OILCROP-SUN Model Relevance for Evaluation of Nitrogen Management of Sunflower Hybrids in Sargodha, Punjab

doi:10.4236/ajps.2013.49212

Paper Menu >>

Journal Menu >>

American Journal of Plant Sciences, 2013, 4, 1731-1735

http://dx.doi.org/10.4236/ajps.2013.49212 Published Online September 2013 (http://www.scirp.org/journal/ajps)

OILCROP-SUN Model Relevance for Evaluation of

Nitrogen Management of Sunflower Hybrids in Sargodha,

Punjab

Ashfaq Ahmad1*, Amjed Ali1,2, Tasneem Khaliq1, Syed Aftab Wajid1, Zafar Iqbal2,

Muhammad Ibrahim3, Hafiz Muhammad Rashad Javeed4, Gerrit Hoogenboom5

1Agro-Climatology Lab, University of Agriculture, Faisalabad, Pakistan; 2University College of Agriculture, University of Sargodha,

Sargodha, Pakistan; 3College of Agriculture, D. G. Khan, Pakistan; 4NFC Institute of Engineering and Technology, Multan, Pakistan;

5Washington State University, Pullman, USA.

Email: *[email protected]

Received April 26th, 2013; revised May 27th, 2013; accepted June 15th, 2013

which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

ABSTRACT

The experiments were conducted to evaluate the performance of crop system (DSSAT) OILCROP-SUN model simu-

lating growth & development and achene yield of sunflower hybrids in response to nitrogen under irrigated conditions

in semi arid environment, Sargodha, Punjab. The model was evaluated with observed data collected in trials which were

conducted during spring season in 2010 and 2011 in Sargodha, Punjab, Pakistan. Split plot design was used in layout of

experiment with three replications. The hybrids (Hysun-33 & S-278) and N levels (0, 75, 150 and 225 kg·ha−1) were

allotted in main and sub plots, respectively. The OILCROP-SUN model showed that the model was able to simulate

growth and yield of sunflower with an average of 10.44 error% between observed and simulated achene yield (AY).

The results of simulation analysis indicated that nitrogen rate of 150 kg·N·ha−1 (N3) produced the highest yield as com-

pared to other treatments. Furthermore, the economic analysis through mean Gini Dominance also showed the domi-

nance of this treatment compared to other treatment combinations. Thus management strategy consisting of treatment

150 kg·N·ha−1 was the best for high yield of sunflower hybrids.

Keywords: Decision Support System for Agro-Technology Transfer; Nitrogen; Achene Yield; Crop Modeling

1. Introduction

Oil seed sector, because of ever rising use of edible oil,

has attained significant magnitude in the cost-cutting

measure of Pakistan. Pakistan is a net importer of edible

oil and is spending millions of dollars on its import every

year. Sunflower crop, because of having high oil and pro-

tein contents, has the potential to overpass this gap that

exists between the domestic demands and supply [1].

Sunflower production is very low, and the possible rea-

son is the non-adoption of newly developed hybrids with

higher nutrition requirements. The growers are applying

less nitrogen per hectare, hence, the sowing of hybrids of

high yield potential with optimum nitrogen dose is con-

sidered as a hopeful approach to increase edible oil pro-

duction as well as to reduce the import bills [2]. Choices

of hybrid play a great role in increasing sunflower pro-

duction. [3] worked on performance of various sunflower

hybrids and found a significant difference in yield and

yield components of various hybrids. He suggested that

hybrids should be selected according to agroclimatic con-

ditions of a particular region to obtain higher return. [4]

concluded in his experiments that cultivar adaptation is

imperative for regions in term of obtaining highest yield.

[5] worked on various sunflower hybrids and concluded

that hybrids differ regarding to yield potential. Yield pa-

rameters were increased by nitrogen supply, whereas,

harvest index and seed oil percentage were diminished

by the N application [6]. Evaluations of a crop simulation

model ascertained confidence in its competence to fore-

cast outcomes veteran in the real world. Several simula-

tion models are being used for the sunflower [7,8]. Crop

growth simulations models are based on scientific know-

ledge which serves as a quantitative tool for evaluation of

agronomic factors effects on yield. Crop simulation mo-

*Corresponding author.

OILCROP-SUN Model Relevance for Evaluation of Nitrogen Management of Sunflower Hybrids in Sargodha, Punjab

1732

dels greatly facilitate optimization of crop and its man-

agement strategies [9]. [10] implemented observed re-

sults obtained from his experiments conducted into DS-

SAT (CSM-CERES-Rice model) to appraise the impact

of plant population and nitrogen levels on leaf area index

and total dry matter as well as yield and yield parameters.

