Research Article Received: 29 July 2013
Revised: 19 May 2014
Accepted article published: 28 May 2014
Published online in Wiley Online Library: 16 July 2014
(wileyonlinelibrary.com) DOI 10.1002/jsfa.6762
System dynamics approach for modeling of sugar beet yield considering the effects of climatic variables Lia Pervin* and Md Saiful Islam Abstract BACKGROUND: The aim of this study was to develop a system dynamics model for computation of yields and to investigate the dependency of yields on some major climatic parameters, i.e. temperature and rainfall, for Beta vulgaris subsp. (sugar beet crops) under future climate change scenarios. RESULTS: A system dynamics model was developed which takes account of the effects of rainfall and temperature on sugar beet yields under limited irrigation conditions. A relationship was also developed between the seasonal evapotranspiration and seasonal growing degree days for sugar beet crops. The proposed model was set to run for the present time period of 1993–2012 and for the future period 2013–2040 for Lethbridge region (Alberta, Canada). The model provides sugar beet yields on a yearly basis which are comparable to the present field data. It was found that the future average yield will be increased at about 14% with respect to the present average yield. CONCLUSION: The proposed model can help to improve the understanding of soil water conditions and irrigation water requirements of an area under certain climatic conditions and can be used for future prediction of yields for any crops in any region (with the required information to be provided). The developed system dynamics model can be used as a supporting tool for decision making, for improvement of agricultural management practice of any region. © 2014 Society of Chemical Industry Keywords: system dynamics model; crop yield prediction; climatic variables; soil moisture; evapotranspiration; growing degree days
INTRODUCTION
J Sci Food Agric 2015; 95: 515–521
For any plant, water is vital for its growth and yielding. Water stress decreases the water potential of plants and photosynthesis, and reduces its growth.5 Potential crop yield in any geographic area is determined by a number of factors in addition to crop evapotranspiration (ETc ). A linear relationship was established between yield and crop evapotranspiration (ETc ) for most irrigated crops such as corn, potato, sugar beet and wheat.4 A study was undertaken by Zhang and Huang6 to investigate climatic impacts on cereal yields (rice, wheat and maize) by analyzing climate–yield relationships for China. Stewart et al.7 and Doorenbos and Kassam8 predicted relative yield reductions from the product of relative evapotranspiration deficits and a crop-specific yield response factor or yield reduction ratio, whereas Howell9 found that if water supply is limited yield has a greater dependence on effective precipitation. Several studies have been conducted under controlled conditions to quantify the influence of temperature, solar radiation or water input on the growth and yield of sugar beet.10,11 Another approach for quantifying the influence of weather variables on crop growth was to establish long-term field experiments at single sites and to record weather and yield data for many years.12,13
∗
Correspondence to: Lia Pervin, Department of Civil and Environmental Engineering, University of Alberta, Canada T6G 2R3. E-mail: [email protected] Department of Civil and Environmental Engineering, University of Alberta, Canada T6G 2R3
www.soci.org
© 2014 Society of Chemical Industry
515
Growth and yield of an agricultural crop are influenced by precipitation, temperature, soil moisture, texture, nutrient availability, the occurrence of pests and diseases and their interactions. Yield potential of sugar beet crops depends primarily on site and year effects; whereas the influence of agronomic practices is much lower.1 According to Märländer,1 the effect of the site can be characterized by its soil and climate and their interactions, and the year effect reflects the weather conditions during the vegetation period, which directly affects the length of the growing season. Sugar beet is one of the regular crops of southern Alberta and it plays an important role in the agricultural industry of Canada. Under optimal conditions (well-fertilized, well-irrigated, well-drained soils, pest-free stand, and uniform and optimum canopy) sugar beet crop requires about 500 mm water per growing season in southern Alberta and mainly the water is used for their growth and cooling purposes.2 The water requirement or evapotranspiration (ET) for sugar beet depends on variety, growth stage, canopy density, climatic conditions, and irrigation and crop management.2 Typically, sugar beet roots grow to an effective water extraction depth of 100 cm and its effective lifetime starts from 1 May and ends on 30 September. In 2011 about 33 307 acres of land were used for sugar beet plantation in Alberta.3 The current licensed water allocation is 3.45 billion m3 for 600 000 ha of irrigation land in southern Alberta, with mean gross diversion volume 2.0 billion m3 per year; this amount ensures only about 333 mm per year mean water supply for this irrigation zone.4
www.soci.org Previous work, however, has focused on individual weather variables and their influence on the yield at individual sites. Crop water production functions have a high degree of uncertainty and vary greatly with a number of factors, including weather, soil characteristics that affect infiltration and redistribution of water, antecedent soil moisture, uniformity of water applications, irrigation water salinity, crop responses to chemical use, diseases and pests, as well as fixed and variable costs of irrigation and crop prices.14,15 ‘Optimum water use for each farm is thus highly variable and unpredictable from year to year’.14 Impact of weather variables on the growth of sugar beet under controlled conditions or single field experiment data are available from a previous study,16 but these data are of only limited validity for other sites or larger areas. In a previous study a linear correlation between sugar beet yield and evapotranspiration based on current maximum potential crop yield data for southern Alberta was established.4 The main focus of that study was to investigate the crop yield reductions from water stress due to limited water supplies. The present study is an effort to develop a system dynamics model to estimate the sugar beet yields under limited water supply conditions, considering the temperature and rainfall variability with time. The selected study area was the Lethbridge irrigation region located in southern Alberta, Canada, as sugar beet is one of the main agricultural crops of that region. The aim was to understand the effects of rainfall and temperature on sugar beet yields under limited irrigation supply by developing a system dynamics model, which is one of the systems thinking tools. It also offers results analysis of simulation modeling made using Vensim software, PLE version.
