Publication: Çarşamba Ovası'nda Buğday, Mısır ve Soya İçin Verim ve Bazı Verim Unsurlarının Matematiksel Modellenmesi
Loading...
Date
Authors
Journal Title
Journal ISSN
Volume Title
Abstract
Bu çalışmada, Çarşamba Ovasının buğday, mısır ve soya bitkileri yetiştirilen tarım topraklarının bazı fiziksel, kimyasal ve bitkilerin bazı agronomik özelliklerinin (bitki boyu, bin tane ağırlığı ve tane verimi) dağılımı; bitkilerin agronomik özellikleriyle toprak özellikleri arasındaki korelasyon ilişkilerin belirlenmesi ve bu ilişkilere dayanarak agronomik ve farklı toprak özellikleri arasındaki regresyon modellerinin oluşturulması; elde edilen modellerin bölge topraklarında bitki veriminin tahmin edilmesinde uygulanabilirliği araştırılmıştır. Bu amaçla, Samsun ilinde yer alan Çarşamba Ovasının 60 köyünde çiftçiler tarafından tarım yapılan arazilerden toprak ve bitki örnekleri alınmıştır. Araştırmanın amacı doğrultusunda, toprakların bazı fiziksel (tekstür, tarla kapasitesi, solma noktası, hacim ağırlığı) ve kimyasal (organik madde, toprak reaksiyonu, elektriksel iletkenlik, kireç içeriği, toplam azot, değişebilir katyonlar, yarayışlı fosfor, katyon değişim kapasitesi, alınabilir Fe, Cu, Zn, Mn); bitkilerin ise agronomik özellikleri belirlenmiştir. Araştırma 2013-2014 yılları arasında aynı arazilerde ve aynı bitkiler ile yürütülmüştür. Buğday ve soya bitkilerine ait korelasyon ilişkileri karşılaştırıldığında, mısır yetiştirilen toprakların fiziksel ve kimyasal özellikleri ile mısırın agronomik özellikleri arasında, istatistiksel açıdan çok önemli ve önemli ilişkiler daha fazla bulunmuştur. Buğday bitki boyu (BBB) ile EC, Ca+Mg, Db, kil, (Db)2, (EC)2, (Kil)2,√(Ca+Mg) toprak özellikleri arasındaki regresyon modelinde, istatistiksel anlamlılık (p=0.035) düzeyinde en yüksek regresyon katsayısı (R=0.731) saptanmıştır. Buğday bin tane ağırlığı (BBTA) ile (EC)2, (OM)2, (Fe)2, (Kil)2, (Db)2, (Kil×Db),√Kil ,√Db , SN parametreleri arasındaki regresyon modelinde ise, en yüksek regresyon katsayısı R=0.794 olarak, p=0.013 istatistiksel anlamlılık değerinde bulunmuştur. Buğday tane verimi (BY) ile EC, CaCO3, Kum, (Kum×Db), (Db×SN), Fe, N, (Db)2, SN, (EC)2,√Db terimleri biçiminde toprak özellikleri arasındaki regresyon modeli istatistiksel olarak anlamlı (p=0.012<0.05) olup, en yüksek regresyon katsayısına (R=0.840) sahip olmuştur. Mısır bitki boyu (MBB) ile toprakların (CaCO3)2, (EC)2, (OM)2, Db, TK,√n , Ca+Mg, CaCO3, OM,√Kum parametreleri arasındaki ilişkide regresyon katsayısı (R=0.543) orta düzeyde olup, istatistiksel olarak anlamlı farklılık saptanmamıştır (p>0.10). Mısır bin tane ağırlığı (MBTA) ile