to yield (R2 from0.57 for SAVI2 to 0.67 for SAVI and TSAVI), to biomass (R2 from 0.49for SAVI2 to 0.64 for PVI and SAVI), to the bulb weight (R2 from 0.46for SAVI2 to 0.62 for OSAVI). No appreciable R2 was found for LAI,number and size of bulbs.
4. Discussion
Cometa hybrid showed a significant higher bulb yield than RedMech (+34%), a higher average bulb weight (+25%) and higher num-ber of bulbs per square meter (+9%). Furthermore, Cometa showeda higher bulb size (+10%), an augmented plant biomass (+33%) anda higher plant height (+16%). Significant differences in bulb weight
Table 4Coefficient of determination (R2) of nine VIs and LAI related to yield and yield components of Cometa and Red Mech at different D.A.S.
WDVI PVI SAVI TSAVI SAVI2 MSAVI OSAVI NDVI GNDVI LAI
62 S. Marino et al. / Scientia Horticulturae 155 (2013) 56–64
between the two hybrids came from the shift of the mean towardhigher values in Cometa (94% was in the range 100–160 g), whilethe 90% of the Red Mech bulb weight was between 80 g and 140 g.Cometa showed a peak shift for the number of bulbs: over 90% wasin the range 50–80, while in Red Mech the peak was in the range40–70 (nm−2).
A shift of the mean was noticed even for the bulb size: one thirdof bulbs of Cometa was in the size class 60–80 mm, while over 96%of Red Mech bulbs fall in one size class 40–60 mm. Since the initialplant density was the same for both hybrids, as well as the agro-nomic conditions of the field, the different final areic number ofbulbs at harvest relies on genetic behavior of the two hybrids. Thedifferences measured are anyway well in the range found by Teiet al. (1996), namely as large as 20%.
According to de Visser and van den Berg (1998), total yieldof onions is the result of the plant density and the mean size ofonions (which is linked to the shape and the weight average). Inour study, the following linear regressions were highly significant:yield-biomass, yield-bulb weight, yield-areic bulb number, yield-bulb size. Within each correlation, yield-biomass and yield-bulbweight, the lines of the hybrids were coincident, implying a similarhybrid response to the agronomic and environmental conditions.Vice versa, Cometa showed a lower slope of the regression yield– number of bulbs m−2 respect to Red Mech. This means that theCometa field showed zones characterized by a higher areic bulbdensity associated with a higher bulb size, and in turns with a highyield levels.
Very often a low plant density increases the number of largebulbs, associated with a decline of yield (Frappell, 1973). In the caseof the present study, agronomic practice, climatic condition and thehybrid Cometa had a positive influence on onion yield, bulb weightand bulb size.
At both dates, Cometa showed significant higher LAI (3.1 vs.2.56) and biomass (10.1 kg m−2 and 7.45 kg m−2) respect to RedMech. As matter of fact, we should expect positive correlationsbetween LAI and biomass; however this was not true, likely becauseLAI values were so high to affect radiometric measurements, suchas LAI measurements. For the same reason, LAI was not correlatedto VIS. Bosch-Serra and Casanova (2000) have found in onion thatNDVI (and presumably all indices based on similar wavelengths)saturated at LAI values close 2.0, probably due to changes in leafposture from nearly vertical to a more horizontal position with timeas plant grows bigger, as well as simultaneous overlapping of leafblades with neighboring plants can be observed. As onion plantdevelops, even before onset of bulbing there is a gradual shift fromleaf blade to leaf sheaths production. Gupta et al. (2000) have con-cluded that, during onion crop growth, the relationship betweenLAI and NDVI (a perfect estimator of biomass) was very complex:NDVI and other biomass VIs became more or less constant whereasLAI still continued to increase; at the end both parameters depletednot accordingly. Gupta et al. (2000) thought leaves of the plant,whose architecture is cylindrical, become thicker and more opaqueto light, increasing LAI values detected. According to Gausman andAllen (1973), spectroradiometric characteristics of onion crop arevery special when compared with other 30 plant species: onionshowed the highest leaf thicknesses and was ranked fourth for theleaf water content; furthermore this plant showed the lowest meanvalues of infinite reflectance and the lowest scattering coefficient,which was not correlated anyway to leaf thickness.