MATERIALS AND METHODS Base model The analysis of variability in crop yields has been approached by considering correlations between yield and evapotranspiration. Relative yield reductions from different levels of water stress were predicted by Bennett and Harms.4 In the present study the following base model was used to obtain potential yields for sugar beet crops:4 ( ) [ ( )] 1 − Ya ∕Ym = ky 1 − ETc ∕ETm (1) where Y a is the predicted crop yield (Mg ha−1 ), Y m is maximum potential yield (Mg ha−1 ), ETc is seasonal crop evapotranspiration (mm), ETm is maximum seasonal crop evapotranspiration (mm) and ky is a crop-specific yield response factor (dimensionless). This method assumes water is the yield-limiting factor, where crops are well adapted to the local growing environment and are produced under a high level of agronomic management.7 In the present study, in addition to the water stress effect, temperature effect was also directly taken into account for computing seasonal crop evapotranspiration. Temperature was subsequently converted to cumulative growing degree days (GDD) above 5 ∘ C basic, as the minimum threshold temperature requirement for sugar beet crop is 5 ∘ C:4 GDD =
[( ) ] Tmax + Tmin ∕2 -Tbase
(2)
516
where T max is the daily maximum temperature, T min is the daily minimum temperature and T base is the base temperature, 5 ∘ C. Rainfall and GDD data at Lethbridge were used as input parameters in the model. A sugar beet yield response factor ky of 1.1 was
wileyonlinelibrary.com/jsfa
L Pervin, MS Islam
used.8 Bennett and Harms4 reported that ETm for the Lethbridge irrigated area is 662 mm, based on data from 1983 to 2010, and Y m for sugar beet crops in southern Alberta is 81.5 Mg ha−1 . These two parameters are being used for developing the system dynamics model for sugar beet yields. The system dynamics model Systems thinking may be defined as ‘a discipline for seeing wholes. It is a framework for seeing interrelationships rather than things, for seeing patterns of change rather than static “snapshots.” During the last thirty years, these tools have been applied to understand a wide range of corporate, urban, regional, economic, political, ecological, and even physiological systems.’17 The basis for analysis in system dynamics is a mental model which becomes a causal loop diagram. The diagrams lead to a formal system model, which is constructed according to traditional theories, employing computer simulation. Systems models represented by the system dynamics method consist of three types of variables: stocks, flows and auxiliary variables. From any crop production project the target is to obtain maximum yield within the available resources. Climatic factors play a very important role here, but irrigation water availability is also a key factor. A system dynamics model can find the interrelationship between crop yield and the other variables on which it depends. Soil moisture is the ‘stock’ of an irrigation system, which depends on some auxiliary variables and changes by the ‘flows’; here flows are soil moisture increment and soil moisture depletion; whereas the allowable soil moisture depletion, effective rainfall, maximum available irrigation, and GDD are the ‘auxiliary variables’. For a certain type of crop in any region there are some parameters that may be considered as ‘constants’. Historical ETm , Y m and ky are treated as ‘constants’ in the model, which are site dependent for that particular crop. A system dynamics model is developed based on the assumption that, in the field, sugar beet crops grow under optimal conditions and depend on climatic factors (rainfall and temperature) and irrigation water supply only. Maximum available irrigation water supply is the boundary condition for the required irrigation amount. As there are no reliable irrigation data available for this study area, assumptions were made for input irrigation amount for each year based on gross irrigation water diversion volume in southern Alberta.4 The proposed sugar beet model simulates at a yearly time step. Therefore, the data had to be scaled to re-create a yearly format but for the effective time period only. Since the growing season of sugar beet starts from early May to the end of September, only this period was taken as the effective time in considering the yearly data. Thus the yearly rainfall comprises the rainfall for the period 1 May to 30 September; similarly all other required data for the model were considered for the same time period only. This could be done realistically because sugar beet is a seasonal crop and it uses water, sunshine and other required nutrition over the season only, and yield is obtain at the end of the season; one yield value per year. The required amount of water or evapotranspiration for sugar beet can be obtained as a cumulative water demand for this effective growing season only. Thus the demand and supply, i.e. evapotranspiration and soil moisture gain, in this model are actually based on the seasonal value but used as yearly data by assigning a null value to the other time periods. The model requires the initial soil moisture content of the field at the beginning of the growing season, as preceding precipitation determines initial soil
© 2014 Society of Chemical Industry
J Sci Food Agric 2015; 95: 515–521
System dynamics approach for modeling of sugar beet yield
www.soci.org
Figure 1. Daily water use during different growth stages of irrigated sugar beet in southern Alberta (source: Alberta Agriculture and Rural Development, 2013), showing that sugar beet uses the maximum amount of water at its full growing stage (last part of July).
water status. In the cold Lethbridge region, October to April is generally covered with snow and unsuitable for normal agricultural activities with, on average, 124 days being frost free.18 Thus only one growing season in a year is possible for sugar beet growers to sensibly utilize the full effective growing season for cultivation. Soil moisture up to the root zone depth is continuously used by the crops and, when the depletion reaches the crop wilting point, the field needs to be replenished with water, either by irrigation or rain or a combination of both, so that soil moisture reaches field capacity. The required irrigation amounts vary with soil texture and growth stage; for silty clay loam soil, the water requirement to replenish the soil to its field capacity at 40% allowable depletion are 44 mm and 88 mm (per irrigation event for sugar beet), during early and peak developmental stages respectively.2 Average sugar beet water use ranges from about 0.1 mm d−1 when the crop emerges to nearly 8 mm d−1 when the crop canopy completely
shades the ground and the tap root is enlarging (Figure 1); water demand decreases as the old leaves start to die and temperatures start to cool down in the fall.2 Figure 1 illustrates that sugar beet uses the maximum amount of water at its full growing stage, possibly as late as the last week of July. Generally, yield is high when the crop is capable of using more water, so reasonably we can relate the seasonal crop evapotranspiration to its seasonal yield. Effective sugar beet irrigation scheduling uses soil water levels in the root zone as a measure of starting and stopping irrigation. Adequate soil water is critical for sugar beet during emergence and growth stages of sugar beet.2 A previous study has shown a linear relationship between GDD and yield for sunflower crops.19 In the present study actual evapotranspiration and GDD data were analyzed and a linear relationship was found for sugar beet crops (using data from 1993
Actual evapotranspiration (mm)
650 600
y = 0.3669x - 34.572 R2 = 0.7088
550 500 450 400 350 300
900
1000 1100
1200
1300
1400 1500 1600
1700 1800
1900
Crop season Growing Degree Days
J Sci Food Agric 2015; 95: 515–521
© 2014 Society of Chemical Industry
wileyonlinelibrary.com/jsfa
517
Figure 2. Relationship between seasonal evapotranspiration and seasonal GDD for sugar beet crops. A best-fitted line is drawn from the actual evapotranspiration and GDD values for Lethbridge region.
www.soci.org
L Pervin, MS Islam
GDD average GDD
normalized GDD
GDD coefficient
Ky sugar beet yield
gap
Ym
existing irrigation system capacity
ETc ETm
max irrigation application
rainfall loss
available soil moisture for crop allowable soil moisture depletion
effective rainfall
soil moisture increment
Soil moisture
soil moisture depletion
Figure 3. System dynamics model developed in Vensim PLE. The model diagram shows all the components of the system. 80
Model sugar beet yield
Model sugar beet yield (Mg ha-1)
Yield (Mg ha-1)
80
Actual sugar beet yield
70 60 50 40 30 20 10 19 9 19 3 94 19 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 01 20 0 20 2 0 20 3 0 20 4 0 20 5 06 20 0 20 7 0 20 8 0 20 9 1 20 0 1 20 1 12
0
y = 0.9009x + 2.6331 R2 = 0.4272
70 60 50 40 30 20 20
Year
40
50
60
70
80
Actual sugar beet yield (Mg ha-1)
Figure 4. Computed and actual3 sugar beet yields (1993–2012). The model output and actual values are shown for each year by vertical columns.
to 2012). Figure 2 shows the relationship between sugar beet seasonal evapotranspiration and seasonal GDD. A factor was calculated by the linear regression method from GDD which affects the seasonal crop evapotranspiration and was found to be 1.109. The developed system dynamics model determines the interdependence between sugar beet yield and the input parameters such as seasonal rainfall, GDD, irrigation water supply and the typical initial soil moisture content of that field. The model is capable of estimating the seasonal crop evapotranspiration as well. Depending on seasonal evapotranspiration and crop yield response factor, it can compute the sugar beet yield in each year.
RESULTS AND DISCUSSION
518
In the system dynamics model developed in Vensim PLE, stock changes by rate. Here soil moisture content is the stock, which is changed by the soil moisture increment and soil moisture depletion (rate). Soil moisture increases by effective rainfall or by applying irrigation water and depleted by crop growth. ETm relates the maximum yield; there exists a gap between the maximum
wileyonlinelibrary.com/jsfa
30
Figure 5. Actual (average of Alberta region) and model sugar beet yields (1993–2012) are plotted on the x- and y-axis respectively. The R2 value is shown on the graph.
evapotranspiration and seasonal crop evapotranspiration and this amount of water should be supplied in the field by using the existing irrigation system. The existing irrigation system always has a certain limited supply capacity. This water is used by crops for evapotranspiration, which is directly related to yield, though yield has a dependency on temperature factors as well. Figure 3 shows the system dynamics model developed in Vensim PLE. The developed model was simulated for the time period 1993–2012 and its behavior was examined; it was observed that the model behaves like the expected reference mode and gives comparative results. Figure 4 shows the computed and actual sugar beet yields (average yields for Alberta region) for 1993–2012. Figure 5 illustrates the model performance in terms of R2 value of plotting, where actual and model sugar beet yields (1993–2012) are plotted on the x- and y-axis, respectively; an R2 value of 0.427 was obtained and the root mean square error was 8.0 for the proposed model.
