Artificial intelligence techniques play a promising role in designing decision-making systems in agriculture. Several artificial intelligence techniques have been adopted to design various agricultural decision support systems worldwide during the past three and a half decades. Over time, the application trend of such techniques has been changed and opened several new frontiers. A scientific study can lead aspirant researchers to choose an appropriate technique before designing such systems in agriculture. Instead of a mere literature survey, a statistical method of trend analysis and forecasting is essential to explore the past, present, and future trajectory of applying such techniques. This paper presents a statistical framework to analyze and forecast the application trend of the five most popular intelligent techniques in agriculture using the time-series data of the past 35 years. In the first step, as a well-established nonparametric method of trend test, the Mann-Kendall statistics have been applied to assess the existence of a trend. In the second step, a forecasting model has been proposed using the Autoregressive Integrated Moving Average approach to predict the future trend and prospect of applications of such systems with precision. Finally, the prediction accuracy has been measured using three popular scaled error metrics.