The evaluation of DSSAT (CSM-CERES-Rice) showed

that the model was able to simulate growth and yield of

rice grown in semi arid environment, with an average

error of 11% between predicted and observed grain yield.

This approach was described as a useful way to optimize

the crop management for higher production per unit area

as well as monetary return. Plants growth modeling and

its applications have been investigated by a large number

of researchers during the last decade [11,12].

The objectives of this study, therefore, were to evalu-

ate the performance of OILCROP-SUN model for nitro-

gen management under irrigated conditions in semi arid

environment and to determine the best management op-

tion to increase sunflower productivity for local environ-

ment of Punjab.

2. Materials and Methods

The experiment was carried out at the Research Area of

University College of Agriculture Sargodha (32˚05''N,

72˚67''E), Pakistan, under irrigated semi arid conditions

during the spring seasons of 2010 and 2011. The experi-

ment was laid out in a Split plot arrangement under

RCBD having three replications, keeping net plot size

4.20 × 8 m. Sunflower hybrids (Hysun-33, S-278) were

kept in main plots and N levels (0, 75, 150 and 225

kg·ha−1) in sub plots. The crop was sown by dibbler me-

thod on 70 cm spaced ridges and at 22.5 cm plant spac-

ing using a recommended seed rate of 7 kg·ha−1. Phos-

phorus and potash was applied at the rate of 100 - 50

kg·ha−1, respectively. Nitrogen, P and K were given in

the form of urea, single super phosphate and sulfate of

potash, respectively. Full dose of P and K and 1/3 of N

was applied at the time of sowing and remaining 2/3 of N

was applied in two equal splits, at first irrigation and R3

stage (immature bud elongates). All other cultural prac-

tices such as weeding, water application and plant pro-

tection measures were kept normal for the crop.

2.1. Plant Sampling and Measurements

Phenology, as well as growth and development were re-

corded during both the vegetative and reproductive phases

in both years. Five plants were selected at random and

tagged in each plot; anthesis (flowering), and physio-

logical maturity dates were noted. First growth sampling

was conducted after 20 days of sowing, then each sam-

pling every 10 days interval. The leaf area was measured

from 10 g fresh leaves from harvested material from each

fifteen days interval. An area meter (JVC Model TK-

S310EG) was used for the measurement of leaf area and

dry weights, LAI and TDM (gm−2) were recorded at each

harvest as explained by [13]. At final harvest, three rows

with a length of 8 m for each plot were harvested. All the

head were threshed mechanically to determine achene

yield of entire plot and converted into kg·ha−1 and final

yield was corrected to 0% moisture. All weather data was

obtained from measurements made at the nearest mete-

orological observatories around the experimental site.

Weather station provided daily maximum and minimum

air temperature (˚C) i.e. mean temperature, total rainfall

(mm) and mean relative humidity (Table 1).

2.2. Calibration and Evaluation of

OILCROP-SUN

Data obtained from experiments conducted during the

years, 2010 and 2011 was used as input file for calibra-

tion and evaluation of the crop-model. The model simu-

lation was performed under optimum growth conditions.

The comparison of model simulated outcome with ob-

served data assesses accuracy of the model [14]. Mete-

orological data of the location, soil as well as plant char-

acteristics and crop management practices data was ob-

tained from each site and used as input data for the model

[15], Genetic coefficients of hybrids sown was calculated

by decision support system for agro-technology transfer

(DSSAT V 4.5), by using observed data of two years ex-

perimentation [16]. The experimental files that were used

as inputs files includes, weather data file for the experi-

mental period (WeatherMan), soil data of respective ex-

periment (SBuild), crop management data file (XBuild)

and crop cultivar coefficients file [17]. As a part of cali-

bration and evaluation process the simulated data for dif-

ferent phonological developmental stages (anthesis and

maturity date), AY, and TDM were compared with the

observed values.

Table 1. Mean monthly weather data for sunflower growing

season March-June in 2010 and 2011.