All vegetation indices have discriminated between the twovarieties showing higher values for Cometa than Red Mech, PVIexcluded which showed a reverse trend. PVI, WDVI, MSAVI andSAVI indices discriminated between the two dates, raising theirvalues at the second sampling data (PVI shows a reverse trend).
These results are explained by the difference in reflectanceamong the two hybrids and the two dates and consequently by
the differences of values in the wavelengths used for the vegeta-tion indices. The higher value of biomass, plant height in Cometahas been found by VIs.
No relationship was found in Cometa among yield, and yieldcomponents vs. leaf area index, and each VIs at 140 and 147 daysafter showing. This is probably related to the characteristics of themost of the plants that have a high biomass and a high LAI whichinvolves a saturation of the VIs and consequently no relationshipto the yield parameters. The accepted explanation for saturation isthat in dense vegetation, the leaf coverage approaches 100%, whilethe biomass and LAI continue to grow. Typically, crops reach 100%canopy cover around mid-vegetative phases. However, almost allcrops continue to accumulate biomass and LAI until they reach crit-ical phases of growth at which point they begin to senesce. NDVI(and other biomass VIs) saturates at LAI of 2.56–3.1 because theamount of red light that can be absorbed by leaves rapidly reachesa peak. In contrast, NIR scattering by leaves continues to increaseeven as LAI exceeds 3. As a result, once a canopy reaches 100%, NIRreflectance will continue to rise but red reflectance will show onlymodest decreases, resulting in only slight changes in the ratio; thedenominator will have a much greater impact on the ratio thanthe numerator, as stated by Thenkabail et al. (2000). As plants agesNDVI and other biomass VIs becomes more or less constant whereasLAI still continues to increase. Furthermore, leaves of the plant,whose architecture is cylindrical, become thicker and more opaqueto light, which increases LAI (Gupta et al., 2000); on the other hand,NDVI does change a little since leaves are nearly vertical and thusoffer less interaction cross-section (Gupta et al., 2000). Bosch Serraand Casanova (2000) have found that NDVI (and presumably allindices based on similar wavelengths) saturated at LAI values close2.0 probably due to changes in leaf posture from nearly vertical toa more horizontal position with time as plant grow bigger, simul-taneous overlapping of leaf blades with neighboring plants can beobserved. As the onion plant develops, even before onset of bulbing,there is a gradual shift from leaf blade to leaf sheaths production.
Another possible reason of lack of correlations among CometaVIs and yield may derive from several factors, e.g., senescence,water plant stress, disease.
The position of the inflexion point in the red edge region(680–780 nm) of the spectral reflectance signature is affected bybiochemical and biophysical parameters and has been used as ameans to estimate foliar chlorophyll or nitrogen content (Cho andSkidmore, 2006). Shifts in the red edge position to longer or shorterwavelengths has been used as a means to estimate changes in foliarchlorophyll content and also as an indicator of vegetation stress(Chang and Collins, 1983; Clevers et al., 2002; Curran et al., 1995;Horler et al., 1983; Lamb et al., 2002; Smith et al., 2004). Shift towardlonger wavelength and the raise of reflectance of the red edge posi-tion may be related to lower chlorophyll content or to the presenceof a disease (Mohd Shafri et al., 2006; Pu et al., 2003). From the rededge analysis (wavelength and reflectance) there was no evidenceof unfair conditions of Cometa at 140 and 147 DAS.
In the present study, though LAI values of Red Mech at 140 DASwere over 2.0, indices were correlated to yield and yield compo-nents. The VIs with the highest R2 values respect to yield wereSAVI and TSAVI; the bulb weight was positively related with allVIs, SAVI2 excluded (R2 = 0.46). SAVI, PVI, WDVI and MSAVI indiceswere linearly correlated to biomass (R2 higher than 0.6). In conclu-sion, SAVI and PVI were the best indices for this onion type.