© 2014 Society of Chemical Industry
J Sci Food Agric 2015; 95: 515–521
System dynamics approach for modeling of sugar beet yield 3
700
3
525
2 1
mm
1
350
3
2 3
4
4
6 6
2 1
4 2
2 1
4 5
3
3
3
4 2 1
4 1
1
1 2
4
2 5
175
www.soci.org
5
5
5 6
6 6 5
6 5
Lethbridge
6
0 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 Time (Year) available soil moisture for crop : model2 1 1 1 1 1 1 ETc : model2 2 2 2 2 2 2 2 soil moisture increment : model2 3 3 3 3 3 soil moisture deplition : model2 4 4 4 4 4 4 max irrigation application : model2 5 5 5 5 5 5 effective rainfall : model2 6 6 6 6 6 6
Figure 6. Variation of soil moisture factors with time. Different parameters are labeled with consecutive numbers, as explained at the bottom of the figure. Figure 8. User interface of USGS Regional Climate downloader; Lethbridge area is shown in the figure (source: USGS Regional Climate Downloader 24 ).
available soil moisture for crop GDD coefficient
ETc
normalized GDD ETm
sugar beet yield
Ky Ym Figure 7. Cause tree of sugar beet yield, showing the model parameters that are directly related to yield.
J Sci Food Agric 2015; 95: 515–521
Future prediction of yields Future modeling was undertaken to investigate how yields of this particular crop (sugar beet) in this particular region might respond to a future climate change scenario. The study was limited to investigating one emission scenario, i.e. A2 only using MPI ECHAM5 model data. Seasonal rainfall and GDD data for the future period (2013–2040) were collected from the MPI ECHAM5 model for A2 emission scenario using USGS Regional and Global Climate downloader. ECHAM5 is the fifth-generation atmospheric general circulation model (GCM), developed at the Max Planck Institute for Meteorology, which is a part of a coupled atmosphere–ocean GCM.21 Sensitivities and use of this model for future climate change scenario have been investigated and reported in previous studies.22,23 The USGS Regional and Global Climate downloader provides dynamically downscaled GCM data; MPI EHCAM5 is one of them,24 from which the required daily data for the Lethbridge region were extracted in a format which is compatible to Vensim PLE using MATLAB programming.
© 2014 Society of Chemical Industry
wileyonlinelibrary.com/jsfa
519
It is observed that the model gives reasonable values of sugar beet yields for the Lethbridge area in comparison to the actual average yields in the Alberta region; this quantitative comparison at the regional scale is meaningful and can be made the basis for justification. The average sugar beet yield for Alberta varies from 37.6 to 67.4 Mg ha−1 ,3 whereas the model value is 29.2–66.3 Mg ha−1 for the same period of time. The actual highest average yield was obtained in 2012 and the lowest yield was found in 2002; the yield in 2006 was found to be the second largest one (in the time period 1993–2012).3 The model gives the highest yield in 2006 and the lowest in 2002, which closely matches the actual yields. The values of crop evapotranspiration (ETc ) from the model are also comparable to the previous data from Bennett and Harms 4 . In a previous study it was found that evapotranspiration values varied from 310 to 630 mm, whereas yields varied from 36 to 82 Mg ha−1 in the Lethbridge area for sugar beet crops.4 In the present study the modeled ETc values varied from 275.8 to 549.9 mm and sugar beet yields ranged from 29.2 to 66.3 Mg ha−1 for the same area, which is close agreement to the results of Bennett and Harms.4 A linear relationship between yield and ETc was also established here; yield increases with the increase of seasonal crop evapotranspiration. Figure 6 illustrates the variation of available soil moisture, ETc , irrigation water supply and increase of soil moisture with time (1993–2012). It is observed from Figure 6 that the available soil moisture varies from 252.7 to 494.3 mm and effective rainfall varies from 37.6 to 441.3 mm. The gap variable shows the additional water requirement for the expected maximum evapotranspiration. In 2006 effective rainfall (only for the growing season) was 136.8 mm and GDD value was 1572.6, while the gap was minimum at that time was only 112.1 mm
with a high evapotranspiration value (549.9 mm); these ensure the highest yield in that year. On the other hand, a maximum gap of 386.2 mm was found in 2002; although the rainfall and GDD value were high enough, the low soil moisture content led to a low value of evapotranspiration and thus resulted in the lowest yield in that year. In a system dynamics model, a causes tree or uses tree is a distinctive option which can find the relationship between the related parameters and show the exact cause of any particular result. Figure 7 illustrates the causes tree diagram for sugar beet yield. The reason behind the high yield in 2006 was investigated by the model and, as described earlier, the high soil moisture content was the primary factor for producing a high yield in 2006. Similarly, the causes of lowest yield in 2002 can be explained by examining Figure 6. If we analyze the actual situation in that particular year, we would find a similar explanation. 2002 was declared a drought year, with the largest agricultural loss and Alberta the hardest hit province, as declared by Agriculture and Agri-Food Canada.20
www.soci.org
L Pervin, MS Islam
90 80
Yield (Mg ha-1)
70 60 Predicted sugar beet yields
50 40 30 20 10
20 1 20 3 1 20 4 15 20 1 20 6 1 20 7 1 20 8 1 20 9 2 20 0 2 20 1 22 20 2 20 3 2 20 4 2 20 5 26 20 2 20 7 2 20 8 2 20 9 3 20 0 3 20 1 3 20 2 3 20 3 3 20 4 3 20 5 3 20 6 3 20 7 3 20 8 3 20 9 40
0
Year
Figure 9. Predicted sugar beet yields for 2013–2040. The model outputs (sugar beet yields) are plotted for each year and are shown by vertical columns.
Figure 8 shows the user interface of the USGS Regional Climate downloader, in which the Lethbridge area is indicated, with its geographical coordinates. For future prediction of yields in 2013–2040 rainfall and GDD (5 ∘ C basis) data were used as input parameters to the previously developed system dynamics model. The maximum irrigation water that can be supplied by the existing irrigation system is kept constant for the future period and is taken as an average of current irrigation water supply amount, i.e. 250 mm. Using the future data for rainfall, GDD and irrigation amount while considering all other factors to be constant or the same as previously, the model was run for the future period 2013–2040. The predicted sugar beet yields are shown in Figure 9. It is observed that high yields (greater than 70 Mg ha−1 ) are found for 2014, 2017, 2021, 2032 and 2039, whereas 2015, 2026, 2033, 2035 and 2036 are low-yielding years (around 30 Mg ha−1 ). For the other years yield varies from 40 to 60 Mg ha−1 .The proposed model suggested that the average sugar beet yield will be increased under future climatic conditions for the Lethbridge agricultural region. The 20-year average of recent yields (1993–2012) is 50 Mg ha−1 and the future average yield is 57 Mg ha−1 (2013–2032), which is about 14% higher than that of recent years. In 2011, Alberta produced sugar beet to a value of $41 986 440 (CAD),3 and recently an investment of nearly $600 000 (CAD) was made to help the Alberta Sugar Beet Growers (ASBG) study the use of sugar beet in the production of sustainable alternatives to petrochemicals.25 This might be a turning point in the sugar beet industry in Alberta, which would require an estimation of future yields in this region. The projected yield will provide an idea of how the sugar beet industry might be affected in future years when over-supply or under-supply occurs. Future estimation would help the detailed planning and management of the sugar beet industry in this region.