Mean

temperature

Total

rainfall

Mean relative

humidity

Months

2010 2011 2010 2011 20102011

 ˚C  mm  % 

March 22.5 21.3 9.2 7.11 58.5 59.4

April 30.2 25.6 4.06 35.06 44.2 46.3

May 32.7 33.8 2.04 8.89 44.9 38.5

June 33.5 33.7 14.74 126.25 44.6 51.0

OILCROP-SUN Model Relevance for Evaluation of Nitrogen Management of Sunflower Hybrids in Sargodha, Punjab

1733

2.3. Statistical Indices

Simulation performance was evaluated by calculating

different statistic indices like root mean square error

(RMSE), mean percentage difference (MPD), error% and

index of agreement [18] with the help of following equa-

tions:



0.5

RMSEp on













nii

MPD100 n



















 

Error %100































where Pi and Oi are predicted and observed values

respectively, O is the observed mean value. The Index of

Agreement (d) as described by [19] that if the d-statistic

value is closer to one, then there is good agreement be-

tween the two variables that are being compared and vice

versa.

3. Results and Discussion

3.1. Model Calibration

The OILCROP-SUN model was calibrated with experi-

mental data collected during 2010 sunflower crop season.

The cultivar coefficients of Hysun-33 and S-278 were

estimated through trial and error and comparison of

simulated and observed data. The final values for the two

cultivar coefficients that determine vegetative and re-

productive growth and development are presented in Ta-

ble 2. A close agreement was obtained between simu-

lated and observed values for sunflower phenology. The

model predicted the dates for days to anthesis with a dif-

ference of one and 2 days between observed and simu-

lated dates for Hysun-33 and S-278 hybrids, respectively.

Similarly, the model predicted the dates for days to

physiological maturity with a difference of 2 and 1 day

between observed and simulated dates for Hysun-33 and

S-278 hybrids, respectively. The simulated and observed

values were in good agreement for Leaf area index and

above ground biomass at different phonological stages.

The lower values for RMSE and higher d-values close to

one reflected that model predicted LAI and above ground

biomass quite well. The d statistics values were (0.94,

0.96) and (0.96, 0.95) along with RMSE values of and

(0.67, 0.47) & (1040, 1132 kg·ha−1) for LAI and TDM

for Hysun-33 and S-278 hybrids, respectively.

3.2. Model Evaluation

The OILCROP-SUN model was calibrated with experi-

mental data collected during 2011 sunflower crop season.

The model predicted the dates for anthesis with RMSE

values from 2.60 and 4.69 days for sunflower hybrids

Hysun-33 and S-278, respectively with average RMSE

value of 3.64 days. Similarly, Mean Percentage Differ-

ence (MPD) values were 3.12 and 6.37 for sunflower

hybrids Hysun-33 and S-278, respectively with average

MPD value of 4.74 (Table 3). The model predicted the

dates for physiological maturity with RMSE values from

5.17 and 4.18 days for sunflower hybrids Hysun-33 and

S-278, respectively with average RMSE value of 4.67

days. Similarly, Mean Percentage Difference (MPD)

values were 3.88 and 3.63 for sunflower hybrids Hy-

sun-33 and S-278, respectively with average MPD value

of 3.75 (Table 4). The simulated and observed values for

LAI and TDM at different phonological stages for dif-

ferent nitrogen levels were in a good agreement. The

value for the d-value for LAI ranged from 0.87 and 0.97,

while the RMSE ranged from 0.42 to 0.53. The d-value

for above-ground biomass ranged from 0.85 to 0.99

while the RMSE ranged from 867 to 1043 kg·ha−1. The

lower values for RMSE and higher d-values close to one

revealed that model predicted LAI and TDM quite well.

However, the RMSE values for achene yield at final

harvest were 347.49 to 346.43 kg·ha−1 for Hysun-33 and

S-278 hybrids, respectively with average RMSE value of

346.96 kg·ha−1. Similarly, Mean Percentage Difference

(MPD) values were 10.01 and 10.88 for sunflower hy-

brids Hysun-33 and S-278, respectively with average

MPD value of 10.44 (Table 5). In general, the results for

model evaluation with the observed data sets indicated

the OILCROP-SUN model was able to simulate yield

Table 2. Cultivar coefficients used with OILCROP-SUN Model for sunflower hybrids.

P1 P

2 P

5 G

2 G

3 O

Genotype

(˚C days) (days) (˚C days) (Nr) (mg·day−1) (%)

Hysun-33 280 2.55 560 746 2.43 53

S-278 255 3.25 545 915 3.22 65

OILCROP-SUN Model Relevance for Evaluation of Nitrogen Management of Sunflower Hybrids in Sargodha, Punjab

1734

Table 3. Comparison of simulated and observed days to anthesis at different planting densities and nitrogen rates during year,

2011.

