A3.4 Yield measurements of sorghum and millet plots

Contents - Previous

The method used must be appropriate for sorghum and millet trials. In contrast to trials on legumes, most sorghum and millet trials focus on production practices. A single variety with a relatively uniform maturity is planted. This means that yield usually can be measured with a single cutting. Plots for trials on tillage-planting and water conservation tend to be large, which increases the size of the harvest. The method also needs to be quick, because sorghum and millet plots tend to be numerous in on-farm research.

All sorghum and millet trials are not the same. Some consist of a large number of plots in a field, whereas others include only two or three plots, The participation of farmers is generally more in simple experiments than in those with a larger number of plots in the field.

A3.4.1 Systematic Quadrat Sub-Sampling for Sorghum and Millet

The procedure, which can be used for both broadcast and row-planted plots, can be used to make fairly accurate estimates of grain yield during a single visit by researchers to the field. This procedure is especially useful for production systems in which harvests are typically carried out in stages. This procedure is remarkably accurate in the Botswana production situation because much more of the variability in sorghum and millet grain yields is due to head numbers per hectare than to average weight per head. The procedure is as follows:

• Sub-Sampling. Sub-sample the plot with a 2m by 2m quadrat. Take 20 quadrat sub samples in a systematic pattern. Usually, this pattern consists of two passes through the plot with 10 equally spaced sub-samples. If the plot is row planted, each pass follows rows near, but not on, the border. In such a case, select rows that are one-fourth of the plot width from the border on both sides of the plot. Preferably, three people work: two carry the quadrat, and one carries the sample bags and records data. The front carrier follows a sampling line and the rear carrier counts paces. The quadrat is placed on the ground, without bias, at the toe of the rear carrier. For row planting, record the number of rows in each sub-sample and average row spacing in each plot.
• Data and Sample Collection. In each sub-sample, counts are made on the number of plants and number of grain heads. Heads are recorded as:

HR -- Heads already harvested before measurement.
MT -- Heads ready for harvest.
GR -- Heads that are green. These can reach maturity.
GZ -- Heads that are completely missing due to livestock feeding.

At this point, there are two options:

• Option 1: Harvest the MT heads and put them in large, brown kraft paper bags. Labelling includes field, plot, and quadrat number. Close and staple bags. Pack sample bags from plot in a large burlap bag.
• Option 2: Harvest the MT heads and bulk all sub-samples from the plot in a large burlap bag. This is the option now recommended.

• Drying, Threshing, and Weighing Samples. Which approach to use here is based on the options used above:

• Option 1: Oven dry or sun dry the sub-samples. For sun drying, the heads should be removed from the bag. For this option, a protected space is required, reached by the sun, in which to spread and dry the samples. Alternatively, heads may be left in the bags to dry, but the bags must be opened fully. When the samples are left in the bags, drying will take a little longer, but it reduces the danger of samples getting mixed up or lost, When samples reach a uniform threshing moisture, the heads are counted and threshed, grain weighed, and the data recorded on the harvest data sheet, A labelled bag is used to hold a sample throughout the final weighing process.
• Option 2: If the grain is wet, it is first dried, and then bulked sub-samples are threshed and weighed at the farm,

Additional points to note are:

• If a plot has been harvested already, FSD staff still can collect a sample of 30 heads that the farmer indicates came from the plot, Check carefully with the farmer about this, When the plot is harvested, a rough estimate of yield can be made by using this head sample and HR counts in the field.
• Other measurements can be combined with quadrat sampling, such as plant height, visual gauging of average height of productive plants to the base of the main panicle, etc.
• Systematic sub-sampling appears complicated, but experience has shown that research technicians and casual labourers can routinely and correctly handle the method, A training session for technicians improves the quality of sub sampling, When farmers harvest trials, sub-sampling should not be used.
• Sub-sampling is appropriate for RMRI or RMFI trials where the number of plots per field is greatest, With this method, the researcher can gather in-depth agronomic information, such as sorghum tiller numbers and head size in a RMFI manure trial, which is needed for some comparisons with station research. Sub sampling can estimate more precisely the effect of treatments than whole-plot harvest, This is because:

-- Plot size and shape are not regular. Inaccurate estimates of plot size affect whole-plot measurements, but not quadrat measurements.
-- Non-treatment causes often mask treatment effects in portions of plots. Such visible conditions as ploughing dead furrows, plot borders, water washouts, ant hills, and trees can be avoided by sub-sampling.

