Pemodelan beberapa Genotipe Tanaman Padi (Oryza Sativa) dalam Merespon Variasi Iklim

Wicaksono, Ade Fajar and Prof. Dr. Ir. Ariffin, M.S. and Dr. NoerRahmi Ardiarini, S.P., M.Si. (2024) Pemodelan beberapa Genotipe Tanaman Padi (Oryza Sativa) dalam Merespon Variasi Iklim. Magister thesis, Universitas Brawijaya.

Abstract

Padi (Oryza sativa) merupakan salah satu dari makanan pokok yang di konsumsi lebih dari 3 milyar masyarakat dunia. Meskipun dengan produksi yang tinggi, Negara Indonesia masih rentan dengan ketahanan pangan. Besarnya jumlah penduduk Indonesia, yakni 278 juta jiwa dan memiliki pertumbuhan sekitar 1,07% per tahunnya, diperkirakan apabila peningkatan produksi tidak dilakukan, maka akan terjadi kelaparan di berbagai daerah. Beberapa kekhawatiran selain dari pertumbuhan penduduk muncul dari perubahan iklim yang terasa saat ini. Perubahan iklim melalui perubahan pola curah hujan, peningkatan konsentrasi CO2 di atmosfer dan peningkatan suhu dapat mempengaruhi hasil panen. Oleh sebab itu, kerentanan dan keberlanjutan dari sektor pertanian sangatlah dipengaruhi oleh kondisi iklim yang berubah. Kondisi ini yang menjadikan berbagai peneliti di belahan dunia untuk mencari jalan pencegahan, adaptasi, serta menduga bagaimana kedepannya produksi tanaman harus diarahkan. Pemanfaatan pemodelan tanaman juga dirasa penting karena dapat menduga pada skala luas dibandingkan harus melakukan percobaan di lapang atau plot yang membutuhkan sumber daya besar serta memiliki resiko dalam pelaksanaan seperti tidak tepatnya praktik manajemen yang dilakukan. Melalui penggunaan suatu model, respon tanaman dan hasil akan dapat diprediksi di tengah perubahan iklim yang terjadi. Ketika model sudah disusun diperlukan proses evaluasi dan validasi. Evaluasi dan validasi sangat diperlukan mengingat iklim yang berbeda akan menyebabkan perbedaan yang nyata antara potensi hasil dengan hasil yang ada di lapang. Oleh sebab itu, penelitian ini akan menyusun dan memvalidasi model yang disusun berdasarkan data hasil panen dan data iklim di dua kecamatan di Kabupaten Malang dan satu kecamatan di Kota Batu. Selain itu, akan dilakukan juga uji stabilitas genotipe yang dilakukan dengan berbagai model stabilitas. Penelitian ini dilaksanakan pada bulan Juni hingga oktober di tiga lokasi yakni Kepanjen, Karangploso, dan Junrejo. Penanaman akan dilakukan di ketiga lokasi tersebut dengan menanam 6 genotipe F7 hasil persilangan padi gogo dan padi sawah yakni genotipe G1, G2, G3, G4, G5, dan G6 serta 3 varietas pembanding yakni situbagendit (SB), cibogo (CB), dan ciherang (CH). Alat dan bahan yang digunakan antara lain alat dan bahan budidaya tanaman padi, software analisis data yakni Rstudio. Karakter yang diamati antara lain tinggi tanaman, hari berbunga, hari panen, jumlah anakan produktif, jumlah malai per rumpun, panjang malai, jumlah biji bernas per malai, jumlah biji total per malai, persentase biji bernas, berat 1000 biji, berat biji per tanaman, dan produktivitas atau panen per hektar. Karakter lain yang akan diamati yakni variabel iklim yang diakses melalui BMKG online pada saat penanaman. Analisis data dalam menyusun model menggunakan sequential sequential forward selection, sequential backward elimination, stepwise regression,correlation base selection, variance inflation factor base elimination, regresi ridge, dan regresi LASSO. Uji validasi model menggunakan root mean squared error (RMSE) dan relative root mean squared error (RRMSE). Analisis data stabilitas menggunakan analisis ragam masing-masing lokasi, analisis ragam gabungan (combine ANOVA), uji stabilitas dan adaptabilitas Finlay-Wilkinson, dan uji stabilitas Eberhart-Russell. Analisis korelasi pearson juga dilakukan untuk mengetahui hubungan variabel iklim terhadap karakter tanaman yang diamati, Analisis data akan menggunakan beberapa package yang ada pada Rstudio yakni olsrr, glmnet, ehagof, metan, dan agricolae. Hasil analisis menunjukkan model yang disusun menggunakan sequential selection variabel tidak dapat menduga produktivitas tanaman padi pada seluruh lokasi. Model lain yang disusun menggunakan VIF dan LASSO tidak dapat menduga produktivitas beberapa genotipe tanaman padi pada lokasi tertentu. Keseluruhan model tidak dapat digunakan untuk menduga produktivitas tanaman padi varietas situbagendit. Namun, model yang disusun dengan metode sequential backward elimination, stepwise regression, correlation base selection, dan regresi ridge mampu menduga produktivitas genotipe G1, G2, G3, G4, G5, G6, CB, dan CH di seluruh lokasi penelitian. Hasil analisis ragam dan analisis ragam gabungan menunjukkan bahwa terdapat perbedaan antar genotipe di masing-masing lokasi maupun di ketiga lokasi. Namun, hasil uji beda nyata jujur menunjukkan hanya karakter berat 1000 biji yang menunjukkan perbedaan antara genotipe F7 hasil persilangan dengan varietas pembanding. Hasil analisis ragam gabungan menunjukkan terdapat interaksi pada karakter tinggi tanaman, hari berbunga, hari panen, panjang malai, jumlah biji bernas, dan berat 1000 biji. Hasil stabilitas dan adaptabilitas menunjukkan tidak terdapat genotipe yang memiliki kestabilan di keseluruhan karakter yang diamati. Namun, genotipe G6 memiliki 10 karakter stabil dan 2 karakter adaptif. Sehingga, genotipe G6 dapat di jadikan calon varietas unggul baru.