N rates Hysun-33 S-278 Average

(kg·ha−1) Sim Obs. Error (%) Sim Obs

Error

(%) Sim Obs Error (%)

0 75 71 5.63 68 60 13.33 72 66 9.48

75 75 72 4.17 68 64 6.25 72 68 5.21

150 75 74 1.35 68 66 3.03 72 70 2.19

225 75 76 −1.32 68 70 −2.86 72 73 −2.09

RMSE 2.60 4.69 3.64

MPD 3.12 6.37 4.74

Table 4. Comparison of simulated and observed physiological maturity at different planting densities and nitrogen rates

during year, 2011.

N rates Hysun-33 S-278 Average

(kg·ha−1) Sim Obs. Error (%) Sim Obs Error (%) Sim Obs Error (%)

0 116 107 8.41 101 94 7.45 109 101 7.95

75 116 112 3.57 101 97 4.12 109 105 3.85

150 116 113 2.65 101 100 1.00 109 107 1.83

225 116 115 0.87 101 103 −1.94 109 109 −0.54

RMSE 5.17 4.18 4.67

MPD 3.88 3.63 3.75

Table 5. Comparison of simulated and observed achene yield (AY) kg·ha−1 at different planting densities and nitrogen rates

during year, 2011.

N rates Hysun-33 S-278 Average

(kg·ha−1) Sim Obs. Error (%) Sim Obs Error (%) Sim Obs Error (%)

0 2730 2120 24.06 2851 2280 25.04 2791 2200 24.55

75 3192 3023 5.59 3258 2979 9.37 3225 3001 7.48

150 3486 3404 2.41 3870 3790 2.11 3678 3597 2.26

225 3720 3445 7.98 4078 3811 7.01 3899 3628 7.50

RMSE 347.49 346.43 346.96

MPD 10.01 10.88 10.44

accurately for sunflower hybrids for treatment of nitro-

gen rates under irrigated conditions for a semi arid envi-

ronment in Sargodha, Pakistan.

4. Conclusion

In Model application, the results for model calibration

and evaluation showed that model simulated values were

close to observed values for phenology, and the growth

and yield of sunflower. This study also showed the OIL-

CROP-SUN model served as a tool for determining the

best nitrogen levels for growing sunflower under irri-

gated conditions in semi-arid environment in Pakistan.

This study illustrates the potential for using crop simula-

tions models as information technology for determining

suitable management strategies for sunflower production

in Sargodha, Punjab, Pakistan. Therefore, we can con-

clude that the OILCROP-SUN model could potentially

assist resource-poor farmers in Pakistan and provide

them with alternate management options. However, we

OILCROP-SUN Model Relevance for Evaluation of Nitrogen Management of Sunflower Hybrids in Sargodha, Punjab 1735

suggest that in order to be able to identify the optimum

management practices for a specific region and a specific

crop, a few years of actual field experiments should be

conducted for model evaluation.

REFERENCES

[1] H. Badar, M. S. Javed, A. Ali and Z. Batool, “Production

and Marketing Constraints Limiting Sunflower Produc-

tion in Punjab (Pakistan),” International Journal of Agri-

culture and Biology, Vol. 4, No. 2, 2002, pp. 267-271.

[2] M. A. Malik, M. F. Saleem, M. Sana and A. Rehman,

“Suitable Level of N, P and K for Harvesting the Maxi-

mum Returns of Sunflower,” International Journal of Ag-

riculture and Biology, Vol. 6, No. 2, 2004, pp. 240-242.

[3] J. Bakht, S. Ahmad, M. Tariq, H. Akbar and M. Shafi,

“Performance of Various Hybrids of Sunflower in Pe-

shawar Valley,” Journal of Agricultural Science, Vol. 3,

2006, pp. 25-29.

[4] D. Karaaslan, A. Hatipoglu, Z. Turk and Y. Kaya, “De-

termination of Potential Sunflower Cultivars for the Irri-

gated Conditions of Diyarbakir,” Helia, Vol. 33 No. 52,

2010, pp. 145-152. doi:10.2298/HEL1052145K

[5] W. Nasim, A. Ahmad, A. Bano, R, Olatinwo, M. Usman,

T. Khaliq. A. Wajid, H. M. Hammad, M. Mubeen and M.

Hussain, “Effect of Nitrogen on Yield and Quality of

Sunflower (Helianthus annuus L.) Hybrids under Sub

Humid Conditions of Pakistan,” American Journal of

Plant Sciences, Vol. 3, 2012, pp. 243-251.