CONCLUSION
520
A system dynamics model was developed for estimation of sugar beet yields, as well as for investigating the interrelationship between yields and evapotranspiration for the Lethbridge irrigation area. To investigate the response of this particular crop to future climate change scenario, yields were predicted for future periods. The model gives comparable results for sugar beet yields; it took into account the effect of growing degree days in calculating yields, in addition to water stress effects. Initially the model was set to run for the period 1993–2012, from which yearly yields
wileyonlinelibrary.com/jsfa
were obtained and, again, for the future prediction of yields, the model was set to run for the period 2013–2040 using the same crop parameters for the same area; the results indicate a 14% increase in average yield in the future period. Analysis shows that yields are directly proportional to evapotranspiration and 500–550 mm evapotranspiration ensures high values of sugar beet yields, but no general relationship was found that could correlate rainfall and growing degree days values with sugar beet yields; extensive data analysis may facilitate examination of this relationship. Analysis of this proposed model demonstrates which climatic parameter (rainfall, temperature) has a greater influence on yields (see Figure 7) and how this affects plant water availability in that region (see Figure 6). This information can improve the understanding of soil water conditions and irrigation water requirements of that area under particular weather conditions. This model could also be used for predicting yields of any region for any crop when the crop parameters as well as the weather parameters and soil moisture of that region are known. Although sugar beet yields for the Lethbridge irrigation area are well established by the proposed model, it did not consider any agronomic effects on yield; also it has been stated that the adopted base model did not consider any agronomic effects, since the investigation of agronomic effects on yield was beyond the scope of the present study. Moreover, the proposed model needs to be well established by applying data relating to other crops from other irrigation regions as well.
ACKNOWLEDGEMENTS The authors would like to thank Alberta Agriculture and Rural Development for sugar beet irrigation information and climate data, and the Max Planck Institute for Meteorology for future climate data. The authors would like to convey their gratitude to Dr Evan Davies, Department of Civil and Environmental Engineering, University of Alberta, for his valuable suggestions.
REFERENCES 1 Märländer B, Zuckerrüben, Increase in production in sugar beet as a result of the optimization of methods of cultivation and varieties and by breeding progress. Bernhardt-Pätzold, Stadthagen (1991).
© 2014 Society of Chemical Industry
J Sci Food Agric 2015; 95: 515–521
System dynamics approach for modeling of sugar beet yield 2 Alberta Agriculture and Rural Development, Crop Water Use and Requirements. [Online]. Available: http://www1.agric.gov.ab.ca/ $department/deptdocs.nsf/all/agdex12726 [6 April 2013]. 3 Alberta Agriculture and Rural Development, The Economics of Sugar Beet Production in Alberta 2011. [Online]. Available: http:// http://www1.agric.gov.ab.ca/$Department/deptdocs.nsf/all/agdex 12666/$FILE/171_821-5.pdf [8 April 2013]. 4 Bennett DR and Harms TE, Crop yield and water requirements for major irrigated crops in Southern Alberta. Can Water Resour J 36:159–170 (2011). 5 Kirkham MB, Plant responses to water deficits, in Irrigation of Agricultural Crops, Vol. 30, ed. by Stewart BA and Nielsen DR. American Society of Agronomy, Madison, WI, pp. 323–342 (1990). 6 Zhang T and Huang Y, Impacts of climate change and inter-annual variability on cereal crops in China from 1980 to 2008. J Sci Food Agric 92:1643–1652 (2012). 7 Stewart JI, Danielson RE, Hanks RJ, Jackson EB, Hagan RM, Pruitt WO et al., Optimizing crop production through control of water and salinity levels in the soil. Utah Water Lab PRWG 151:191 (1977). 8 Doorenbos J and Kassam AH, Yield response to water. FAO Irrig Drainage 33:193 (1979). 9 Howell TA, Enriching water use efficiency in irrigated agriculture. Agron J 93:281–289 (2001). 10 Terry N, Developmental physiology of sugar beet. I. The influence of light and temperature on growth. J Exp Bot 19:795–811 (1968). 11 Abdollahian-Noghabi M and Froud-Williams RJ, Effect of moisture stress and re-watering on growth and dry matter partitioning in three cultivars of sugar beet. Aspects Appl Biol 52:71–78 (1998). 12 Freckleton RP, Watkinson AR, Webb DJ and Thomas TH, Yield of sugar beet in relation to weather and nutrients. Agric Forest Meteorol 93:39–51 (1999). 13 Stockfisch N, Koch HJ, Märländer B, Influence of weather on dry matter production of sugar beet. Pflanzenbauwiss 6:63–71 (2002).
www.soci.org 14 English M, Deficit irrigation. I. Analytical framework. J Irrig Drain Div 116:399–412 (1990). 15 Howell TA, Relationships between crop production and transpiration, evapotranspiration, and irrigation, in Irrigation of Agricultural Crops, Vol. 30, ed. by Stewart BA and Nielsen DR. American Society of Agronomy. Madison, WI, pp. 391–434 (1990). 16 Kenter C and Hoffmann C, Impact of weather on yield formation of sugar beet in Germany. Adv Sugar Beet Res 5:19–32 (2003). 17 Senge P, The Fifth Discipline: The Art and Practice of the Learning Organization (2nd edn) Century 86, London (2006). 18 Agroclimatic Atlas of Alberta: Agricultural Climate Elements. [Online]. Available: http://www1.agric.gov.ab.ca/$department/deptdocs.nsf/ all/sag6301 [28 January 2014]. 19 Quadiri G, Hassan FU and Malik Ma, Growing degree days and yield relationship in sunflower. Int J Agric Biol 4:564–568 (2007). 20 Agri-Food Canada, Lessons learned from the Canadian drought years 2001 and 2002. [Online]. Available: http://www4.agr.gc.ca/AAF C-AAC/displayafficherdo? id=1326987176314&lang=eng [22 June 2013]. 21 Max Planck Institute for Meteorology. [Online]. Available: http:// www.mpimet.mpg.de/en/ wissenschaft/modelle/echam/echam5. html [29 January 2014]. 22 Roeckner E, Brokopf R, Esch M, Giorgetta M, Hagemann S, Kornblueh L et al., Sensitivity of simulated climate to horizontal and vertical resolution in the ECHAM5 atmosphere model. J Climate 19:3771–3791 (2006). ˝ 23 Muller WA and Roeckner E, ENSO teleconnections in projections of future climate in ECHAM5/MPI-OM. Clim Dynam 31:533–549 (2008). 24 USGS Regional and Global Climate. [Online]. Available: http:// regclim.coas.oregonstate.edu/visualization/rcd/regional-climatedownloader/index.html#project=regcm [29 January 2014]. 25 Alberta Sugar Beet Growers (ASBG). [Online]. Available: http:// www.asbg.ca/news/news-release [6 May 2014].
