Theses and Dissertations http://www.erepo.iihr.ernet.in/handle/123456789/277 2013-05-15T12:25:10Z 2013-05-15T12:25:10Z R, Hanumanthaiah http://www.erepo.iihr.ernet.in/handle/123456789/595 2012-12-15T20:30:12Z 2012-01-01T00:00:00Z

Title: Evolving Statistical models for crop-logging studies in Brinjal (Solanum Melongena L.) Authors: R, Hanumanthaiah Abstract: Crop yield forecast before harvest is likely to provide valuable information to farmers, policymakers/government on sales, storage, and export, price fixation, grading, and marketing for advance planning so as to ensure sustainable crop production during the years ahead. Researchers are also interested to know explicitly by which stage of the crop, yield could be predicted more accurately and what are all the significant crop-logging parameters. Crop improvement research is also be benefited, as selection can be made in the early stages based on the significant crop-logging parameters identified. To this end, statistical models were developed using Multiple Linear Regression (MLR) and Artificial Neural Network (ANN) methods. Statistical models developed showed that Brinjal crop yield could be predicted well in advance as early as 26 Days After Planting (DAP) using three biometrical traits (plant height, plant girth and plant spread north south ) to an extent of 71 %. As the DAP increases prediction of yield could be possible to an extent of 88 %. Identification and removal of outliers in the data set increased the prediction of MLR models in the range of 31 %, 33 %, 34 and 8% respectively across four crop growth stages. ANN approach which was also used to predict the yield resulted in R2 values 83 % (stage 1), 89 % (stage 2), 88 % (stage 3) and 68 % (stage 4), which was high as compared to MLR (except for stage 4). Cross-validation of MLR and ANN models for all four stages, showed good results as the prediction power was in the range of 78 to 87 % R2- for MLR and 64 % to 85 %- for ANN. Hence, it is recommended to study the role of outliers before developing crop yield forecasting models and also to exploit ANN approaches by capturing the inherent nonlinearity among biometrical traits.

2012-01-01T00:00:00Z G S, Ravi http://www.erepo.iihr.ernet.in/handle/123456789/594 2012-12-15T20:30:12Z 2011-01-01T00:00:00Z

Title: Statistical models for stability research in cucumber Authors: G S, Ravi Abstract: Crop improvement research is mainly aimed to exploit initially the genetic diversity available in the germplasm and culminate with identifying stable lines for release as variety. Its main aim is to estimate the average response of the genotypes and also to test the consistency of the yield responses over years/environments. The presence of genotype X environment interaction (GEI) makes it difficult to assess the genetic potential of a variety. In the present study, three different approaches were used to develop stability models for assessing the stability of 33 cucumber lines tested over three consecutive years in one location (Bangalore), based on eight yield and related biometrical traits. GEI was highly significant for all the traits and the genotypes had divergent response to environmental changes. Presence of significant linear GE interactions in both yield per plot (Kg.) and number of fruits per plot indicated that there is still more potential for crop improvement over years. Measures of stability when used to group the 33 genotypes into genotypes suitable for ideal environment, for favorable environment and for poor environment, fortified a distinct difference in their grouping using two approaches. percentage of misclassification was in the range of 20-100% due to the non-utility of Freeman-Perkins (FP) method. Parametric and non-parametric measures were computed to assess the extent of contribution of each of the 33 genotypes to GE interaction. It was observed that the lines CH-36-71-6 and CH-32-36-6 were most stable in yield character. Also, a combined index by giving desired importance to all the traits was developed to rank the genotypes. Results showed that CH-20-1-10, CH-1-42-10 and CH-32-36-6 under ER model and CH-28-32-6, CH-20-1-10 and CH-32-36-6 under FP model were top 3 stable genotypes. Finally, by considering relative performance of a genotype various nonparametric measures computed showed that CH-20-1-10 and CH-28-32-6 were found to be most stable. SAS programming codes and STAB-IIHR were generated for data analysis. Further, in any crop improvement research, as the breeders may expect that a genotype/variety should also possess desirable characters of other yield related traits. Hence, an index based on the combined ecovalance value and relative performance of a genotype as compared to others (for a character under study), using rank based nonparametric measures may be still more practically meaningful so as to come out with stable lines either for release as variety or as a promising line in the ensuing crop hybridization trails

