Fri, 19 May 2017 10:14:15 +0000
Crop production is affected by a number of factors some of which are environmental factors. For example, weather influences crop growth and development. Variability in the crop production arises from differences in soil properties, crop agronomic management practices such as planting, fertiliser application, irrigation, tillage, etc. can be used to reduce the negative impact of weather.
Therefore, crop forecasting is an important tool for optimising crop yield but requires extensive use of statistical techniques as is demonstrated in this article.
Crop Forecast Survey (CFS)
A crop forecast is a statement of the estimated amount of yield or production on the basis of known facts at a given time. Last week, the Hon. Minister of Agriculture announced the Crop Forecast Survey (CFS) results. I wish therefore wish to deal with statistical applications in the conduct of the CFS. I am aware that some people have criticised and may continue criticising the use of CFS. It is important to understand what is statistical issues are involved in the CFS. The Hon, Ministers statement estimated crop production for the 2016/2017 agricultural season as well as the country’s national food balance for the marketing season covering the period 1st May 2017 to 30th April 2018. The crop production estimates that are announced are based on a universally applied scientific survey method that is used every year. The survey is jointly conducted by the Ministry of Agriculture and the Central Statistical Office (CSO) and covers all the districts of the country. The survey targets a number of food and cash crops such as Maize, Cassava, Sorghum, Rice, Millet, Sunflower, Groundnuts, Soya beans, Cotton, Irish potato, Virginia Tobacco, Burley tobacco, Mixed beans, Bambara nuts, Cowpeas, Velvet beans, Sweet potatoes, Paprika, Pineapples, Wheat, Barley, Popcorn, Sugarcane. The survey covers a number of variable such as Area planted (ha), Area expected to be harvested (ha), Expected production (MT), Yield (MT), Expected total sales (MT), Quantity of basal and top fertilisers used (MT). It is in the interests of the members of the public, farming community and the economy as whole that crop forecast surveys are conducted and results disseminated as was done last week.
The crop forecast survey also includes productivity statistics for each crop. For example, in terms of maize, 2,852,687 metric tonnes were produced in the 2011/2012 farming season and the average yield was 2.24 metric tonnes per hectare. In the 2013/14 farming season, 3,103,483 metric tonnes of were produced and the average yield was 2.26 per hectare. In the 2016/17 farming season, 3,606,549 metric tonnes have been forecasted and the average yield is estimated as 2.19 metric tonnes per hectare. Based on these three farming seasons, it can be concluded that the yield per hectare in 2013/14 farming season was better than the 2016/17 in spite of the total production having been lower. Such a conclusion is made possible with the help of statistics.
Maize production has been forecasted to increase to 3,606,549 metric tonnes from 2,873,052 metric tonnes in the 2015/2016 season, an increase of 25.53 percent. Small and medium scale farmers have recorded an average maize yield rate of 2.12 metric tonnes per hectare, whilst large scale farmers have recorded an average maize yield rate of 5.24 metric tonnes per hectare. This shows that large scale farmers are getting a higher yield per hectare than our small-scale and medium scale farmers. This is an area that requires great improvement.
The Hon. Minister of Agriculture reported that the contributing factors to the high production of maize are increased area planted, favourable agro meteorological conditions increase in the usage of fertiliser on maize by farmers which increased by 26.42 percent in 2016/2017 compared to last season. The area under maize cultivation increased by 20.5 percent to 1,644,741 hectares in the 2016/2017 season from 1,364,977 hectares last season. The area expected to be harvested also increased by 23.86 per cent to 1,433,944 metric tonnes from 1,157,755 metric tonnes. All these increases a determined using statistics.
Methods of Crop Forecast
There are several methods of yield forecasting. The following are the types of forecasting methods
i.e. qualitative methods – judgmental methods (Forecasts generated subjectively by the forecaster, Educated guesses) and quantitative methods (Forecasts generated through mathematical modeling). Observations and measurements are made throughout the crop growing season, on the above stated factors together with estimates of percentage of damage from pests (such as army worms) and fungi. Experts also include estimates of the percentage of weeds infestation, and so on in estimating crop production. Based on the data obtained, yield can be forecasted using regression methods, or by the knowledge from local expertise. Other two methods used to forecast crop yield are the use of remote sensing and- crop simulation models. The objective of the yield forecast is to give a precise, scientific sound and independent forecasts of crops’ yield as early as possible during the crops’ growing season by considering the effect of the weather and climate. The differences between forecasts and final estimates are in the timing of the release. Forecasts are made before the entire crop has been harvested whereas estimates are made after the crop has been harvested.
It has to be noted that historically, farmers have been always been making “forecasts” in order to plan their agronomic practices. Forecasting crop yield means also knowing or forecasting other important parameters. For example, quantifying the area planted at the starting of the growing season and quantifying the area harvested.