5. Conclusion
The Cometa showed better performance for yield and bulb sizecompared to Red Mech, as well as for plant biomass and yield com-ponents. The geo-referenced data have allowed to find a strong
Author's personal copy
S. Marino et al. / Scientia Horticulturae 155 (2013) 56–64 63
correlation among production vs. biomass and yield vs. yield com-ponents in both hybrids, while no correlation was found among LAIand yield and yield parameters in both hybrids. On the contrary,the VIs significantly discriminate between Cometa and Red Mech,highlighting the differences in biomass. In Cometa the VIs werenot related to bulb yield and yield components probably becauseof high LAI values, plant architecture, thickness, light. In Red Mechhybrid, only at 140 DAS, the VIs were significantly related to yield,bulb weight and biomass, such as SAVI and PVI.
For one of the two studied hybrids, the VIs appeared to be apowerful tool for non-destructive determination of onion produc-tion, well before the harvesting period. Information provided byremote/proximal sensing techniques may be helpful for farmers tomanage the onion field crop, for all the dealers who sign a farm-ing contract to check the crop response well before the harvestingperiod, and for stakeholders to foresee the market price trends.
Future studies will be necessary to extend the investigation toother hybrids and to the other crop stages.
Acknowledgements
The authors are gratefully indebted to Dr. Giovanni Cafiero andMaurizio Tosca for their excellent technical assistance. The authorsthank prof. Filippo De Curtis (University of Molise) for many help-ful discussions, derived from field experience on onion cultivation.We are deeply indebted to Mr. Angelo D’Ermes, owner of the fieldwhere the experiment was carried out. This project has receivedfinancial support from CNR–PRIN 2008.
References
Baret, F., Guyot, G., 1991. Potential and limits of vegetation indices for LAI and PARassessment. Remote Sens. Environ. 35, 161–173.
Baret, F., Guyot, G., Major, D.,1989. TSAVI: a vegetation index, which minimizes soilbrightness effects on LAI and APAR estimation. In: Proceedings of 12th CanadianSymposium on Remote Sensing and IGARSS’89, 10–14 July. Vancouver, Canada,pp. 1355–1358.
Beck, P.S.A., Jonsson, P., Høgda, K.A., Karlsen, S.R., Eklundh, L., Skidmore, A.K., 2007.A ground-validated NDVI dataset for monitoring vegetation dynamics and map-ping phenology in Fennoscandia and the Kola peninsula. Int. J. Remote Sens. 28(19), 4311–4330.
Belward, A.S., 1991. Spectral characteristics of vegetation, soil and water in thevisible, near-infrared and middle-infrared wavelengths. In: Bellward, A.S.,Valenzuela, C.R. (Eds.), Remote Sensing and Geographical Information Systemsfor Resources Management in Developing Countries. Kluwer Academic Publish-ers, Dordrecht, pp. 31–53.
Bleasdale, J.K.A., 1959. The yield of onions and red beet as affected by weeds. J. Hortic.Sci. 34, 7–13.
Bosch Serra, A.D., Casanova, D., 2000. Estimation of onion (Allium Cepa L.) biomassand light interception from reflectance measurements at field level. Prod. XXVIHC – Prt 9. Acta Hortic. 519 519, 53–69.
Brewster, J.L., 1990. Physiology of crop growth and bulbing. In: Rabinowith, H.D.,Brewster, J.L. (Eds.), Onions and allied crops: botany, physiology and genetics, 1.CRC Press, Boca Raton, pp. 53–88.
Calabrò G., Tosca M., 2007. SpecPro software per il processing dei dati spettrra-diometrici release alfa 2.95. predisposto per le attività dei laboratori di spettroradiometria. CD-Rom.
Cammarano, D., Fritzgerald, G., Basso, B., O’leary, G., Grace, P.R., Fiorentino, C., 2011.Use of the canopy chlorophyl content index (CCCI) for remote estimation ofwheat nitrogen content in rainfed environments. Agron. J. 103, 1597–1603.