pH, CaCO3, EC, OM, ,√(CaCO_3 ) , Kum, (Db×Kum), (Db)2, SN parametreleri arasındaki regresyon modeli istatistiksel olarak anlamlı olup (p=0.012), diğer modellerle karşılaştırıldığında, en yüksek regresyon katsayısına (R=0.819) sahip olmuştur. Mısır tane verimi (MY) ile OM, N, P, K, Ca+Mg, Na, Zn, SN, Db, silt, (Db×SN) parametreleri arasındaki ilişki yüksek düzeyde istatistiksel anlamlılık (p=0.001) göstermiş, regresyon katsayısı ise (R=0.894) çok yüksek olarak belirlenmiştir. Soya bitki boyu (SBB) ile (EC)2, (OM)2, N, Cu, K, Db, Kil, (Db×Kil), SN, BYS,√Db toprak özellikleriyle oluşturulan regresyon modeli istatistiksel anlamlılık (p=0.091<0.10) eğilimi dahilinde olup, regresyon katsayısı yüksek düzeyde (R=0.766) saptanmıştır. Soya bin tane ağırlığı (SBTA) ile (KDK)2, (CaCO3)2, N, P, K, Na, (OM)2, (EC)2,√Db , (Db×Kum), SN, TK,√Kum , silt parametreleri arasındaki regresyon modelinde, diğer modellerle karşılaştırıldığında, en yüksek regresyon katsayısı (R= 0.782) elde edilmiştir. Soya tane verimi (SY) ile EC, CaCO3, OM, N, P, K, (OM)2, (EC)2, (CaCO3)2, Zn, Mn,√(N×P×K) toprak özellikleri arasında oluşturulan regresyon modeli çok yüksek istatistiksel anlamlılık (p<0.001) göstermiş, regresyon katsayısı (R=0.921) ise çok yüksek olarak belirlenmiştir. Modellerin geçerliliğinin belirlenmesinde, hata kareler ortalamasının karekökü (HKOK), uygunluk indeksi (d), modelin etkinliği (ME), mutlak hata (MH), ortalama aritmetik hata (OAH) ve nispi hata (NH) parametreleri birlikte değerlendirilmiştir. İstatistiksel parametrelerden kullanılarak modellerin geçerliliğinin değerlendirilmesi sonucunda, buğday, mısır ve soya bitkilerinin agronomik özellikleri ile araştırılan toprak özellikleri arasında oluşturulan modellerin uygulanabilirliğinin mümkün olabileceği gözükmektedir. Buğday bitki boyu ile toprakların fiziksel-kimyasal özellikleri; tane verimiyle fiziksel-kimyasal özellikleri; bin tane ağırlığı ile kimyasal özellikleri arasında elde edilen modellerin etkinliği optimal düzeyde olup, agronomik özelliklerin tahmininde kullanılması mümkün olabilmektedir. Mısır bitkisinin boyu ile sırasıyla toprakların fiziksel ve fiziksel-kimyasal özellikleri; bin tane ağırlığıyla fiziksel-kimyasal özellikleri; tane verimiyle sırasıyla fiziksel ve kimyasal özellikleri arasında oluşturulan modeller daha etkin ve güvenirliği yüksek olmuştur. Soya bitkisine ait yapılan regresyon modellerinde ise, soya bitki boyu ile sırasıyla toprağın kimyasal ve fiziksel-kimyasal özellikleri; bin tane ağırlığı ile kimyasal özellikleri; tane verimiyle fiziksel özellikleri arasındaki modellerin güvenirliliği daha yüksek saptanmıştır.