Traditional Crop
Forecasting
It has to be noted that crop forecasting has been and is still being done using indigenous methods that are not necessarily documented but are simply being passed on from one generation to another. It is encouraging to see that in Zambia an Act of Parliament was enacted known as the Protection of Traditional Knowledge, Genetic Resources and Expressions of Folklore Act No. 16 of 2016. The Act provides for a transparent legal framework for the protection of, access to, and use of, traditional knowledge, genetic resources and expressions of folklore, which also guarantees equitable sharing of benefits and effective participation of holders; to recognise the spiritual, cultural, social, political and economic value of traditional knowledge, genetic resources and expressions of folklore of holders. Clearly, the Act recognises the value of traditional and indigenous knowledge. When and if properly implemented, I am confident that this Act will assist in strengthening our indigenous knowledge system in Zambia.
The philosophical underpinning in all indigenous forecasts has a comfortable home in statistics. In spite of many of our members not being formally schooled in statistics, they are able to use a number of statistical concepts such as trend analysis, time series analysis, seasonality, random or irregular events without calling them by these names given by statisticians.
Sometimes the quantities of fruits available can be used to make some predictions in weather and crop forecast. For example, abundance of masuku fruits in some communities is a perfect prediction of above normal rainfall in the coming season and hence higher likelihood of good yield. With such knowledge communities have prepared themselves for calamities long before weather experts communicate to them. Crude as they may be, they give some reasonable forecasts that can be confirmed using modern ways of making weather forecasts by experts. Mundy and Compton (1991) state that to-date, modern science has not come up with a conclusive stance for or against the claims of indigenous weather forecasting although some believe that modern science could gain valuable insights from indigenous knowledge.
Historically and to date indigenous communities in different parts of Zambia have continued to rely on indigenous knowledge to forecast crops, conserve the environment and deal with natural disasters. The communities particularly those in drought and flood prone areas have generated a vast body of indigenous knowledge on disaster prevention, management and mitigation through early warning and preparedness systems. This knowledge can be harnessed as it still plays a key role in our communities. These indigenous forecasting techniques have further underscored that fact that access climate information is fundamental to understanding the climate risks as well as identifying and assessing the viability of adaptation options. It has also been recognised that, along with information on climate risks, knowledge of the impacts of exposure to those risks can be instrumental in motivating communities and organisation to begin the process of adapting.
Indigenous knowledge of seasonal weather forecasting could be useful in decision making at village level to best exploit the seasonal distribution of rainfall in order to increase or stabilise crop yields. The presence of locusts is used as predictor of poor crop harvest. Equally, occurrence of more grasshoppers in a particular year indicates less rainfall and hunger but appearance of army worms around October signifies abundant rainfall in the upcoming season. Disposition of the new moon indicates more disease outbreaks and erratic rainfall. Indigenous forecasting is mainly based on relative experience acquired by elders. These experiences are part of the data they collect and analyse with any formal statistical analysis but simply based on experiences.
I believe that the Protection of Traditional Knowledge, Genetic Resources and Expressions of Folklore Act No. 16 of 2016 gives us an opportunity to harness our indigenous crop forecasting techniques for the greater benefit of our communities if and when appropriately implemented. Indigenous knowledge systems are used by small scale rural farmers in crop forecasting. They mainly make use of observations of the crops, area planted
The National Food
Balance Sheet (NFBS)
The National Food Balance Sheet (NFBS) is one of the most important planning tools used at beginning of marketing season. It provides market outlook in terms of supply and demand of staple food crops and forms the basis for public and private sector planning for agriculture marketing. The CFS enables determination of the NFBS. It is a key ingredient in compilation of NFBS. According to the forecast, there is an expectation of a total of 4,175,866 to be available stock compared to total requirement of 2,997,350. This shows a maize surplus of 1,178,516 but deficits of 40,000 for paddy rice and 146,765 for wheat. Clearly, these estimates can be used for planning and policy measures taken by the Government such as export ban on maize and maize products, the Food Reserve Agency (FRA) to purchase up to 500,000 metric tonnes of commodities including maize for strategic food reserves.`
Availability (A) take into account the Opening stock as at 1st May 2017 and the forecasted Total production. In terms of the maize Requirements (B), it takes into account the Human consumption including the strategic reserve stocks and Industrial requirements (stock feed and breweries), Losses, and Structural cross-border. These statistics are then used to determine the surplus/deficit in the simple equation: Surplus/deficit = A – B.
Conclusion
Crop focus survey is an import tool for planning both at individual and national level. The National Food Balance Sheet (NFBS) can only be produced if we have statistically valid national estimates of production and opening stocks. Qualitative monitoring systems will only be able to generate qualitative assessments but not quantitative estimates. Qualitative crop monitoring systems or assessments will never produce reliable quantitative estimate for compiling the NFBS. It is almost certain that, currently, there is no substitute to a survey that is based on a statistically valid random sample for generating crop forecast estimates. Purpose of Crop Forecast is to collect information on anticipated area, production, and sales of major crops. This information is used to assess expected food security situation at national level, assess performance of major cash crops, facilitate crop production trend analysis, and compute agriculture’s contribution to Gross Domestic Product (GDP). As a result of the reported crop forecast results disseminated, join me next week as I look at the use of statistics in crop storage and marketing.