2011-01-01T00:00:00Z N, Vijay http://www.erepo.iihr.ernet.in/handle/123456789/593 2012-12-15T20:30:12Z 2010-01-01T00:00:00Z

Title: Investigation and development of non-linear statistical models for disease forecasting in grapes Authors: N, Vijay Abstract: An attempt was made to develop nonlinear stochastic models for disease forecasting in Grapes. Grape (Vitis SPP.) is an important crop for the farmers for getting higher returns and with consumer for delicacy and as a medicinal fruit. Downy mildew is one of the most destructive vine diseases known leading to total crop losses. Thus, the remunerative and successful cultivation of grapes has been hampered. To avoid this, by understanding prevailing weather conditions which influence the onset, initiation and progression of disease, judicious need based control measures are the need of the hour. To this end, the present investigation was carried out to understand the role of weather factors on downy mildew incidence in Grapes (cv. Anab-e-Shahi) and disease progression over time epoch by developing suitable statistical models. Efforts were made to develop models individually for backward and fore pruning periods, resulting in meaningful interpretation to the researchers. Also, an attempt was made to investigate statistical considerations involved in the error structure and subsequent methodologies to be followed, while developing non-linear models. Using the nonlinear models developed, an index was also developed to compute quantitative information about the biological parameters concerning intrinsic infection rate and maximum mildew severity over time-epoch. Statistical models developed for backward pruning data (May-June) showed that maximum temperature, Evaporation and relative humidity at 7.30 hrs, observed with a time lag of one week, collectively explain about 89.4% of the variation in downy mildew incidence. Statistical models developed for fore pruning data (September-October) showed that minimum temperature, relative humidity at 7.30 hrs and 13.30 hrs, observed with a time lag of one week, collectively explain 88% of the variation in weekly downy mildew incidence. Logistic and Gompertz nonlinear stochastic statistical models developed expressed the disease progression to the extent of 97-99%. These models when used to compute quantitative information about the biological parameters concerning intrinsic infection rate and maximum mildew severity over time-epoch showed that, in general, for backward and fore pruning data, the rate of disease severity was maximum during the fourth- fifth week and fifth- sixth week after pruning, respectively. Hence, appropriate management strategies for controlling the disease should be oriented within the period identified in the investigation, separately for backward and fore pruning. Resultant nonlinear models were used to compute the Area under Disease Progressive Curve (AUDPC). A perusal indicates that the values obtained by logistic and Gompertz are ranged from 48 to 84 and 25 to 65 respectively for backward pruning data. However, for the fore pruning data the results showed that AUDPC values were higher as it ranged from 78 to 86 and 61 to 65 respectively. These results indicate that the downy mildew rate of progression in Fore pruning is much severe than in backward pruning. SAS programming codes were generated for model building. The message arising out of this present investigation is that proper prophylactic measures, if taken by considering the model resulted significant weather factors along with knowledge about disease progression over time as depicted by nonlinear models, separately for backward and fore pruning, not only results in an efficient and economic management strategies for controlling downy mildew incidence in grapes (cv. Anab-e-Shahi) but also considerably reduce crop yield loss thereby providing better return to the farmers. The graphical representation of nonlinear models fitted is depicted as below.