In this study, the distribution of some physical, chemical soil properties and some agronomic characteristics of crops (plant height, 1000 seed weight and grain yield) of agricultural lands of Çarşamba Plain where wheat, corn and soybean plants were grown; the determination of correlation relationships between agronomic characteristics of plants and soil properties and the setting regression models between agronomic and different soil characteristics based on these correlations; the applicability of obtained models in estimating plant yield in district soil were investigated. For this purpose, soil and plant samples were taken from agricultural land cultivated by farmers in sixty villages of Çarşamba Plain in Samsun. In accordance with this purpose, some physical (texture, field capacity, wilting point, bulk density) and chemical properties of soils (organic matter, soil reaction, electrical conductivity, lime content, total nitrogen, exchangeable cations, available phosphorus, cation exchange capacity, available Fe, Cu, Zn and Mn) and agronomic characteristics of plants were determined. The study was carried out between 2013-2014 on the same land with the same plants. When the correlation relationships between wheat and soybean plants were compared, the correlations between physical and chemical properties of corn grown soils and agronomic characteristics were found to be statistically more significant. In the regression model between wheat plant height and soil properties as EC, Ca+Mg, Db, Clay, (Db)2, (EC)2, (Clay)2,√(Ca+Mg) soil properties, the highest regression coefficient was determined (R=0.731) at the level of statistical significance (p=0.035). In the regression model between 1000 seed weight of wheat and (EC)2, (OM)2, (Fe)2, (Clay)2, (Db)2, (Clay×Db),√Clay ,√Db , WP parameters, the highest regression coefficient was found to be R=0.794 at p=0.013 the level of satistical significance. Regression model between wheat grain yield and EC, CaCO3, Sand, (Sand×Db), (Db×WP), Fe, N, (Db)2, WP, (EC)2, √Db parameters was statistically significant (p=0.012 <0.05) and had the highest regression coefficient (R=0.840). Regression coefficient (R=0.543) was at medium level in relations between corn plant height and (CaCO3)2, (EC)2, (OM)2, Db, FC,√N , Ca+Mg, CaCO3, OM,√Kum parameters and statistically insignificant (p>0.10). When compared whit the other models, regression model between 1000 seed weight of corn and pH, CaCO3, EC, OM, √EC,√(CaCO_3 ) , Sand, (Db×Sand), (Db)2 and WP parameters was statistically significant (p=0.012) and had the highest regression coefficient (R=0.819). Relationship between corn grain yield and OM, N, P, K, Ca+Mg, Na, Zn, WP, Db, silt, (Db×WP) parameters showed high level of statistical significance (p= 0.001) and regression coefficient (R=0.894) was found to be very high. Regression model formed between soybean plant height and soil properties as (EC)2, (OM)2, N, Cu, K, Db, Clay, (Db×Clay), WP, AW,√Db was statistically significant and regression coefficient was high (R=0.766). The highest regression coefficient (R=0.782) was obtained in regression model between 1000 seed weight of soybean and (EC)2, (CaCO3)2, N, P, K, Na, (OM)2, (EC)2,√Db , (Db×Sand), WP, FC,√Sand , silt parameters, when compared with the other models. Regression model between soybean grain yield and EC, CaCO3, OM, N, P, K, (OM)2, (EC)2, (CaCO3)2, Zn, Mn,√(N×P×K) parameters showed statistically significance (p<0.001) and regression coefficient was determined to be very high (R=0.921). In determining the validity of models, root mean square error (RMSE), index of agreement (d), model efficiency (ME), mean absolute error (MAE), mean bias error (MBE), mean relative error (MRE) parameters were evaluated together. As a result, it seems possible that the models formed between agronomic characteristics of wheat, corn and soybean plants and soil properties investigated may be applied. The efficiency of models obtained between wheat plant height and chemical-physical soil properties; wheat grain yield physical and chemical properties; 1000 seed weight of wheat and chemical soil properties was at optimal level and it is possible to use these models in predicting agronomic characteristics. Models formed between corn plant height and physical and physical-chemical soil properties respectively, 1000 seed weight of corn and physical-chemical soil properties; grain yield of corn and physical and chemical soil properties respectively were more effective and reliable. In regression models of soybean plant, reliability of models between soya plant height and chemical and physical-chemical soil properties respectively; 1000 seed weight and chemical soil properties; yield and physical properties was determined higher.