Carter, G.A., 1991. Primary and secondary effects of water content of the spectralreflectance of leaves. Am. J. Bot. 78, 16–924.
Chang, S.H., Collins, W., 1983. Confirmation of the airborne biogeophysical mineralexploration technique using laboratory methods. Econ. Geol. 78, 723–736.
Cho, M.A., Skidmore, A.K., 2006. A new technique for extracting the red edge posi-tion from hyperspectral data: the linear extrapolation method. Remote Sens.Environ. 101, 181–193.
Clay, D.A., Kim, K., Chang, J., Clay, S.A., Dalsted, K., 2006. Characterizing water andnitrogen stress in corn using remote sensing. Agron. J. 98, 579–587.
Clevers, J.G.P.W., 1989. The application of a weighted infrared-red vegetation indexfor estimating leaf area index by correcting soil moisture. Remote Sens. Environ.29, 25–37.
Clevers, J.G.P.W., De Jong, S.M., Epema, G.F., Van der Meer, F.D., Bakker, W.H., Skid-more, A.K., Scholte, K.H., 2002. Derivation of the red edge index using the MERISstandard band setting. Int. J. Remote Sens. 23 (16), 3169–3184.
Coolong, T.W., Randle, W.M., 2003. Ammonium nitrate fertility levels influenceflavour development in hydroponically grown ‘Granex 33′ onion. J. Sci. FoodAgric. 83, 477–482.
Curcio, J.A., Petty, C.C., 1951. The near infrared absorption spectrum of liquid water.J. Opt. Soc. Am. 47, 302–304.
Curran, P.J., Windham, W.R., Gholz, H.L., 1995. Exploring the relationship betweenreflectance red edge and chlorophyll concentration in slash pine leaves. TreePhysiol. 15, 203–206.
Danson, F.M., Steven, M.D., Malthus, T.J., Clarck, J.A., 1992. High-spectral resolutiondata for determining leaf water content. Int. J. Remote Sens. 13, 461–470.
de Visser, C.L.M., van den Berg, W., 1998. A method to calculate the size distributionof onions and its use in an onion growth model. Sci. Hortic. 77, 129–143.
Dubuis, P., Mauch, F., 2003. Sulphur Nutrition Deficiency and Plant Resistance toPathogens. Progress in Plant Sulfur Research 1997–2003. COST Action 829, FAL,Braunschweig, Germany.Faostat web site http://www.faostat.fao.org/site/567/DesktopDefault.aspx?PageID=567#ancor
Fillella, I., Penuelas, J., 1994. The red edge position and shape as indicators of plantchlorophyll content, biomass and hydric status. Int. J. Remote Sens. 15 (7),1459–1470.
Frappell, B.D., 1973. Plant spacing of onions. J. Hort. Sci. 48, 19–28.Galvão, L.S., Roberts, D.A., Formaggio, A.R., Numata, I., Breunig, F.M., 2009. View
angle effects on the discrimination of soybean varieties and on the relationshipsbetween vegetation indices and yield using off-nadir Hyperion data. RemoteSens. Environ. 113, 846–856.
Gausman, H.W., Allen, W.A., 1973. Optical parameters of leaves of 30 plant species.Plant Physiol. 52 (1), 57–62.
Gibbons, P., Freudenberger, D., 2006. An overview of methods used to assess vege-tation condition at the scale of the site. Ecol. Manage. Restor. 7 (S1), 10–17.
Gitelson, A.A., Kaufman, Y.J., Merzlyak, M.N., 1996. Use of a green channel inremote sensing of global vegetation from EOS-MODIS. Remote Sens. Environ.58, 289–298.
Glaze, N.C., 1987. Cultural and mechanical manipulation of Cyperus spp. Weed Tech-nol. 1, 61–65.
Gómez, K.A., Gómez, A.A., 1984. Statistical Procedures for Agricultural Research.Wiley, New York, 680 pp.