• Harvesting by research staff can have a negative effect on the farmer's view of the trials. To help with implementation and assessment of experimental technology, researchers should strive to pass ownership of farmer-implemented trials to farmers, Farmers gain a greater sense of ownership if they harvest their own trials,

• Calculate Yield. Grain yield estimates, tiller numbers, and head weights are most easily calculated by entering the data in dBase 3 Plus database package and using the program that is described in Norman and Siebert [1990], Alternatively, grain yield can be calculated by the following formula:

Average head weight (kg) = total weight in sample of heads taken (kg) / number of heads

Total potential grain yield (kg/ha) = (av. head weight x av. number HR heads) + (av. head weight x av. number MT heads) + (av. head weight x av. number GZ heads)

'Potential' is indicated because the GZ heads have been eaten by animals and are not available to the farmer for harvest.

Total grain yield for farmer (kg/ha) = (av. head weight x av. number HR heads) + (av. head weight x av. number MT heads)

Note: An estimate of harvest from heads that are still green can be added if the visit is relatively early and the researcher and farmer believe that the green heads will develop and be harvested at a later date,

Grain yield from green heads (kg/ha) = (av. head weight x av. number GR heads)

For example, FSD staff find several large plots in which about half of the sorghum is ready for harvest but still in the field, The farmer has harvested the earliest heads to prevent damage from wild animals. Using the systematic quadrat subsampling method (see Section A3.4.1), FSD staff count on an average of 5,8 plants/quadrat, averages of 8.5 MT heads/quadrat in the field, 2.2 HR stems/quadrat indicating heads that have been harvested, 3.7 GZ stems/quadrat showing signs of animal grazing of heads, and 5.5 GR heads/quadrat that should still mature and be harvested. The FSD staff harvests the MT heads from the subsamples and collects a sample of HR heads from the threshing floor as directed by the farmer. The MT heads weigh an average of 0.032 kg and the HR heads an average of 0.015 kg. Note that the HR head weight is used for GZ or previously grazed heads, as well. Each quadrat sub-sample is 2m by 2m,

Harvest plant stand (plants/ha.) = (5.8 x 10000)/4 = 14,500 plants/ha

The MT harvest (kg/ha) = ((8,5 x 10000)/4) x 0,032) = 680 kg./ha

The HR harvest (kg/ha) = ((2.2 x 10000)/4) x 0.015) = 83 kg/ha

The GR harvest (kg/ha) = ((5.5 x 10000)/4) x 0.032) = 440 kg/ha

The GZ harvest (kg/ha) = ((3,7 x 10000)/4) x 0.015) = 139 kg/ha

The actual harvest to the farmer (kg/ha) = MT + HR + GR = 1,203 kg/ha

If it is discovered later, for whatever reason, that the field was abandoned before the GR heads matured, the actual harvest = MT + HR = 763 kg/ha.

