Exploratory analysis of permanent manurial trials in rice

Abstract

Permanent Manurial Trials are experiments performed at fixed sites in a long run to assess the role of nutrient sources in catalyzing crop production. The experimental data from Permanent Manurial Trial in rice at RARS, Pattambi was explored in a view to identify suitable set of fertilizer treatments in optimizing grain yield.The experiment was commenced in 1973 under Kerala Agricultural University State Plan Project and is being conducted continually over years in two cropping seasons viz., Kharif and Rabi. Eight fertilizer treatments under test were either organic, inorganic or combination thereof. As treatment responses can be deliberately measured through yield, grain yield data maintained at research station for the period 1973 to 2015 formed the basis of study. Treatment-wise rice yield data was subjected to various statistical analyses, to screen superior fertilizer treatments. The influence of weather on yield responses were also assessed. Exploratory Data Analysis using graphical and non-graphical methods as a means to familiarize yield data was attempted to realize yielding behavior of treatments. Summary statistics viz., mean and median explicitly showed the superiority of treatment T1 ( Cattle manure at 18000 kg ha -1 to supply 90 kg N ha-1) followed by treatment T5 (Cattle manure at 9000 kg ha-1+ ammonium sulphate to supply 45 kg N ha-1+ superphosphate to supply 45 kg P2O5 ha-1 +45 kg K2O ha-1 as muriate of potash). Box plot of yield data remarked consistent yielding performance for the same treatments during both Kharif and Rabi seasons. Yield trend was assessed by regressing crop yield on time factor and no significant yield trend was observed. Analysis of variance was done for each of the experiments during both seasons and the post hoc test was effected through DMRT. All the experimented seasons showed significant fertilizer effect on grain yield. The treatments subgroups as formed through DMRT for each year was further used to score treatments. Lowest scores speaking of better treatment performance was recorded for T1 in both the seasons followed by T5. Analysis of groups of experiments was further necessitated to generalize on the yielding behavior of each treatment as results followed from analysis of variance for individual experiments do not give a confirmatory account in this regard; owing to its inclusion of seasonal fluctuations (season/ year) in experimental error. The results showed superiority of treatment T1 and T5 during both seasons. Influence of weather on grain yield was studied for three growth stages of rice viz., early tillering to panicle initiation, panicle initiation to flowering and flowering to milk stage. The results followed from correlation and regression analysis pointed to minimal role of weather in affecting grain yield, attributed to the crop management practices followed in the research station. Regression models for treatment-wise yield on weather indices formulated through principal component analysis were found poor fit as observed through R2 value. Consistency of treatments were assessed through two consistency measures and treatments T1 and T5 were observed to have average consistency. Sustainability Yield Index (SYI) index was further used to identify treatments that have yielded in a sustainable manner in comparison with the maximum yield reckoned for each of the treatments over a longer period. Maximum SYI was recorded for treatments T1 followed by T5 during both seasons. Combinatorial treatments of organic and inorganic fertilizers had higher index values compared to purely inorganic treatments. Statistically optimum fertilizer requirement was obtained by compiling the results followed from the aforementioned analyses. Treatment T1 (Cattle manure at 18000 kg ha ammonium -1 to supply 90 kg N ha-1) and T5 (Cattle manure at 9000 kg ha-1+ sulphate to supply 45 kg N ha-1 + superphosphate to supply 45 kg P2O5 ha-1 + 45 kg K2O ha-1 as muriate of potash) were chosen as the optimal fertilizer schedules for rice.