Gupta, R.K., Prasad, T.S., Vliayan, D., 2000. Relationship between LAI andNDVI for IRS LISS and LANDSAT TM bands, 2000. Adv. Space Res. 26 (7),1047–1050.
Guyot, G., 1989. Les Signatures Spectrales des Surfaces Naturelles. Paradigme,Caen178–179.
Haack, D., Jampoler, S., 1995. Colour composite comparisons for agricultural assess-ments’. Int. J. Remote Sens. 16 (9), 1589–1598.
Haneklaus, S., Bloem, E., Schnug, E., 2001. Sulphur Induced Resistance (Sir): AnInnovative Concept For Improving Crop Quality in Sustainable Agricultural Pro-duction Systems. International Symposium on Elemental Sulfur for AgronomicApplications and Desert Greening, Abu Dhabi, United Arab Emirates.
Hardisky, M.A., Klemas, V., Smart, R.M., 1983. The influence of soil salinity growthform and leaf moisture on the spectral radiance of Spartina alterniflora canopies.Photogramm. Eng. Remote Sens. 49, 77–83.
Havey, M.J., 1991. History of releases from the onion breeding program of the UnitedStates, Department of Agriculture. Allium Nwsl. 1, 50–54.
Havey, M.J., 1993. Onion breeding. In: Kalloo, G., Berg, B. (Eds.), Genetic Improvementof Vegetable Crops. Pergamon Press, Oxford, pp. 35–49.
Hay, R.K.M., Walker, A.J., 1989. An Introduction to the Physiology of Crop Yield.Longman Scientific & Technical, UK.
Hartridge-Esh, K.A., Bennet, J.P., 1980. Effects of seed weight, plant density andspacing on yield responses of onions. J. of Hortic. Sci. 55, 247–252.
Horler, D.N.H., Dockray, M., Barber, J., 1983. The red edge of plant leaf reflectance.Int. J. Remote Sens. 4 (2), 273–288.
Huete, A.R., 1988. A soil adjusted vegetation index (SAVI). Remote Sens. Environ. 25,295–309.
Huete, A.R., Jackson, R.D., Post, D.F., 1985. Spectral response of plant canopies withdifferent soil background. Remote Sens. Environ. 17, 37–53.
Ji-Hua, M., Bing-Fang, W., 2008. Study on the crop condition monitoring meth-ods with remote sensing. The Interna-tional Archives of the Photogrammetry.Remote Sens. Spatial Inform. Sci. 37 (B8), 945–948.
Jones, H., Davis, G., 1944. Inbreeding and Heterosis and their Relation to the Devel-opment of New Varieties of Onions. US Dept. Agric. Tech. Bull. 874.
Knipling, E.B., 1970. Physical and physiological basis for the reflectance of visibleand near infrared radiation from vegetation. Remote Sens. Environ. 1, 155–159.
Kumar, S., Imtiyaz, M., Kumar, A., 2007a. Effect of differential soil moisture and nutri-ent regimes on postharvest attributes of onion (Allium cepa L.). Sci. Hortic. 112,121–129.
Kumar, S., Imtiyaz, M., Kumar, A., Singh, R., 2007b. Response of onion (Alliumcepa L.) to different levels of irrigation water. Agric. Water Manage. 89,161–166.
Lamb, D.W., Steyn-Ross, M., Schaare, P., Hanna, M.M., Silvester, W., Steyn-Ross, A.,2002. Estimating leaf nitrogen concentration in ryegrass (Lolium spp.) pasture
Author's personal copy
64 S. Marino et al. / Scientia Horticulturae 155 (2013) 56–64
using the chlorophyll red-edge: theoretical modelling and experimental obser-vations. Int. J. Remote Sens. 23 (18), 3619–3648.
Leone, A.P., Menenti, M., Buondonno, A., Letizia, A., Maffei, C., Sorrentino, G., 2007. Afield experiment on spectrometry of crop response to soil salinity. Agric. WaterManage. 80 (12), 39–48.