The harvest that was lost to uncontrolled animals = GZ = 139 kg/ha,

A3.4.2 Whole-Plot Harvest for Sorghum and Millet

• Measure Size and Shape of Plot. Re-measure at the end of the season for RMRI and RMFI trials, For FMFI trials, the researcher may decide that pegging from the beginning of the season is accurate enough.
• Row Planting. Measure average row spacing, if the plot is row planted. Average row spacing equals the width of the plot in metres divided by the number of rows in the plot.
• Plot Boundaries. Where they are not clear, plot boundaries need to be marked, In large plots, tall sighting poles help greatly in marking boundaries, In small- and intermediate sized plots, string lines can be used to mark boundaries. Because the need for precise marking is less in FMFI trials, decisions about boundaries usually can be left to the host farmer.
• Farmer Harvesting. Discuss the needs of the trial with the farmer if he/she will be harvesting alone, When research staff are helping with the harvest, discuss the schedule with the farmer, Arrange logistics, such as where the harvest from the plots will be stacked for drying, If possible, researcher or field staff should visit the field on the day when the farmer starts harvesting. At that time, they should label the harvest bags and agree on which bags will be used for each plot.
• Harvest Criteria. Instruct harvesters on the criteria for harvesting, such as whether to harvest green millet heads, heads with smut, and so on.
• Single Cutting. Usually plan to harvest a plot in a single cutting. If more than one cutting is planned, adjust the harvest criteria appropriately.
• Threshing and Weighing. After drying, thresh and weigh grain from each plot. The farmer is asked to transport the harvest of each plot to the threshing floor, thresh the grain, and put it in a labelled bag. Researchers provide the farmer with bags, Identification on the bag must be simple and clear, Research staff will schedule a visit to weigh the threshed harvest from each plot,
• Calculate Yield. This is done as follows:

Grain yield (kg/ha) = (plot yield (kg) x 10,000)/plot size in square metres,

For example, if 18 kilograms are harvested and the plot size is 10m by 25m metres, then:

Grain yield (kg/ha) = (18 kg x 10,000)/(10 metres x 25 metres) = 720 kg/ha,

Additional points to note are:

• Whole-plot harvesting is easy, inexpensive, and appropriate for FMFI trials. That is why this method is recommended when farmers harvest trials alone or together with research staff. Because farmers are helping with harvesting, a larger number of research sites can be managed, If farmers work with the harvest, they also are more likely to feel they own the trial.
• The additional work required to make the whole-plot method precise enough for some studies can be tedious, Plot size and shape must be measured accurately. Sometimes, portions of the plot affected by causes from outside the trial need to be subtracted from the plot area before calculating yield.
• This method is most suitable with small- and intermediate-sized plots, Keeping many harvest samples of large plots separate is difficult, and the chance of a mix-up increases. Using this method, the researcher cannot easily collect the data to make in-depth agronomic analyses. Information on tiller number and harvest plant stands would need separate sampling and measurement.

A3.4.3 Row-Segment Measurement for Sorghum and Millet

• Sub-Sampling. Sub-sampling and data collection are similar to those of the systematic quadrat sub-sampling method (see Section A3.4.1). The main difference is the use of a measuring stick instead of a square quadrat.
• Sample Collection. Cutting, drying, threshing, and weighing samples are the same as for Option 2 of systematic quadrat sub-sampling (see Section A3.4.1).
• Calculate Yield. Grain yield estimates can be calculated by using the following equation:

Grain yield (kg/ha) = [(sample Yield per plot (kg) x 10,000)] / [(number sub-samples x segment length (metres) x row spacing (metres)]

For example, if 2 kilograms are harvested from 20 measured segments in a plot, each segment is 2 metres long, and average row spacing is 0.75 metres then:

Grain yield (kg/ha) = (2 x 10,000)/(20 x 2 x 0.75) = 667 kg/ha.

Note the following additional points:

• Because this method is similar to the systematic quadrat sub- sampling method, most comments given in that section (Section A3.4.1) pertain here as well. Tiller number and head weights also can be calculated,
• The use of a measuring stick instead of a quadrat has three effects:

-- The stick is more manageable, so work can be done by two technicians instead of three.
-- The stick is easy to transport -- a big plus for staff travelling on bicycle or foot.
-- However, the quadrat is more decisive than the stick about defining the boundary of the sub-sample. Harvest data collected with quadrats are generally more reliable than data collected with measuring sticks.