521
J Sci Food Agric 2015; 95: 515–521
© 2014 Society of Chemical Industry
wileyonlinelibrary.com/jsfa
Revised: 19 May 2014
Accepted article published: 28 May 2014
Published online in Wiley Online Library: 16 July 2014
(wileyonlinelibrary.com) DOI 10.1002/jsfa.6762
System dynamics approach for modeling of sugar beet yield considering the effects of climatic variables Lia Pervin* and Md Saiful Islam Abstract BACKGROUND: The aim of this study was to develop a system dynamics model for computation of yields and to investigate the dependency of yields on some major climatic parameters, i.e. temperature and rainfall, for Beta vulgaris subsp. (sugar beet crops) under future climate change scenarios. RESULTS: A system dynamics model was developed which takes account of the effects of rainfall and temperature on sugar beet yields under limited irrigation conditions. A relationship was also developed between the seasonal evapotranspiration and seasonal growing degree days for sugar beet crops. The proposed model was set to run for the present time period of 1993–2012 and for the future period 2013–2040 for Lethbridge region (Alberta, Canada). The model provides sugar beet yields on a yearly basis which are comparable to the present field data. It was found that the future average yield will be increased at about 14% with respect to the present average yield. CONCLUSION: The proposed model can help to improve the understanding of soil water conditions and irrigation water requirements of an area under certain climatic conditions and can be used for future prediction of yields for any crops in any region (with the required information to be provided). The developed system dynamics model can be used as a supporting tool for decision making, for improvement of agricultural management practice of any region. © 2014 Society of Chemical Industry Keywords: system dynamics model; crop yield prediction; climatic variables; soil moisture; evapotranspiration; growing degree days
INTRODUCTION
J Sci Food Agric 2015; 95: 515–521
For any plant, water is vital for its growth and yielding. Water stress decreases the water potential of plants and photosynthesis, and reduces its growth.5 Potential crop yield in any geographic area is determined by a number of factors in addition to crop evapotranspiration (ETc ). A linear relationship was established between yield and crop evapotranspiration (ETc ) for most irrigated crops such as corn, potato, sugar beet and wheat.4 A study was undertaken by Zhang and Huang6 to investigate climatic impacts on cereal yields (rice, wheat and maize) by analyzing climate–yield relationships for China. Stewart et al.7 and Doorenbos and Kassam8 predicted relative yield reductions from the product of relative evapotranspiration deficits and a crop-specific yield response factor or yield reduction ratio, whereas Howell9 found that if water supply is limited yield has a greater dependence on effective precipitation. Several studies have been conducted under controlled conditions to quantify the influence of temperature, solar radiation or water input on the growth and yield of sugar beet.10,11 Another approach for quantifying the influence of weather variables on crop growth was to establish long-term field experiments at single sites and to record weather and yield data for many years.12,13
∗
Correspondence to: Lia Pervin, Department of Civil and Environmental Engineering, University of Alberta, Canada T6G 2R3. E-mail: [email protected] Department of Civil and Environmental Engineering, University of Alberta, Canada T6G 2R3
www.soci.org
© 2014 Society of Chemical Industry
515
Growth and yield of an agricultural crop are influenced by precipitation, temperature, soil moisture, texture, nutrient availability, the occurrence of pests and diseases and their interactions. Yield potential of sugar beet crops depends primarily on site and year effects; whereas the influence of agronomic practices is much lower.1 According to Märländer,1 the effect of the site can be characterized by its soil and climate and their interactions, and the year effect reflects the weather conditions during the vegetation period, which directly affects the length of the growing season. Sugar beet is one of the regular crops of southern Alberta and it plays an important role in the agricultural industry of Canada. Under optimal conditions (well-fertilized, well-irrigated, well-drained soils, pest-free stand, and uniform and optimum canopy) sugar beet crop requires about 500 mm water per growing season in southern Alberta and mainly the water is used for their growth and cooling purposes.2 The water requirement or evapotranspiration (ET) for sugar beet depends on variety, growth stage, canopy density, climatic conditions, and irrigation and crop management.2 Typically, sugar beet roots grow to an effective water extraction depth of 100 cm and its effective lifetime starts from 1 May and ends on 30 September. In 2011 about 33 307 acres of land were used for sugar beet plantation in Alberta.3 The current licensed water allocation is 3.45 billion m3 for 600 000 ha of irrigation land in southern Alberta, with mean gross diversion volume 2.0 billion m3 per year; this amount ensures only about 333 mm per year mean water supply for this irrigation zone.4
www.soci.org Previous work, however, has focused on individual weather variables and their influence on the yield at individual sites. Crop water production functions have a high degree of uncertainty and vary greatly with a number of factors, including weather, soil characteristics that affect infiltration and redistribution of water, antecedent soil moisture, uniformity of water applications, irrigation water salinity, crop responses to chemical use, diseases and pests, as well as fixed and variable costs of irrigation and crop prices.14,15 ‘Optimum water use for each farm is thus highly variable and unpredictable from year to year’.14 Impact of weather variables on the growth of sugar beet under controlled conditions or single field experiment data are available from a previous study,16 but these data are of only limited validity for other sites or larger areas. In a previous study a linear correlation between sugar beet yield and evapotranspiration based on current maximum potential crop yield data for southern Alberta was established.4 The main focus of that study was to investigate the crop yield reductions from water stress due to limited water supplies. The present study is an effort to develop a system dynamics model to estimate the sugar beet yields under limited water supply conditions, considering the temperature and rainfall variability with time. The selected study area was the Lethbridge irrigation region located in southern Alberta, Canada, as sugar beet is one of the main agricultural crops of that region. The aim was to understand the effects of rainfall and temperature on sugar beet yields under limited irrigation supply by developing a system dynamics model, which is one of the systems thinking tools. It also offers results analysis of simulation modeling made using Vensim software, PLE version.
