Residue Production

Method of data collection and data needed to compute the method:

Crop residues can be weighed at harvest from a known area using the methods described in the Tropical Soil Biology and Fertility manual (Anderson and Ingram, 1993) Sections 3.1.2 to 3.2.2 which are copied in Box 1.

Box 1. TSBF section 3.1.2 Herbaceous plants and short duration crops

Total biomass is measured by harvesting, drying, and weighing a number of small subsamples, or quadrats. Quadrat size depends on plant spacing: 0.5 m x 0.5 m is a convenient size for most grasslands; 1 m x 1 m may be appropriate for crops such as maize. Sample number (n) should be sufficient to reduce the standard error to about 10% of the mean. Use of 20 to 30 samples per treatment, distributed amongst replicates, is usually sufficient. Sample location should best be random, but systematic with a random start is acceptable.

Procedure

Cut all herbaceous vegetation within the quadrat at 2 cm above the ground (to avoid soil contamination), and sort into live (green) biomass and standing dead if possible.

Collect the litter from the ground for an estimate of litter standing crop.

Dry all samples as soon as possible to prevent decomposition.

Species composition in mixed communities is estimated by the dry-weight-ranking technique. The technique is based on a multiple regression for the dry weight of a mixed sample of herbage on the weights of the three heaviest species in the mixture. Experience indicates that it is easier to assess visually the rank order of the species in a quadrat than to estimate accurately their biomass. Tests in a large number of different communities have shown that the regression coefficients are fairly consistent between communities, and therefore do not need to be recalculated each time. The technique does not work well in communities completely dominated by one species, and tends to ignore rare species. The modified form given here, where the total biomass within the quadrat is also given a visual score, gives better results in communities where the total biomass is patchily distributed. Quadrat size should be small enough that species ranking is simple, but large enough that most quadrats have at least three species in them. Quadrats 0.5 m x 0.5 m square are usually adequate in grasslands. About 50 quadrats should be assessed per treatment.

Procedure

Walk around the plot to obtain a clear visual impression of what the minimum (1) and maximum (5) quadrat biomass looks like.

Locate the quadrats randomly or systematically after a random start.

In each quadrat (i) give the total biomass a score (w) between 1 and 5 according to whether it is near the minimum or maximum for the plot.

In each quadrat, give the species (j) which contributes most to the total quadrat biomass a rank score (rij) of 1, the second heaviest species a rank of 2, and the third heaviest species a score of 3. If a single species contributes more than about 70% of the biomass, give it ranks 1 and 3, or 1 and 2 (or even 1, 2 and 3 if it is the only species in the quadrat). Similarly, the second species could get a 2 and 3 if necessary.

(Excerpted from Anderson and Ingram, 1993, pp. 27-28)

When all the quadrats have been scored and ranked, calculate the score for each species:

Calculation

where Ʃwi is the sum of the quadrat scores for the quadrats where species j obtained rank r.

Add up the species scores to give a total score.

Determine the % contribution by species j to the total biomass:

Species j contribution to biomass (%) = (species score / total score) x 100.

Further reading

Gillen, R.L. and Smith, E.L. (1986) Evaluation of the dry-weight-rank method for determining species composition in a tallgrass prairie. Journal of Range Management 39, 283-285.

Jones, R.M. and Hargreaves, J.N.G. (1979) Improvements to the dry-weight-rank method for measuring botanical composition. Grass and Forage Science 34, 181-184.

Sandland, R.L., Alexander, J.C. and Haydock, K.P. (1982) A statistical assessment of the dry-weight-rank method of pasture sampling. Grass and Forage Science 37, 263-272.

3.2 ABOVE-GROUND INPUTS

3.2.1 Tree and shrub litter

In a comprehensive review of tropical litter fall data, Proctor (1983) observed that the results of many published studies were not comparable. This resulted from inadequate siting and replication of traps in relation to site heterogeneity, sampling for periods of less than a year and lack of standardization of small litter fractions (laves, twigs, reproductive structures and “trash”).

Litter trap construction

Litter traps are bags or boxes supported just clear of the ground with an aperture of 0.25- 1 m2.  A circular construction is best as it minimizes edge effects. Woven plastic bags are light weight for use in remote sites and can be tensioned into shape using lines attached to D-rings sewn around the mouth of the bag. The traps must allow free drainage of rain water but have a mesh size of approximately 1 mm or less to retain fine litter fractions. Trays on the ground surface can be used to measure litter-fall from dwarf shrubs etc., but animal activity, drainage and wind can present problems.

Similar considerations apply to collections of litter from quadrats on the ground, which may be necessary for estimating falls of palm fronds and larger woody litter. Trash fractions, however, which often have low mass but high nutrient content, will be lost by this method.