A3.5 Yield measurements of legume plots

A3.5.1 Whole-Plot Approach

Several reasons why whole-plot harvesting is used to measure grain yield in onfarm legume trials are:

• A major part of on-farm work with cowpea, mung bean, and groundnut has been testing varieties and other crop comparisons. In these trials, plots can be distinguished readily by plant or seed type,
• Legume plots are not big; trials include only a few plots per farm and tend to be FMFI. Farmers can readily manage the whole-plot harvest of this type of trial.
• Most importantly, cowpea, mung bean, and tepary bean have an indeterminate growth habit -- at least partially -- so that more than one harvest cutting is needed. For research staff to sub-sample a plot repeatedly is not practical,
• Groundnut plots do not require repeated sampling. However, these plots are small, and grain yield is more easily measured if farmers harvest whole trial plots.

A3.5.2 Systematic Quadrat Sub-Sampling for Legume Leaf Spinach

Systematic quadrat sub-sampling is used for measuring leaf spinach yield for crops such as cowpea. The procedure is as follows:

• Sub-Sampling. The sub-sampling pattern is similar to that for sorghum and millet (see Section A3.4.1).
• Data and Sample Collection. In each sub-sample, the number of plants are counted and their leaves are harvested. Take all the leaves to measure the potential harvest. Alternatively, consult with the farmer and choose the intensity of defoliation that is typical. Bulk or combine the sub-samples from each plot.
• Prepare Spinach and Weight. Ask the farmer to prepare dried spinach and weigh.

• Option 1: Weigh the plot samples while they are still fresh. Take a small sample (0.5 kilograms) from the bulked sample, Give the rest of the bulked material to the farmer to use as he/she pleases. Ask the farmer to prepare the dried spinach from the small sample, Weigh the small samples again when ready as spinach,
• Option 2: Ask the farmer to prepare dried spinach from the bulked sample for each plot, Weigh it when it is ready.

• Calculate Yield. Depending on the option, the procedure is:

• Option 1:

Spinach weight per 100 square metres = [bulked fresh sample (kg) x (small sample dried (kg)/small sample fresh (kg)) x 100] / plot size (square metres)

For example, if 7.4 kilograms are harvested from a 20m by 10m plot. The 0.5 kilogram small sample, when prepared as spinach, weighs 0.15 kilogram, then:

Spinach weight per 100 square metres = (7.4 x 100 x (0.15/0.5))/(20 x 10) = 1.11 kg.

• Option 2:

Dried spinach weight per 100 square metres = [(prepared spinach weight (kg) x 100)] / plot size (square metres)

For example, if 0.95 kilogram spinach is prepared from leaf harvested from a 15m by 12m plot, then:

Dried spinach weight per 100 square metres = (0.95 x 100)/(15 x 12) = 0.53 kg.

Note: Because leaf sampling might affect grain yield, the researcher should either harvest leaf samples on parts of the plot where grain yield will not be measured or harvest leaf samples equally on all treatments of the legume trial. If leaf sampling is only on some treatments, the others will be favoured, and this might bias and invalidate the grain legume yield comparison of the trial.

A3.5.3 Whole-Plot Harvest for Legume Grain.

• Whole-Plot Harvest Procedure. Whole-plot harvest for legumes is similar to that for sorghum and millet where farmers harvest the grain (see Section A3.4.2).
• Cuttings. The indeterminacy of their growth will require more than one cutting to measure grain yield for cowpea, mung bean, and tepary bean. There are two approaches to dealing with cuttings:

• Option 1: Each cutting is kept separate, If the researcher wants to compare legume treatments for each cutting, the schedule of harvesting by the farmer and weighing by research staff must be organized well. To compare correctly, farmers should harvest each treatment at each cutting, even if the yield is still small.
• Option 2: The cuttings are combined, For most trials, researchers can bulk the different cuttings to get one yield per plot, Even here, farmers often would like to use part of the harvest before the final harvest has been completed. To help the farmers, research staff can organize a schedule of visits to check the harvesting procedure and weigh grain harvests to that date.