MATERIALS AND METHODS Base model The analysis of variability in crop yields has been approached by considering correlations between yield and evapotranspiration. Relative yield reductions from different levels of water stress were predicted by Bennett and Harms.4 In the present study the following base model was used to obtain potential yields for sugar beet crops:4 ( ) [ ( )] 1 − Ya ∕Ym = ky 1 − ETc ∕ETm (1) where Y a is the predicted crop yield (Mg ha−1 ), Y m is maximum potential yield (Mg ha−1 ), ETc is seasonal crop evapotranspiration (mm), ETm is maximum seasonal crop evapotranspiration (mm) and ky is a crop-specific yield response factor (dimensionless). This method assumes water is the yield-limiting factor, where crops are well adapted to the local growing environment and are produced under a high level of agronomic management.7 In the present study, in addition to the water stress effect, temperature effect was also directly taken into account for computing seasonal crop evapotranspiration. Temperature was subsequently converted to cumulative growing degree days (GDD) above 5 ∘ C basic, as the minimum threshold temperature requirement for sugar beet crop is 5 ∘ C:4 GDD =
[( ) ] Tmax + Tmin ∕2 -Tbase
(2)
516
where T max is the daily maximum temperature, T min is the daily minimum temperature and T base is the base temperature, 5 ∘ C. Rainfall and GDD data at Lethbridge were used as input parameters in the model. A sugar beet yield response factor ky of 1.1 was
wileyonlinelibrary.com/jsfa
L Pervin, MS Islam
used.8 Bennett and Harms4 reported that ETm for the Lethbridge irrigated area is 662 mm, based on data from 1983 to 2010, and Y m for sugar beet crops in southern Alberta is 81.5 Mg ha−1 . These two parameters are being used for developing the system dynamics model for sugar beet yields. The system dynamics model Systems thinking may be defined as ‘a discipline for seeing wholes. It is a framework for seeing interrelationships rather than things, for seeing patterns of change rather than static “snapshots.” During the last thirty years, these tools have been applied to understand a wide range of corporate, urban, regional, economic, political, ecological, and even physiological systems.’17 The basis for analysis in system dynamics is a mental model which becomes a causal loop diagram. The diagrams lead to a formal system model, which is constructed according to traditional theories, employing computer simulation. Systems models represented by the system dynamics method consist of three types of variables: stocks, flows and auxiliary variables. From any crop production project the target is to obtain maximum yield within the available resources. Climatic factors play a very important role here, but irrigation water availability is also a key factor. A system dynamics model can find the interrelationship between crop yield and the other variables on which it depends. Soil moisture is the ‘stock’ of an irrigation system, which depends on some auxiliary variables and changes by the ‘flows’; here flows are soil moisture increment and soil moisture depletion; whereas the allowable soil moisture depletion, effective rainfall, maximum available irrigation, and GDD are the ‘auxiliary variables’. For a certain type of crop in any region there are some parameters that may be considered as ‘constants’. Historical ETm , Y m and ky are treated as ‘constants’ in the model, which are site dependent for that particular crop. A system dynamics model is developed based on the assumption that, in the field, sugar beet crops grow under optimal conditions and depend on climatic factors (rainfall and temperature) and irrigation water supply only. Maximum available irrigation water supply is the boundary condition for the required irrigation amount. As there are no reliable irrigation data available for this study area, assumptions were made for input irrigation amount for each year based on gross irrigation water diversion volume in southern Alberta.4 The proposed sugar beet model simulates at a yearly time step. Therefore, the data had to be scaled to re-create a yearly format but for the effective time period only. Since the growing season of sugar beet starts from early May to the end of September, only this period was taken as the effective time in considering the yearly data. Thus the yearly rainfall comprises the rainfall for the period 1 May to 30 September; similarly all other required data for the model were considered for the same time period only. This could be done realistically because sugar beet is a seasonal crop and it uses water, sunshine and other required nutrition over the season only, and yield is obtain at the end of the season; one yield value per year. The required amount of water or evapotranspiration for sugar beet can be obtained as a cumulative water demand for this effective growing season only. Thus the demand and supply, i.e. evapotranspiration and soil moisture gain, in this model are actually based on the seasonal value but used as yearly data by assigning a null value to the other time periods. The model requires the initial soil moisture content of the field at the beginning of the growing season, as preceding precipitation determines initial soil
© 2014 Society of Chemical Industry
J Sci Food Agric 2015; 95: 515–521
System dynamics approach for modeling of sugar beet yield
www.soci.org
Figure 1. Daily water use during different growth stages of irrigated sugar beet in southern Alberta (source: Alberta Agriculture and Rural Development, 2013), showing that sugar beet uses the maximum amount of water at its full growing stage (last part of July).
water status. In the cold Lethbridge region, October to April is generally covered with snow and unsuitable for normal agricultural activities with, on average, 124 days being frost free.18 Thus only one growing season in a year is possible for sugar beet growers to sensibly utilize the full effective growing season for cultivation. Soil moisture up to the root zone depth is continuously used by the crops and, when the depletion reaches the crop wilting point, the field needs to be replenished with water, either by irrigation or rain or a combination of both, so that soil moisture reaches field capacity. The required irrigation amounts vary with soil texture and growth stage; for silty clay loam soil, the water requirement to replenish the soil to its field capacity at 40% allowable depletion are 44 mm and 88 mm (per irrigation event for sugar beet), during early and peak developmental stages respectively.2 Average sugar beet water use ranges from about 0.1 mm d−1 when the crop emerges to nearly 8 mm d−1 when the crop canopy completely
shades the ground and the tap root is enlarging (Figure 1); water demand decreases as the old leaves start to die and temperatures start to cool down in the fall.2 Figure 1 illustrates that sugar beet uses the maximum amount of water at its full growing stage, possibly as late as the last week of July. Generally, yield is high when the crop is capable of using more water, so reasonably we can relate the seasonal crop evapotranspiration to its seasonal yield. Effective sugar beet irrigation scheduling uses soil water levels in the root zone as a measure of starting and stopping irrigation. Adequate soil water is critical for sugar beet during emergence and growth stages of sugar beet.2 A previous study has shown a linear relationship between GDD and yield for sunflower crops.19 In the present study actual evapotranspiration and GDD data were analyzed and a linear relationship was found for sugar beet crops (using data from 1993
Actual evapotranspiration (mm)
650 600
y = 0.3669x - 34.572 R2 = 0.7088
550 500 450 400 350 300
900
1000 1100
1200
1300
1400 1500 1600
1700 1800
1900
Crop season Growing Degree Days
J Sci Food Agric 2015; 95: 515–521
© 2014 Society of Chemical Industry
wileyonlinelibrary.com/jsfa
517
Figure 2. Relationship between seasonal evapotranspiration and seasonal GDD for sugar beet crops. A best-fitted line is drawn from the actual evapotranspiration and GDD values for Lethbridge region.
www.soci.org
L Pervin, MS Islam
GDD average GDD
normalized GDD
GDD coefficient
Ky sugar beet yield
gap
Ym
existing irrigation system capacity
ETc ETm
max irrigation application
rainfall loss
available soil moisture for crop allowable soil moisture depletion
effective rainfall
soil moisture increment
Soil moisture
soil moisture depletion
Figure 3. System dynamics model developed in Vensim PLE. The model diagram shows all the components of the system. 80
Model sugar beet yield
Model sugar beet yield (Mg ha-1)
Yield (Mg ha-1)
80
Actual sugar beet yield
70 60 50 40 30 20 10 19 9 19 3 94 19 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 01 20 0 20 2 0 20 3 0 20 4 0 20 5 06 20 0 20 7 0 20 8 0 20 9 1 20 0 1 20 1 12
0
y = 0.9009x + 2.6331 R2 = 0.4272
70 60 50 40 30 20 20
Year
40
50
60
70
80
Actual sugar beet yield (Mg ha-1)
Figure 4. Computed and actual3 sugar beet yields (1993–2012). The model output and actual values are shown for each year by vertical columns.
to 2012). Figure 2 shows the relationship between sugar beet seasonal evapotranspiration and seasonal GDD. A factor was calculated by the linear regression method from GDD which affects the seasonal crop evapotranspiration and was found to be 1.109. The developed system dynamics model determines the interdependence between sugar beet yield and the input parameters such as seasonal rainfall, GDD, irrigation water supply and the typical initial soil moisture content of that field. The model is capable of estimating the seasonal crop evapotranspiration as well. Depending on seasonal evapotranspiration and crop yield response factor, it can compute the sugar beet yield in each year.
RESULTS AND DISCUSSION
518
In the system dynamics model developed in Vensim PLE, stock changes by rate. Here soil moisture content is the stock, which is changed by the soil moisture increment and soil moisture depletion (rate). Soil moisture increases by effective rainfall or by applying irrigation water and depleted by crop growth. ETm relates the maximum yield; there exists a gap between the maximum
wileyonlinelibrary.com/jsfa
30
Figure 5. Actual (average of Alberta region) and model sugar beet yields (1993–2012) are plotted on the x- and y-axis respectively. The R2 value is shown on the graph.
evapotranspiration and seasonal crop evapotranspiration and this amount of water should be supplied in the field by using the existing irrigation system. The existing irrigation system always has a certain limited supply capacity. This water is used by crops for evapotranspiration, which is directly related to yield, though yield has a dependency on temperature factors as well. Figure 3 shows the system dynamics model developed in Vensim PLE. The developed model was simulated for the time period 1993–2012 and its behavior was examined; it was observed that the model behaves like the expected reference mode and gives comparative results. Figure 4 shows the computed and actual sugar beet yields (average yields for Alberta region) for 1993–2012. Figure 5 illustrates the model performance in terms of R2 value of plotting, where actual and model sugar beet yields (1993–2012) are plotted on the x- and y-axis, respectively; an R2 value of 0.427 was obtained and the root mean square error was 8.0 for the proposed model.
