Calibration and validation of the INFOCROP model for moisture stress situation in cotton

Introduction

Cotton (Gossypium hirsutum L.) "The king of fibre" and the leading fibre crop of the world is grown over an area of 34.2 m.ha with a production of 25.3 mt and productivity of 733.0 kg per ha. In India, it is grown on an area of 111.61 ha with a production of 312 million bales and the productivity is 475.0 kg per ha. Whereas in Karnataka, it is grown over an area of 5.34 lakh hectares with a production of 10.0 lakh bales and a productivity of 318.0 kg lint per hectare (Anon., 2010) [10] . In recent times several tools or systems research relating to computer and information technology have become available. Crop growth simulation model is one such tool. They are based on quantitative understanding of underlying processes and integrate the effects of soil, weather, crop, pests and management factors on growth and yield. These models need expensive and time-consuming field experimentations as they can be used to extrapolate the results of the research conducted in one season or one location to other seasons, locations or management. They also provide a means to quantify the effect of drought at different growth stages, soil and management on crop growth, yield and sustainability of agricultural production. For conducting such studies, the models are evaluated after caliberating and validating them under local conditions. There are more than 25 crop simulation models developed around the world in cotton crop. One such model namely INFOCROP has been developed based on parent models viz., SUCROS, ORYZA and MACROS (Aggarwal et al., 2006) [2] . A generic model 'INFOCROP' has been calibrated and validated using factors like weather, soil, genotypes, date of sowing, nitrogen level as basic inputs (Hebbar et al., 2007) [6] . The model has simulated the phenology more accurately and accuracy of simulated yield was 94% and biomass 89% across the agro-climatic zones. Using the crop model an integrated method was developed along with remote sensed area estimation and GIS techniques incorporating soil and weather parameters to predict the cotton production at regional level.

Materials and Methods

The INFOCROP used is from the Compact Visual Fortran version 6.

A) Effect of dates of sowing on growth and yield

Table on effects of dates of sowing on growth and yield is presented in Table 1 (a) and (b). Each year four dates of sowing were under taken at an interval of 20 days and data sets collected in each year was incorporated into the INFOCROP model and output was compared (Actual Vs Simulated).

B) Effects of moisture regimes on growth and yield

In the present study, the impact of moisture stress on growth and yield parameters were computed as the mean values in field and ROS experiment for the four moisture levels (M1 to M4). This set of data was used as input data for the simulation. The actual and simulated data is presented in Table 1c.

Results and Discussion

The data sets (observed Vs simulated) generated out of similar study is presented in Table 1a and 1b. In general, there was decrease in parameters associated with growth and phenology as the date of sowing delayed from D1 to D4. The model showed more number of days for phenological observation compared to observe. The values of yield and yield components were more in simulated than observed except boll weight. The simulated boll weight is less than observed boll weight. The values for many of these parameters listed decreased with the delay in sowing (D1 to D4).