For example, a farmer makes three harvests on each plot of a trial involving cowpeas. For one plot, the harvest yields were 5.2 kg, 3.1 kg, and 0 kg. For a second plot, harvest yields were 0.1 kg, 2.2 kg, and 12.1 kg. Upon careful measurement, FSD staff found the first plot to be 12m by 15.5m and the second plot to be 10.5m by 12.0m.

Grain yield of plot one (kg/plot) = 8.3 kg/plot

When adjusted for actual plot size = (5.2 + 3.1 + 0)/(12 x 15.5) = 0.0446 kg/square metre or = 446 kg/ha

Grain yield of plot two (kg/plot) = 14.4 kg/plot

When adjusted for actual plot size = (0.1 + 2.2 + 12.1)/(10.5 x 12.0) = 1.142 kg/square metre, or= 1,142 kg/ha

A3.5.4 Whole-Plot Harvest for Groundnuts

Groundnut whole-plot harvest is similar to that for sorghum and millet, where farmers harvest the grain in one cutting (see Section A3.4.2). For groundnuts, grain yield for the shelled (i.e., shells removed) harvest is reported. A shelling percentage estimate is one of the yield quality measures for groundnut.

To obtain the groundnut shelling percentage, a small sample (e.g., 0.5 kilograms) of unshelled nuts that have been dried for shelling is weighed. Then shell and weigh the shelled sample:

Shelling percentage = (100 x weight shelled (kg)) / weight unshelled (kg)

Shelling percentage can be used to convert unshelled harvest weights to estimates of shelled weights. To get an estimate of yield when a farmer has not yet shelled his harvest: weigh the unshelled harvest, weigh a small shelling percentage sub-sample (0.5 kilograms), shell, and then reweigh the shelled sub-sample:

Shelled harvest (kg) = [unshelled harvest (kg) x shelling percentage] / 100

For example, a farmer has 22 kilograms of unshelled nuts. A 0,5 kilogram sample of unshelled nuts weighs 0.24 kilogram after shelling.

Shelling percentage = (100 x 0.24)/0.50) = 48.00

Shelled harvest (kg) = (22 x 48.00)/100 = 10.56 kg

Grain yield for all legumes is calculated in much the same way as for sorghum and millet. Because plots in legume trials are small and the whole-plot harvest method is used, estimates of yield are sensitive to errors such as overlapping treatments. Consequently, it is important to discuss the requirements of a plot trial with farmers, to inspect plots, and to measure accurately plot size and grain harvest when yield estimates are important.

Note: For groundnut, when shelled harvest weight is required, it is best to leave the bulk of the harvest in the shell and use the approach described above. This is because groundnuts tend to store better in the shell.

A3.5.5 Method for Fodder Crop Plots

Fodder yield is generally measured by systematic sub-sampling, Most fodder trials are FMFI, with only a few plots on each farm. With farmers doing the work, a whole-plot measurement would be preferred. For economic and farmer assessments, larger plots for fodder production and use studies are required. Because fodder is bulky, weighing the entire harvest will not be practical. For this reason, the systematic quadrat sub-sampling method is used. Points to note about this method are the following:

• Procedures for fodder are similar to the leaf harvest method in legumes (Section A3.5.2).
• Set the schedule for harvest measurement with the farmer, Sub-sampling must be done before the farmer harvests the remainder of the plots.
• Time the harvest for each fodder crop to coincide with the stage of growth when the nutritive value is optimal.
• Cut the sub-samples and immediately carry them to a safe point outside the plot. Sub samples from one plot can be bulked together and weighed, This is a fresh weight. Fodder yield should be reported as dry weight. Spread out the bulked harvest samples for drying. Note that the recommendation for standard fodder production is to allow cut hay to dry to some extent in place. This reduces the weight and burden in transporting. Sub-samples are not heavy and removing them to a safe drying point lowers the risk of accidental loss.
• For precise estimates of fodder dry matter yield, research staff need to weigh a harvested sample at one-week intervals beginning about two weeks after the sample is harvested. The exact beginning time is dependent on the weather conditions being conducive for drying, To manage this work, only a representative portion of each bulked sample harvested from the field is separated and weighed. This new drying sub-sample can be put in a large burlap bag and hung to dry, A label on the bag can be used to record dates and weights. Reweigh the drying sub-sample a week later, If the weight is less than that of the previous weight, the sub-sample is still drying. Repeat the weighing until each drying sub-sample weight is constant. Use the constant weight as the dry weight of fodder in the sub-sample.
• Calculate dry matter (DM) yield as follows:

Proportion dry matter = sample dry weight (kg)/sample fresh weight (kg).

Fodder (kg DM)/harvest sample = bulked sample (kg fresh) x proportion dry matter,

Plot fodder yield (kg DM/plot) = [(fodder (kg DM) x plot size (sq. metres))] / sample size (square metres)

Fodder yield per hectare (kg DM/ha) = [(plot fodder Yield (kg DM/plot) x 10,000)] / plot size (square metres)

For example, 10 quadrat sub-samples (2m x 2m) are harvested from a plot measuring 30m by 30m. The bulked fresh weight of the samples is 42 kilograms. A drying sub sample, weighing 5 kilograms, is dried to a steady weight of 1,2 kilograms,

Proportion dry matter = 1,2/5 = 0,24

Fodder (kg DM)/harvest sample = 42 x 0,24 = 10,08 kg DM/harvest sample,

Plot fodder yield (kg DM/plot) = (10.08 x (30 x 30))/(10 x 2 x 2) = 226,8 kg DM/plot

Fodder yield per hectare (kg DM/ha) = (226,8 x 10,000)/(30 x 30) = 2520 kg DM/ha

• A less precise measure of fodder yield is obtained by following the procedure above with one small change, Instead of repeatedly weighing a drying sample, allow the bulked sample to dry for a set period of six weeks or more. After this period, assume the hay is dry, return, and weigh the entire bulked sample. The weight is for the harvest sample? that is, kilograms of dry matter/harvest sample,

Note: According to research experience, fodder plots are seldom harvested at the optimal time, Researchers should be warned, if harvest is delayed much past the optimal stage, senescence and leaf drop will dramatically reduce harvest yield and harvest quality.

A3.6 Yield estimation for mixed cropping

Mixed cropping research covers a range of trials. Trials differ in the type of growth habit and type of yield that is measured. For example, in a sorghum-melon mix study, yield measurement is for grain in sorghum and for vegetative growth in watermelon, In a sorghum-cowpea intercropping trial, measurement is for grain in both crops,

Many mixed cropping trials are designed to collect relatively precise agronomic data under farm conditions. In these cases research staff will measure yield, In the sorghum-melon mix trial, researchers use a systematic quadrat sub-sampling to measure sorghum yield together with an estimate of percent ground cover to measure melon growth, In a row-planted sorghum-legume intercropping study, measurements are made on selected rows. Yields are measured in a sorghum-legume mixed cropping trial using the whole-plot harvest method for each crop separately in the mix. These mixtures are broadcast seeded,

For example, as in the yield calculation illustration given for systematic quadrat subsampling (Section A.3.4.1), the FSD team finds 8.5 MT sorghum heads/quadrat in the field, 2.2 HR head stems/quadrat, 5.5 GR heads/quadrat, and 3,7 GZ head stems/quadrat. Sorghum plant numbers average 5.8/quadrat. The quadrat is 2m by 2m. Sorghum head samples weigh 0.032 kg./head for MT heads and 0.015 kg./head for HR heads from the threshing floor, As before (Section A3.4.1) assume the weight of the GZ heads is the same as that for the HR heads. In addition, at each quadrat sub-sampling FSD staff estimate watermelon plant ground cover as well as count the number of pigweed plants (Amaranthus spp.), a particularly pernicious weed on these fields. For ten quadrat sub-samples, percent ground cover estimates for watermelon are: 0, 0, 85, 55, I 5, 0, 25, 30, 65, and 25. Pigweed plant counts in the same quadrat sub-samples are 2, 4, 1, 2, 3, 0, 0, 1, 9, and 0.