© 2014 Society of Chemical Industry
J Sci Food Agric 2015; 95: 515–521
System dynamics approach for modeling of sugar beet yield 3
700
3
525
2 1
mm
1
350
3
2 3
4
4
6 6
2 1
4 2
2 1
4 5
3
3
3
4 2 1
4 1
1
1 2
4
2 5
175
www.soci.org
5
5
5 6
6 6 5
6 5
Lethbridge
6
0 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 Time (Year) available soil moisture for crop : model2 1 1 1 1 1 1 ETc : model2 2 2 2 2 2 2 2 soil moisture increment : model2 3 3 3 3 3 soil moisture deplition : model2 4 4 4 4 4 4 max irrigation application : model2 5 5 5 5 5 5 effective rainfall : model2 6 6 6 6 6 6
Figure 6. Variation of soil moisture factors with time. Different parameters are labeled with consecutive numbers, as explained at the bottom of the figure. Figure 8. User interface of USGS Regional Climate downloader; Lethbridge area is shown in the figure (source: USGS Regional Climate Downloader 24 ).
available soil moisture for crop GDD coefficient
ETc
normalized GDD ETm
sugar beet yield
Ky Ym Figure 7. Cause tree of sugar beet yield, showing the model parameters that are directly related to yield.
J Sci Food Agric 2015; 95: 515–521
Future prediction of yields Future modeling was undertaken to investigate how yields of this particular crop (sugar beet) in this particular region might respond to a future climate change scenario. The study was limited to investigating one emission scenario, i.e. A2 only using MPI ECHAM5 model data. Seasonal rainfall and GDD data for the future period (2013–2040) were collected from the MPI ECHAM5 model for A2 emission scenario using USGS Regional and Global Climate downloader. ECHAM5 is the fifth-generation atmospheric general circulation model (GCM), developed at the Max Planck Institute for Meteorology, which is a part of a coupled atmosphere–ocean GCM.21 Sensitivities and use of this model for future climate change scenario have been investigated and reported in previous studies.22,23 The USGS Regional and Global Climate downloader provides dynamically downscaled GCM data; MPI EHCAM5 is one of them,24 from which the required daily data for the Lethbridge region were extracted in a format which is compatible to Vensim PLE using MATLAB programming.
© 2014 Society of Chemical Industry
wileyonlinelibrary.com/jsfa
519
It is observed that the model gives reasonable values of sugar beet yields for the Lethbridge area in comparison to the actual average yields in the Alberta region; this quantitative comparison at the regional scale is meaningful and can be made the basis for justification. The average sugar beet yield for Alberta varies from 37.6 to 67.4 Mg ha−1 ,3 whereas the model value is 29.2–66.3 Mg ha−1 for the same period of time. The actual highest average yield was obtained in 2012 and the lowest yield was found in 2002; the yield in 2006 was found to be the second largest one (in the time period 1993–2012).3 The model gives the highest yield in 2006 and the lowest in 2002, which closely matches the actual yields. The values of crop evapotranspiration (ETc ) from the model are also comparable to the previous data from Bennett and Harms 4 . In a previous study it was found that evapotranspiration values varied from 310 to 630 mm, whereas yields varied from 36 to 82 Mg ha−1 in the Lethbridge area for sugar beet crops.4 In the present study the modeled ETc values varied from 275.8 to 549.9 mm and sugar beet yields ranged from 29.2 to 66.3 Mg ha−1 for the same area, which is close agreement to the results of Bennett and Harms.4 A linear relationship between yield and ETc was also established here; yield increases with the increase of seasonal crop evapotranspiration. Figure 6 illustrates the variation of available soil moisture, ETc , irrigation water supply and increase of soil moisture with time (1993–2012). It is observed from Figure 6 that the available soil moisture varies from 252.7 to 494.3 mm and effective rainfall varies from 37.6 to 441.3 mm. The gap variable shows the additional water requirement for the expected maximum evapotranspiration. In 2006 effective rainfall (only for the growing season) was 136.8 mm and GDD value was 1572.6, while the gap was minimum at that time was only 112.1 mm
with a high evapotranspiration value (549.9 mm); these ensure the highest yield in that year. On the other hand, a maximum gap of 386.2 mm was found in 2002; although the rainfall and GDD value were high enough, the low soil moisture content led to a low value of evapotranspiration and thus resulted in the lowest yield in that year. In a system dynamics model, a causes tree or uses tree is a distinctive option which can find the relationship between the related parameters and show the exact cause of any particular result. Figure 7 illustrates the causes tree diagram for sugar beet yield. The reason behind the high yield in 2006 was investigated by the model and, as described earlier, the high soil moisture content was the primary factor for producing a high yield in 2006. Similarly, the causes of lowest yield in 2002 can be explained by examining Figure 6. If we analyze the actual situation in that particular year, we would find a similar explanation. 2002 was declared a drought year, with the largest agricultural loss and Alberta the hardest hit province, as declared by Agriculture and Agri-Food Canada.20
www.soci.org
L Pervin, MS Islam
90 80
Yield (Mg ha-1)
70 60 Predicted sugar beet yields
50 40 30 20 10
20 1 20 3 1 20 4 15 20 1 20 6 1 20 7 1 20 8 1 20 9 2 20 0 2 20 1 22 20 2 20 3 2 20 4 2 20 5 26 20 2 20 7 2 20 8 2 20 9 3 20 0 3 20 1 3 20 2 3 20 3 3 20 4 3 20 5 3 20 6 3 20 7 3 20 8 3 20 9 40
0
Year
Figure 9. Predicted sugar beet yields for 2013–2040. The model outputs (sugar beet yields) are plotted for each year and are shown by vertical columns.
Figure 8 shows the user interface of the USGS Regional Climate downloader, in which the Lethbridge area is indicated, with its geographical coordinates. For future prediction of yields in 2013–2040 rainfall and GDD (5 ∘ C basis) data were used as input parameters to the previously developed system dynamics model. The maximum irrigation water that can be supplied by the existing irrigation system is kept constant for the future period and is taken as an average of current irrigation water supply amount, i.e. 250 mm. Using the future data for rainfall, GDD and irrigation amount while considering all other factors to be constant or the same as previously, the model was run for the future period 2013–2040. The predicted sugar beet yields are shown in Figure 9. It is observed that high yields (greater than 70 Mg ha−1 ) are found for 2014, 2017, 2021, 2032 and 2039, whereas 2015, 2026, 2033, 2035 and 2036 are low-yielding years (around 30 Mg ha−1 ). For the other years yield varies from 40 to 60 Mg ha−1 .The proposed model suggested that the average sugar beet yield will be increased under future climatic conditions for the Lethbridge agricultural region. The 20-year average of recent yields (1993–2012) is 50 Mg ha−1 and the future average yield is 57 Mg ha−1 (2013–2032), which is about 14% higher than that of recent years. In 2011, Alberta produced sugar beet to a value of $41 986 440 (CAD),3 and recently an investment of nearly $600 000 (CAD) was made to help the Alberta Sugar Beet Growers (ASBG) study the use of sugar beet in the production of sustainable alternatives to petrochemicals.25 This might be a turning point in the sugar beet industry in Alberta, which would require an estimation of future yields in this region. The projected yield will provide an idea of how the sugar beet industry might be affected in future years when over-supply or under-supply occurs. Future estimation would help the detailed planning and management of the sugar beet industry in this region.