English Abstract

Rice (Oryza sativa) is one of the staple foods consumed by more than 3 billion people in the world. Even with high production, Indonesia is still vulnerable to food security. With a population of 278 million people and growth of around 1.07% per year, it is estimated that if production is not increased, there will be famine in various regions. Several concerns apart from population growth arise from climate change that is currently being felt. Climate change through changes in rainfall patterns, increased CO2 concentrations in the atmosphere and increased temperatures can affect crop yields. Thus, the vulnerability and sustainability of the agricultural sector is greatly influenced by changing climate conditions. This condition has led various researchers in various parts of the world to look for ways to prevent, adapt, and predict how future crop production should be directed. The use of plant modeling is also considered important because it can make predictions on a wide scale rather than having to carry out experiments in the field or plots which require large resources and have risks in implementation such as inaccurate management practices. With a model, plant responses and yields can be predicted amidst climate change. Once the model has been prepared, an evaluation and validation process is required. Evaluation and validation are very necessary considering that different climates will cause real differences between potential results and results in the field. Therefore, this research will compile and validate a model based on harvest data and climate data in two sub-districts in Malang Regency and one sub-district in Batu City. Apart from that, genotype stability tests will also be carried out using various stability models. This research was carried out from June to October in three locations, namely Kepanjen, Karangploso, and Junrejo. Planting will be carried out in these three locations by planting 6 F7 genotypes resulting from crossing upland rice and lowland rice, namely genotypes G1, G2, G3, G4, G5, and G6 as well as 3 comparison varieties, namely situbagendit (SB), cibogo (CB), and ciherang (CH). The tools and materials used include rice cultivation tools and materials, data analysis software, namely Rstudio. Characters observed include plant height, day to flowering, day to harvest, number of productive tillers, number of panicles per hill, length of panicle, number of grains per panicle, total number of grains per panicle, percentage of grains, weight of 1000 grains, weight of grains per plant , and productivity or harvest per hectare. Another character that will be observed is the climate variable which is accessed via BMKG online at the time of planting. Data analysis in constructing the model uses sequential forward selection, sequential backward elimination, stepwise regression, correlation base selection, variance inflation factor base elimination, ridge regression, and LASSO regression. The model validation test uses root mean squared error (RMSE) and relative rootv mean squared error (RRMSE). Stability data analysis used analysis of variance for each location, combined analysis of variance (combined ANOVA), FinlayWilkinson stability and adaptability test, and Eberhart-Russell stability test. Pearson correlation analysis was also carried out to determine the relationship between climate variables and the observed plant characters. Data analysis will use several packages available in Rstudio, namely olsrr, glmnet, ehagof, metan, and agricolae. The results of the analysis show that the model prepared using sequential variable selection cannot predict the productivity of rice plants in all locations. Meanwhile, models prepared using VIF and LASSO cannot predict the productivity of several rice plant genotypes at certain locations. The entire model cannot be used to estimate the productivity of the Situbagendit rice variety. However, the model prepared using the sequential backward elimination, stepwise regression, correlation base selection and ridge regression methods was able to estimate the productivity of G1, G2, G3, G4, G5, G6, CB and CH genotypes at all research locations. The results of analysis of variance and combined analysis of variance showed that there were differences between genotypes in each location and in the three locations. However, the results of the honest significant difference test showed that only the weight of 1000 grains showed the difference between the F7 genotype resulting from the cross and the comparison variety. Meanwhile, the results of the combined analysis of variance showed that there was an interaction between the characteristics of plant height, day of flowering, day of harvest, panicle length, number of grainy grains, and weight of 1000 grains. The results of stability and adaptability show that there is no genotype that has stability in all the characters observed. However, the G6 genotype has 10 stable characters and 2 adaptive characters. So, the G6 genotype can be used as a candidate for a new superior variety.

Item Type:	Thesis (Magister)
Identification Number:	042404
Divisions:	S2/S3 > Magister Ilmu Tanaman, Fakultas Pertanian
Depositing User:	Unnamed user with username nova
Date Deposited:	05 Dec 2024 04:35
Last Modified:	05 Dec 2024 04:35
URI:	http://repository.ub.ac.id/id/eprint/233398

Text (DALAM MASA EMBARGO) Ade Fajar Wicaksono.pdf Restricted to Registered users only Download (7MB)

Actions (login required)

View Item