Sorghum plant density (plants/ha.) = 14,500 plants/ha.
The estimated actual sorghum grain crop (kg/ha) = MT + HR + GR = 1203 kg/ha.
The sorghum grain yield lost to uncontrolled animals (kg/ha) = GZ = 139 kg/ha.
The watermelon ground cover (%) = (0+0+85+55+15+0+25+30+65+25)/10 = 30%.
Pigweed plant density = (2+4+1+2+3+0+0+1+9+0)/10 = 2.2 plants/quadrat = 5,500 plants/ha.

From these calculations, it can be concluded:

The intercrop = 14,500 sorghum plants/ha. + 30% watermelon ground cover.
The intercrop yield = 1203 kg. sorghum grain/ha. + 30% watermelon ground cover.
The botanical cover = 14,500 sorghum plants/ha. + 5,500 pigweed plants/ha. + 30% watermelon ground cover

A3.7 Method for estimating whole-farm production and average grain yield

Information on crop production for the whole farm can be used to describe and diagnose needs of households and inter-season variation in production. Average grain yield for the farm gives a baseline value to compare with results of trials. The Single Interview Visit Method is used. The procedure used is as follows:

• Select farmers to be interviewed following completion of threshing,
• At an interview with the farmer -- usually about 30 minutes -- record in kilograms or volume measures the amount of each crop harvested in shelled/threshed terms.
• Ask the farmer to point out the land that he or she has cultivated this season,
• Prepare a sketch of the cultivated part of the field,
• Measure all sides of the cultivated land with a measuring wheel, If the cultivated block has an irregular shape, divide the sketch into rectangular blocks, as large as possible. Measure the sides of these blocks.
• Calculate the land area cultivated as follows:

Area cultivated (square metres) = average length (metres) x average width (metres) for block of land that is cultivated.

If the block is irregular in shape, use the average of opposite sides as the approximate average length and width.

Cultivated land (ha) = area cultivated (sq. m)/10,000

• Calculate production as follows:

Farm grain production (kg threshed/shelled) = sum of weights (kg threshed/shelled) for grain crops harvested by the farmer.

Note the following additional points:

• The above formula implies adding production of sorghum and millet and legume crops. Although this is not strictly correct to do, it is probably not so serious an error, because the bulk of the production is sorghum.
• Because farmers express production in volume units, the following conversions may be helpful:

-- Standard grain bag holds 70 kilograms of grain.
-- Standard 20 litre bucket holds 14.2 kilograms of grain.

• Calculate the yield as follows:

Grain yield (kg threshed/shelled per hectare) = farm grain production (kp threshed/shelled) / land cultivated (ha)

• Calculate the yield for one crop (crop A) as follows:

Use the production (kg threshed/shelled of crop A) of crop A from interview with farmer and identify and measure the area planted to crop A.

Yield of crop A (kg threshed/shelled crop A per hectare) = farm production of crop A (kg threshed/shelled) / land cultivated with crop A (ha)

This type of interview can be quick and surprisingly accurate. Measuring the field takes more time than the interview but does not require the farmer to be present, The interview is useful to estimate wholefarm production, not individual plot yield. The method works well in estimating grain production of sorghum and millet, legumes, and so forth, but is less accurate for melon and spinach crop yields. Researchers do not need to notify a farmer before the interview visit, but advanced notification is preferable. The interview obviously must follow completion of threshing and shelling.

Contents - Previous