In present study (2010) (Table 1c) indicated decrease in yield and yield components with decrease in soil moisture content. The actual yield decreased from 1603 to 1405 kg/ha from M1 to M4. The simulated yield also showed similar trend (3938 to 1969 kg/ha). In simulation models, phenology is calculated using heat units in cotton (Hebbar et al., 2002) [4] . In simulation model, it is customary to represent the growth stage from sowing to maturity on same scale. In INFOCROP, this scale is given from 0 to 2. The development stage 0 (zero) represents sowing and 2 represents maturity. The developmental stage for anthesis is given as 1.0. In most cases, cotton at early growth stages suffers from intermittent drought, water logging or insect attack and at later stages due to the occurrence of multiple flushes of bolls. The INFOCROP model, simulated more number of days for anthesis and maturity except in case of Bt cotton hybrids during 2005-2006 (Table 1a). Cotton (Kiran Kadam et al., 2024) [9] basically is an indeterminate crop where the vegetative and reproductive phases overlap resulting in intra plant competition for photosynthesis between developing bolls and vegetative parts (main stem, lateral branches and leaves). In present day Bt cotton hybrids, there is increase in determinancy. In the beginning of the reproductive phase, almost all the squares formed will set into bolls. As a result, there is more sink activity and comparatively less source. As a result, there is increased determinance. In data sets used for model validation large variation was seen for time to flowering and maturity (Hebbar et al., 2004) [10] . In crop model's phenology is almost important parameter that decides the growth and yield and it is computed in terms of thermal units. The mean TDM (from field and ROS experiments) decreased from 2783 kg/ha in irrigated to 2368 kg/ha in rainfed. By incorporation of these values in the INFOCROP model, it simulated TDM 3938 kg/ha in irrigated (M1) to 1969 kg/ha in rainfed (M4). Thus, the deviation in actual TDM recorded to simulated is from 29% in irrigated (M1) to 16% in rainfed (M4). The LAI in the field experiment decreased from 4.19 in irrigated (M1) to 3.35 in rainfed condition (M4). In ROS, it decreased from 3.26 in irrigated to 2.31 in rainfed. By incorporation of these valued in INFOCROP, it simulated LAI of 4.44 in irrigated (M1) to 3.0 in rainfed (M4). The intermediate moisture levels (M2 and M3) recorded LAI values in between these. Thus, the deviation in actual LAI observed to simulated is from 5% in irrigated (M1) to 10% in M4. In ROS, these deviation ranged from 26.5% in irrigated (M1) to 23% in rainfed (M4). Similar results were concluded by Hebbar et al., (2004) [10] . LAI simulated under irrigated condition was more accurate compared to rainfed condition. Model was sensitive to the differences seen in LAI between rainfed condition. During cropping season, the crop experiences heavy rains resulting in more moisture in the field and also at times periods of dry spells occur. Though, model simulates the effects of these stresses correctly, sometimes it become to sensitive which could have possibly under estimated the LAI in rainfed condition (Hebbar et al., 2004) [10] . The mean yield (from field and ROS experiment) decreased from 2783 kg/ha in irrigated (M1) to 2368 kg/ha in rainfed condition (M4). By incorporation of these values in INFOCROP model, it simulated yield 3938 kg/ha in irrigated to 1969 kg/ha in rainfed. Thus, the deviation in actual yield recorded to simulated is from 32% in irrigated (M1) to 22% in rainfed (M4). INFOCROP model simulates 85% yield to the observed yield under moisture stress condition (Hebbar et al., 2004) [10] . Similarly, the model simulates yield, TDM, LAI, boll weight and number of bolls per hectare at different moisture levels. The mean values in the earlier studies (Table 1) (Patil et al., 2004) showed that the output simulated by INFOCROP is in agreement with the actual data recorded in these studies. However, the model deviated more for parameters like boll weight. Model simulated yield and biomass to an accuracy of 85% and 74% respectively (Hebbar et al., 2004) [10] . Under rainfed condition, several factors like biotic and abiotic factors attributed in reduction of yield. Thus the final yield is very complex phenomenon and it depends on the amount of fruit loss and the plants ability to compensate for the loss (Sadras, 1996). This model simulated most of these aspects reasonably. The difficulty in recording of lost leaves, squares and bolls in between the observations is partially responsible for not getting good agreement between simulated and observed biomass at maturity. Thus, INFOCROP model can simulate the growth and yield under different moisture levels. The water stress, in the present model as determined by the ratio of actual water uptake and potential transpiration will influence the rate of germination through thermal time and the leaf temperature during growth (Aggarwal et al., 2000) [1] .

DAS

References

1. Aggarawal PK, Kalra N, Chander S, Pathak H. INFOCROP: A generic model for simulating the effect of weather, soil, variety, management and pests on growth and yield of annual crops in tropical environments. Users manual. 2000. p.24.
2. Aggarawal PK, Kalra N, Chander S, Pathak H. Infocrop: a dynamic simulation model for assessment of crop yields losses due to pests and environmental impact of agro-ecosystems in tropical environments. I. Model description. Agric Syst. 2006;89:1-25.
3. Anonymous. All India Co-ordinated Cotton Improvement Project. Annu Rep. 2010;1:2-5.
4. Hebbar KB, Venugopalan MV, Rao MRK, Gadade GD, Chatterji S, Mayee CD. Effect of sowing dates and fertilizer levels on phenology, growth and yield of cotton. Indian J Plant Physiol. 2002;7:380-383.
5. Hebbar KB, Venugopalan MV, Rao MRK, Seshasai MV, Rao MVR, Patil BC, et al. Predicting cotton production using INFOCROP cotton simulation model, remote sensing and spatial agroclimatic data. Current Sci. 2008;95:1570-1579.
6. Hebbar KB, Perumal NK, Khadi BM. Photosynthesis and plant growth response to transgenic Bt cotton (G. hirsutum L.) hybrid under field condition. Photosynthetica. 2007;45:254-258.
7. Sadras VO. Cotton responses to simulated insect damage: radiation use efficiency, canopy architecture and leaf nitrogen content as affected by loss of reproductive organs. Field Crop Res. 1996;48:199-208.
8. Hebbar SS, Ramachandrappa BK, Nanjappa HV, Prabhakar MJ. Studies on NPK drip fertigation in field grown tomato (Lycopersicon esculentum Mill.). European Journal of agronomy. 2004 Jun 1;21(1):11727.
9. Kadam K, Chinchane VN, Deshmukh S, Rani R. Correlation and path analysis of yield and fibre quality traits in upland cotton (Gossypium hirsutum L.). Int. J. Adv. Biochem. Res. 2024;8(3):61-66. DOI: 10.33545/26174693.2024.v8.i3a.675
10. Ladoux B, Anon E, Lambert M, Rabodzey A, Hersen P, Buguin A, Silberzan P, Mege RM. Strength dependence of cadherin-mediated adhesions. Biophysical journal. 2010 Feb 17;98(4):534-42.