[6] T. Khaliq, A. Ahmad, A. hussain and M. A. Ali, “Impact

of Nitrogen Rate on Growth, Yield and Radition Use Ef-

ficiency of Maize under Varying Environments,” Paki-

stan Journal of Agricultural Science, Vol. 45, No. 3, 2008,

pp. 1-7.

[7] S. C. Chapman, G. L. Hammer and H. Meinke, “A Sun-

flower Simulation Model: 1. Model Development,” Agro-

nomy Journal, Vol. 85, No. 3, 1993, pp. 725-735.

doi:10.2134/agronj1993.00021962008500030038x

[8] T. Horie, “Simulation of Sunflower Growth, Formulation

and Parameterization of Dry Matter Production, Leaf

Photosynthesis, Respiration and Partitioning of Photo-

synthates,” Bulletin of the National Institute of Agricul-

tural Sciences, Vol. 24, 1997, pp. 47-70.

[9] IBSRAM, “Soil Water and Nutrient Management Re-

search. A New Agenda (CGIAR Document MT/94/07),”

International Board for Soil Research and Management,

Thailand, 1994.

[10] S. Ahmad, A. Ahmad, C. Manuela, T. Soler, H. Ali, M.

Zia-ul-Haq, J. Anothai, A. Hussain, G. Hoogenboom and

M. Hasanuzzaman, “Application of the CSM-CERES-

Rice Model for Evaluation of Plant Density and Nitrogen

Management of Fine Transplanted Rice for an Irrigated

Semi Arid Environment,” Precision Agriculture, Vol. 13,

No. 2, 2012, pp. 200-218.

doi:10.1007/s11119-011-9238-1

[11] T. Fourcaud, X. Zhang, A. Stokes, H. Lambers and C.

Korner, “Plant Growth Modeling and Applications: The

Increasing Importance of Plant Architecture in Growth

Models,” Annals of Botany, Vol. 101, 2008, pp. 1053-

1063. doi:10.1093/aob/mcn050

[12] B. Li, M. Jaeger and Y. Guo, “Plant Growth Modeling

Simulation, Visualization and Applications,” Proceedings

of PMA09, IEEE Computer Society, Los Alamitos, 2008,

pp. 78-86.

[13] R. Hunt, D. R. Causton, B. Shipley and A. P. Askew, “A

Modern Tool for Classical Plant Growth Analysis,” An-

nals of Botany, Vol. 90, No. 4, 2002, pp. 485-488.

doi:10.1093/aob/mcf214

[14] G. Hoogenboom, J. W. Jones, P. W. Wilkens, C. H. Por-

ter, W. D. Batchelor, L. A. Hunt, K. J. Boote, U. Singh, O.

Uryasev, W. T. Boown, A. Gijsman, A. du Toit, J. W.

White and G. Y. Tsuji, “Decision Support System for

Agrotechnology Transfer,” V. 4.0 [CD-ROM], University

of Hawaii, Honolulu, 2004.

[15] G. A. Pereyra-Irujo and L. A. N. Aguirrezabal, “Sun-

flower Yield and Oil Quality Interactions and Variability:

Analysis through a Simple Simulation Model,” Agricul-

tural and Forest Meteorology, Vol. 143, No. 3, 2007, pp.

252-265. doi:10.1016/j.agrformet.2007.01.001

[16] N. K. Tyagi, D. K. Sharma and S. K. Luthra, “Determina-

tion of Evapo-Transpiration and Crop Coefficients of

Rice and Sunflower with Lysimeter,” Agricultural Water

Management, Vol. 45, No. 1, 2008, pp. 41-54.

[17] N. Diodato and G. Bellocchi, “Modelling Solar Radiation

over Complex Terrains Using Monthly Climatological

Data” Agricultural and Forest Meteorology, Vol. 144, No.

1-2, 2007, pp. 111-126.

doi:10.1016/j.agrformet.2007.02.001

[18] D. Wallach and B. Goffinet, “Mean Square of Prediction

in Models for Studying Ecological and Agronomic Sys-

tems,” Biometrics, Vol. 43, 1989, pp. 561-573.

doi:10.2307/2531995

[19] C. J. Willmott, G. S. Akleson, R. E. Davis, J. J. Feddema,

K. M. Klink, D. R. Legates, J. Odonnell and C. M. Rowe,

“Statistics for the Evaluation and Comparison of Mod-

els,” Journal of Geophysical Research, Vol. 90, No. C5,

1985, pp. 8995-9005. doi:10.1029/JC090iC05p08995