2010-01-01T00:00:00Z K J, Sowmya http://www.erepo.iihr.ernet.in/handle/123456789/591 2012-12-15T20:30:08Z 2012-01-01T00:00:00Z

Title: Studies on physiological, biochemical and molecular aspects of seed invigouration in cucumber (Cucumis sativus L.) Authors: K J, Sowmya Abstract: Cucumber is a popular Cucurbitaceous vegetable and the ever growing demand of this vegetable throughout the year exerts challenge for continuous production even during off seasons. The present study was conducted to standardize the seed invigouration protocol, to know the physiological, biochemical and molecular changes due to seed invigouration, to test the performance of invigourted seeds under aboitic stress conditions and to test the storage potentiality of invigourated cucumber seeds. The crop specific efficacy of all the popular methods of priming viz., hydropriming, osmopriming, chemopriming and biopriming, were standardized based on the performance of seed quality attributes. Results revealed that the best priming temperature was 25±1 oC, the optimum duration of soaking was 48h, best osmotica was PEG 6000 @-1.5 Mpa, the best chemicals were KNO3 @1% or Ethrel 1000 ppm followed by GA3 @ 100 ppm, KH2PO4 @10-1M, Thiourea @1% KH2PO4 @10-1M, NaCl @10-1M; and Oxalic acid @10-1M in their order of merit and the best biological agents were Cowdung slurry followed by PSB, Vermiwash, Tricoderma viridae and Azospirillum in the order of performance. All the physiological attributes such as first count germination, final count germination, Bartlett’s Rate Index (BRI), Coefficient of velocity (CV) of germination, germination energy (GE), Mean seedling length, Mean seedling dry weight, Seedling vigour index -I and Seedling vigour index - II were significantly higher (91.67%, 92.33%, 0.544, 73.54%, 85.50% , 31.68 cm, 11.69 mg, 2929 and 1081, respectively) in KNO3 primed seeds. KNO3 primed seed recorded significantly higher Total dehydrogenase activity, total soluble protein, amylase activity, catalyse (CAT) activity, peroxidise activity (POX) and lower Electrical conductivity and total soluble sugars in the seed leach ate. The total soluble seed protein profile of native PAGE and SDS PAGE had revealed polymorphism with respect to appearance (total of 26 bands and 27 bands) and disappearance of peptides at specific Rm values (0.033 to 0.846 and 0.100 to 0.966, respectively) in primed and unprimed seeds. Priming induced proteins were expressed in all the priming treatments which can be employed as a markers for optimum priming. Esterase and peroxidise isozymes expression also varied in primed and unprimed seeds. Primed seeds expressed specific is forms of isozymes compared to unprimed. At the end of 48 hr, primed seeds exhibited more amount DNA content per se due to advancement in cell cycle when compared to unprimed seeds. Primed seeds showed better tolerance to temperature, moisture and saline stress by exhibiting advancement in germination, higher seedling vigour, higher fresh and dry seedling weight compared to unprimed seeds. Among priming treatments, higher (88.63%) FEM, BRI (0.479), PSP (86.13) and PDW (2.06 g) was indicated in seeds primed with KNO3 @ 1% and it was lower (67.50%, 0.391, 53.13% and 1.77 g) in unprimed seeds (control). High vigour seeds primed with KNO3 @1%, packed in super grain bag registered higher (89.00 and 94.50%, 86.00 and 93.50%, 2697 and 2902; 949 and 1091) germination, filed emergence, SVI- I and SVI- II, at the end of the storage period of 80 days when stored under ambient and refrigerated condition, respectively. In addition to KNO3, cow dung slurry primed seeds were also recorded significantly higher seed quality attributes on par with KNO3 treatment. In order to test the longevity of primed seeds during storage, primed seeds were sealed in various packaging material and stored at ambient and refrigerated conditions. Various vigour parameters were compared with unprimed seeds at 20 days interval. Among storage treatments, seeds stored under refrigerated condition showed slightly higher germination and field emergence (82.25% and 80.58%) compared to ambient condition (81.67% and 79.50%), respectively. Results indicated primed seeds can be stored for a short period of 80 days under ambient conditions of Bangalore without significant reduction in seed quality attributes. However, refrigerated storage is advised for long term storage of primed cucumber seeds.

2012-01-01T00:00:00Z