Li, A.N., Liang, S.L., Wang, A.S., Qin, J., 2007. Estimating crop yield from multitemporalsatellite data using multivariate regression and neural network. Photogramm.Eng. Remote Sens. 73 (10), 1149–1157.
Liu, J.G., Mason, P., 2009. Essential Image Processing and GIS for Remote Sensing.Wiley-Blackwell460.
Magruder, R., Webster, R., Jones, H., Randall, T., Snyder, G., Brown, H., Hawthorn, L.,Wilson, A., 1941. Descriptions of Types of Principal American Varieties of Onions.US Dept. Agric. Misc. Publ. No. 435, Wash., DC.
Major, D.J., Baret, F., Guyot, G., 1990. A ratio vegetation index adjusted for soil bright-ness. Int. J. Remote Sens. 11, 727–740.
Marotti, M., Piccaglia, R., 2002. Characterization of flavonoids in different cultivarsof onion (Allium cepa L.). J. Food Sci. 67 (3), 1229–1232.
Martin de Santa Olalla, F., Dominguez-padilla, A., Lopez, R., 2004. Production andquality of onion crop (Allium cepa L) cultivated in semi-arid climate. Agric. WaterManage. 68, 77–89.
Mather, P.M., Koch, M., 2011. Computer Processing of Remotely-Sensed Images. AnIntroduction. Wiley-Blackwell, Chichester, p. 434.
McCallum, J., Porter, N., Searle, B., Shaw, M., Bettjeman, B., McManus, M., 2005. Sul-fur and nitrogen fertility affects flavour of field-grown onions. Plant Soil 269,151–158.
McGeary, D.J., 1985. The effects of plant density of the shape, size, uniformity, sol-uble solids content and yield of onions suitable for pickling. J. of Hortic. Sci. 60,83–87.
Menges, R.M., Tamez, S., 1981. Response of onion (Allium cepa) to annual weeds andpost-emergence herbicides. Weed Sci. 29 (1), 7–79.
Mohd Shafri, H.Z., Mohd Salleh, M.A., Ghiyamat, A., 2006. Hyperspectral remote sens-ing of vegetation using red edge position techniques. Am. J. Appl. Sci. 3 (6),1864–1871.
Nguyen, H.T., Byun-Woo, L., 2006. Assessment of rice leaf growth and nitrogen statusby hyperspectral canopy reflectance and partial least square regression. Eur. J.Agr. 24, 349–356.
Pasricha, N.S., Abrol, Y.P., 2003. Food production and plant nutrient sulphur. In:Abrol, Y.P., Ahmad, A. (Eds.), Sulphur in Plants. Kluwer Academic Publishers,London, pp. 29–44.
Penuelas, J., Filella, I., Briel, C., Serrano, L., Savé, R., 1993. The reflectance at the950–970 nm region as an indicator of plant water status. Int. J. Remote Sens.14, 1887–1905.
Penuelas, J., Filella, I., Serrano, L., Savé, R., 1996. Cell wall elasticity and Water Index(R970 nm/R900 nm) in wheat under different nitrogen availabilities. Int. J. ofRemote Sens. 17, 373–382.
Penuelas, J., Pinio l, J., Ogaya, R., Filella, I., 1997a. Estimation of plant water con-centration by the reflectance index WI (R900/R970). Int. J. Remote Sens. 18,2869–2875.
Penuelas, J., Pinol, J., Ogaya, R., Filella, I., 1997b. Estimation of plant water con-centration by the reflectance index WI (R900/R970). Int. J. Remote Sens. 18,2869–2875.
Pu, R., Gong, P., Biging, G.S., Larrieu, M.R., 2003. Extraction of red edge optical param-eters from hyperion data for estimation of forest leaf area index. IEEE Trans.Geosci. Remote Sens 41, 916–921.
Qarallah, B., Shoji, K., Kawamura, T., 2008. Development of a yield sensor for mea-suring individual weights of onion bulbs. Biosyst. Eng. 100, 511–515.