In this study, the distribution of some physical, chemical soil properties and some agronomic characteristics of crops (plant height, 1000 seed weight and grain yield) of agricultural lands of Çarşamba Plain where wheat, corn and soybean plants were grown; the determination of correlation relationships between agronomic characteristics of plants and soil properties and the setting regression models between agronomic and different soil characteristics based on these correlations; the applicability of obtained models in estimating plant yield in district soil were investigated. For this purpose, soil and plant samples were taken from agricultural land cultivated by farmers in sixty villages of Çarşamba Plain in Samsun. In accordance with this purpose, some physical (texture, field capacity, wilting point, bulk density) and chemical properties of soils (organic matter, soil reaction, electrical conductivity, lime content, total nitrogen, exchangeable cations, available phosphorus, cation exchange capacity, available Fe, Cu, Zn and Mn) and agronomic characteristics of plants were determined. The study was carried out between 2013-2014 on the same land with the same plants. When the correlation relationships between wheat and soybean plants were compared, the correlations between physical and chemical properties of corn grown soils and agronomic characteristics were found to be statistically more significant. In the regression model between wheat plant height and soil properties as EC, Ca+Mg, Db, Clay, (Db)2, (EC)2, (Clay)2,√(Ca+Mg) soil properties, the highest regression coefficient was determined (R=0.731) at the level of statistical significance (p=0.035). In the regression model between 1000 seed weight of wheat and (EC)2, (OM)2, (Fe)2, (Clay)2, (Db)2, (Clay×Db),√Clay ,√Db , WP parameters, the highest regression coefficient was found to be R=0.794 at p=0.013 the level of satistical significance. Regression model between wheat grain yield and EC, CaCO3, Sand, (Sand×Db), (Db×WP), Fe, N, (Db)2, WP, (EC)2, √Db parameters was statistically significant (p=0.012 <0.05) and had the highest regression coefficient (R=0.840). Regression coefficient (R=0.543) was at medium level in relations between corn plant height and (CaCO3)2, (EC)2, (OM)2, Db, FC,√N , Ca+Mg, CaCO3, OM,√Kum parameters and statistically insignificant (p>0.10). When compared whit the other models, regression model between 1000 seed weight of corn and pH, CaCO3, EC, OM, √EC,√(CaCO_3 ) , Sand, (Db×Sand), (Db)2 and WP parameters was statistically significant (p=0.012) and had the highest regression coefficient (R=0.819). Relationship between corn grain yield and OM, N, P, K, Ca+Mg, Na, Zn, WP, Db, silt, (Db×WP) parameters showed high level of statistical significance (p= 0.001) and regression coefficient (R=0.894) was found to be very high. Regression model formed between soybean plant height and soil properties as (EC)2, (OM)2, N, Cu, K, Db, Clay, (Db×Clay), WP, AW,√Db was statistically significant and regression coefficient was high (R=0.766). The highest regression coefficient (R=0.782) was obtained in regression model between 1000 seed weight of soybean and (EC)2, (CaCO3)2, N, P, K, Na, (OM)2, (EC)2,√Db , (Db×Sand), WP, FC,√Sand , silt parameters, when compared with the other models. Regression model between soybean grain yield and EC, CaCO3, OM, N, P, K, (OM)2, (EC)2, (CaCO3)2, Zn, Mn,√(N×P×K) parameters showed statistically significance (p<0.001) and regression coefficient was determined to be very high (R=0.921). In determining the validity of models, root mean square error (RMSE), index of agreement (d), model efficiency (ME), mean absolute error (MAE), mean bias error (MBE), mean relative error (MRE) parameters were evaluated together. As a result, it seems possible that the models formed between agronomic characteristics of wheat, corn and soybean plants and soil properties investigated may be applied. The efficiency of models obtained between wheat plant height and chemical-physical soil properties; wheat grain yield physical and chemical properties; 1000 seed weight of wheat and chemical soil properties was at optimal level and it is possible to use these models in predicting agronomic characteristics. Models formed between corn plant height and physical and physical-chemical soil properties respectively, 1000 seed weight of corn and physical-chemical soil properties; grain yield of corn and physical and chemical soil properties respectively were more effective and reliable. In regression models of soybean plant, reliability of models between soya plant height and chemical and physical-chemical soil properties respectively; 1000 seed weight and chemical soil properties; yield and physical properties was determined higher.
Description
Tez (doktora) -- Ondokuz Mayıs Üniversitesi, 2018
Libra Kayıt No: 124205
Libra Kayıt No: 124205
Citation
WoS Q
Scopus Q
Source
Volume
Issue
Start Page
End Page
203