CONCLUSION
520
A system dynamics model was developed for estimation of sugar beet yields, as well as for investigating the interrelationship between yields and evapotranspiration for the Lethbridge irrigation area. To investigate the response of this particular crop to future climate change scenario, yields were predicted for future periods. The model gives comparable results for sugar beet yields; it took into account the effect of growing degree days in calculating yields, in addition to water stress effects. Initially the model was set to run for the period 1993–2012, from which yearly yields
wileyonlinelibrary.com/jsfa
were obtained and, again, for the future prediction of yields, the model was set to run for the period 2013–2040 using the same crop parameters for the same area; the results indicate a 14% increase in average yield in the future period. Analysis shows that yields are directly proportional to evapotranspiration and 500–550 mm evapotranspiration ensures high values of sugar beet yields, but no general relationship was found that could correlate rainfall and growing degree days values with sugar beet yields; extensive data analysis may facilitate examination of this relationship. Analysis of this proposed model demonstrates which climatic parameter (rainfall, temperature) has a greater influence on yields (see Figure 7) and how this affects plant water availability in that region (see Figure 6). This information can improve the understanding of soil water conditions and irrigation water requirements of that area under particular weather conditions. This model could also be used for predicting yields of any region for any crop when the crop parameters as well as the weather parameters and soil moisture of that region are known. Although sugar beet yields for the Lethbridge irrigation area are well established by the proposed model, it did not consider any agronomic effects on yield; also it has been stated that the adopted base model did not consider any agronomic effects, since the investigation of agronomic effects on yield was beyond the scope of the present study. Moreover, the proposed model needs to be well established by applying data relating to other crops from other irrigation regions as well.
ACKNOWLEDGEMENTS The authors would like to thank Alberta Agriculture and Rural Development for sugar beet irrigation information and climate data, and the Max Planck Institute for Meteorology for future climate data. The authors would like to convey their gratitude to Dr Evan Davies, Department of Civil and Environmental Engineering, University of Alberta, for his valuable suggestions.
REFERENCES 1 Märländer B, Zuckerrüben, Increase in production in sugar beet as a result of the optimization of methods of cultivation and varieties and by breeding progress. Bernhardt-Pätzold, Stadthagen (1991).
© 2014 Society of Chemical Industry
J Sci Food Agric 2015; 95: 515–521
System dynamics approach for modeling of sugar beet yield 2 Alberta Agriculture and Rural Development, Crop Water Use and Requirements. [Online]. Available: http://www1.agric.gov.ab.ca/ $department/deptdocs.nsf/all/agdex12726 [6 April 2013]. 3 Alberta Agriculture and Rural Development, The Economics of Sugar Beet Production in Alberta 2011. [Online]. Available: http:// http://www1.agric.gov.ab.ca/$Department/deptdocs.nsf/all/agdex 12666/$FILE/171_821-5.pdf [8 April 2013]. 4 Bennett DR and Harms TE, Crop yield and water requirements for major irrigated crops in Southern Alberta. Can Water Resour J 36:159–170 (2011). 5 Kirkham MB, Plant responses to water deficits, in Irrigation of Agricultural Crops, Vol. 30, ed. by Stewart BA and Nielsen DR. American Society of Agronomy, Madison, WI, pp. 323–342 (1990). 6 Zhang T and Huang Y, Impacts of climate change and inter-annual variability on cereal crops in China from 1980 to 2008. J Sci Food Agric 92:1643–1652 (2012). 7 Stewart JI, Danielson RE, Hanks RJ, Jackson EB, Hagan RM, Pruitt WO et al., Optimizing crop production through control of water and salinity levels in the soil. Utah Water Lab PRWG 151:191 (1977). 8 Doorenbos J and Kassam AH, Yield response to water. FAO Irrig Drainage 33:193 (1979). 9 Howell TA, Enriching water use efficiency in irrigated agriculture. Agron J 93:281–289 (2001). 10 Terry N, Developmental physiology of sugar beet. I. The influence of light and temperature on growth. J Exp Bot 19:795–811 (1968). 11 Abdollahian-Noghabi M and Froud-Williams RJ, Effect of moisture stress and re-watering on growth and dry matter partitioning in three cultivars of sugar beet. Aspects Appl Biol 52:71–78 (1998). 12 Freckleton RP, Watkinson AR, Webb DJ and Thomas TH, Yield of sugar beet in relation to weather and nutrients. Agric Forest Meteorol 93:39–51 (1999). 13 Stockfisch N, Koch HJ, Märländer B, Influence of weather on dry matter production of sugar beet. Pflanzenbauwiss 6:63–71 (2002).
www.soci.org 14 English M, Deficit irrigation. I. Analytical framework. J Irrig Drain Div 116:399–412 (1990). 15 Howell TA, Relationships between crop production and transpiration, evapotranspiration, and irrigation, in Irrigation of Agricultural Crops, Vol. 30, ed. by Stewart BA and Nielsen DR. American Society of Agronomy. Madison, WI, pp. 391–434 (1990). 16 Kenter C and Hoffmann C, Impact of weather on yield formation of sugar beet in Germany. Adv Sugar Beet Res 5:19–32 (2003). 17 Senge P, The Fifth Discipline: The Art and Practice of the Learning Organization (2nd edn) Century 86, London (2006). 18 Agroclimatic Atlas of Alberta: Agricultural Climate Elements. [Online]. Available: http://www1.agric.gov.ab.ca/$department/deptdocs.nsf/ all/sag6301 [28 January 2014]. 19 Quadiri G, Hassan FU and Malik Ma, Growing degree days and yield relationship in sunflower. Int J Agric Biol 4:564–568 (2007). 20 Agri-Food Canada, Lessons learned from the Canadian drought years 2001 and 2002. [Online]. Available: http://www4.agr.gc.ca/AAF C-AAC/displayafficherdo? id=1326987176314&lang=eng [22 June 2013]. 21 Max Planck Institute for Meteorology. [Online]. Available: http:// www.mpimet.mpg.de/en/ wissenschaft/modelle/echam/echam5. html [29 January 2014]. 22 Roeckner E, Brokopf R, Esch M, Giorgetta M, Hagemann S, Kornblueh L et al., Sensitivity of simulated climate to horizontal and vertical resolution in the ECHAM5 atmosphere model. J Climate 19:3771–3791 (2006). ˝ 23 Muller WA and Roeckner E, ENSO teleconnections in projections of future climate in ECHAM5/MPI-OM. Clim Dynam 31:533–549 (2008). 24 USGS Regional and Global Climate. [Online]. Available: http:// regclim.coas.oregonstate.edu/visualization/rcd/regional-climatedownloader/index.html#project=regcm [29 January 2014]. 25 Alberta Sugar Beet Growers (ASBG). [Online]. Available: http:// www.asbg.ca/news/news-release [6 May 2014].
521
J Sci Food Agric 2015; 95: 515–521
© 2014 Society of Chemical Industry
wileyonlinelibrary.com/jsfa