Randle, W., 1992a. Onion germplasm interacts with sulfur fertility for plant sulfurutilization and bulb pungency. Euphytica 59, 151–156.
Randle, W., 1992b. Sulfur nutrition affects nonstructural water-soluble carbohy-drates in onion germplasm. HortScience 27, 52–55.
Randle, W.M., 2000. Increasing nitrogen concentration in hydroponic solu-tions affects onion flavor and bulb quality. J. Am. Soc. Hortic. Sci. 125,254–259.
Randle, W.M., Kopsell, D.E., Kopsell, D.A., 1998. Considerations for implementingpungency field testing and its practical implications. In: Proceedings of the 1998National Onion Research Conference, pp. 171–173.
Rondeaux, G., Steven, M., Baret, F., 1996. Optimization of soil-adjusted vegetationindices. Remote Sens. Environ. 55, 95–107.
Rouse, J.W., Haas Jr., R.H., Schell, J.A., Deering, D.W., 1974. Monitoring vegetationsystems in the great plains with ERTS. In: Proc. ERTS-1 Symp., Greenbelt, MD.10–15 December 1973, vol. 1, third ed. NASA NASA SP-351, Washington, DC, pp.309–317.
Salisbury, F.B., Ross, C., 1969. Plant Physiology. Wadsworth, San Francisco, CAL.Savitzky, A., Golay, M.J.E., 1964. Smoothing and differentiation of data by simplified
least squares procedures. Anal. Chem. 36 (8), 1627–1639.Schotten, C.G.J., Van Rooy, W.W.L., Janssen, L.L.F., 1995. Assessment of the capa-
bilities of multi-temporal ERS-1 SAR data to discriminate between agriculturalcrops. Int. J. Remote Sens. 16 (4), 2619–2637.
Sharma, G.P., Verma, R.C., Pathare, P., 2005. Mathematical modeling of infrared radi-ation thin layer drying of onion slices. J. Food Eng. 71, 282–286.
Shock, C.C., Feibert, E.B.G., Saunders, L.D., 2004. Plant population and nitro-gen fertilization for subsurface drip-irrigated onion. HortScience 39,1722–1727.
Smith, K.L., Steven, M.D., Colls, J.J., 2004. Use of hyperspectral derivative ratios inthe red edge region to identify plant stress responses to gas leak. Remote Sens.Environ. 92, 207–217.
Tei, F., Aikman, D.P., Scaife, A., 1996. Growth of lettuce, onion and red beet 2. GrowthModel. Ann. Bot. 78, 645–652.
Thenkabail, P.S., Smith, R.B., De Pauw, E., 2000. Hyperspectral vegetation indices fordetermining agricultural crop characteristics. Remote Sens. Env. 71, 158–182.
Wessman, C.A., 1990. Evaluation of canopy biochemistry. In: Hobbs, R.J., Mooney,H.A. (Eds.), Remote Sensing of Biosphere Functioning. Springer-Verlag, NewYork, pp. 135–156.
Wessman, C.A., 1994. Estimating canopy biochemistry through imaging spectrome-try. In: Hill, J., Mégier, J. (Eds.), Imaging Spectrometry. A Tool for EnvironmentalObservations. Kluwer Academic Publishers, Dordrecht, pp. 57–69.
Yoo, K.S., Pike, L., Crosby, K., Jones, R., Leskovar, D., 2006. Differences in onion pun-gency due to cultivars, growth environment, and bulb sizes. Sci. Hortic. 110,144–149.
Zhao, D., Reddya, K.R., Kakani, V.G., Reddy, V.R., 2005. Nitrogen deficiency effects onplant growth, leaf photosynthesis and hyperspectral reflectance properties ofsorghum. Eur. J. Agron. 22, 391–403.
![Phane Properties of [2.2]Paracyclophane/Dehydrobenzoannulene Hybrids](https://static.documents.page/doc/80x56/6360b55f4b9aa63a9e006a27/phane-properties-of-22paracyclophanedehydrobenzoannulene-hybrids.jpg)