Procedure

Randomly locate litter traps (for material other than branches) within moderately homogeneous plots, or in a stratified random pattern (with 10 traps per subplot) in sites where it is necessary to include major variation in topography, soils and vegetation structure. (Excerpted from Anderson and Ingram, 1993, pp. 29 - 30)

Note: To achieve a 5% standard error about the mean, Newbould (1967) recommends the use of at least 20 traps/plot. In very heterogeneous sites, however, higher numbers of traps may be required.

Collect litter every 2 weeks and air dry it. More frequent collections may be necessary for litters which decompose rapidly, e.g. some tree legumes, while less frequent collections may be made under dry conditions (though the possibility of the litter becoming contaminated with dust and/or animal faeces should be recognised). [If the information in this box is taken verbatim from another document, it should be cited]

Sort the dried material into:

• leaves (including petioles and foliar rachises);
• small woody litter (twigs < 2 cm in diameter and bark);
• reproductive structures (flowers and fruits could be differentiated);
• trash (sieve fraction < 5 mm).

(For palm fronds, the leaflets, the rachis below 2 cm, and the remaining parts of the rachis should be weighed and recorded separately.)

Oven-dry subsamples of litter to obtain correction factors for moisture content (see Section 6.1)

Express all fractions defined above on an oven dry basis in g/m2/year or t/ha/year with 95% confidence limits.

Estimate branch fall from large (e.g. 100 m2) ground quadrats. Break twigs at the 2 cm diameter point, weigh the > 2 cm diameter material, subsample for oven-dry mass and other determinations as required.

3.2.2 Herbaceous litter and above-ground crop residues

The minimum level of sampling is at maximum and minimum biomass associated with major seasonal changes or perturbations; i.e. sampling four times a year under climatic regimes with a strongly bimodal pattern of rainfall. This will underestimate litter inputs as a consequence of material turning over between sampling dates and sampling at regular intervals every few weeks is recommended.

Procedure

Determine herbaceous litter (including grasses and forest ground flora) by harvesting quadrats in conjunction with biomass estimates (Section 3.1.2).

Separate litter, where possible, by plant species for the most frequent 80 % of species and bulked for the remaining 20%. (This may be impractical in very species-rich communities.)

Determine crop residues after harvest and at the time of ploughing or other manipulation.

Oven-dry litter subsamples to obtain correction factors for moisture content (see Section 6.1). Litter heavily contaminated by soil may need to be corrected for 'ash' content as well.

Express all fractions defined above on an oven dry basis in g/m2/year or t/ha/year with 95% confidence limits.

(Excerpted from Anderson and Ingram, 1993, pp. 29-30)

Unit of analysis:

The unit of analysis is the dry weight of plant biomass per area of land. A measured value of 100 grams per square meter is equivalent to 1 metric ton per hectare.

Limitations regarding estimating and interpreting:

One challenge with directly measuring the total amount of residue biomass is that if measurement is done once at harvest then it may miss the biomass from plants that lose leaves through the growing season, which may rot before harvest. Leaf traps collected monthly or weekly can be used to measure such biomass (see TSBF section 3.2.1 copied above). This may be especially important for establishing linkages between productivity and soil organic matter.

Method of data collection and data needed to compute the method:

In some contexts, farmers may be able to estimate the amounts of crop residues produced, such as when the residues are cut and stored for livestock feed. In such cases, crop residues can be estimated by farmers for a measured land area in a similar way as yield estimates are carried out (see “Crop productivity” indicator for more details). 

Unit of analysis:

The unit of analysis is the dry weight of plant biomass per area of land. A measured value of 100 grams per square meter is equivalent to 1 metric ton per hectare. Farmers are likely to share residue production using local units, such as ox-carts, bales or heaps. These can be approximately converted to kg by carefully measuring the weight of several local units.

Limitations regarding estimating and interpreting:

In most contexts, farmers are not  able to quantitatively estimate crop residue production in any reliable way, especially where residues are left in the field. In such contexts, it is better to consider another method or metric (such as farmer rating of residue production).

Method of data collection and data needed to compute the method:

If the residue biomass is not measured but grain yield is measured, then the residue biomass can be estimated from the harvest index for the variety of the crop. The harvest index is simply the portion of all biomass that is harvested as grain. As with yield, the crop residue biomass can be estimated through remote sensing of NPP as the non-grain portion of NPP, usually inferred by the harvest index.

Unit of analysis:

The residue biomass can be estimated as follows:

R = (G – G*HI)/HI

where R is the residues, G is the grain harvested, and HI is the harvest index. For more details on harvest index see Kawano (1990).

Limitations regarding estimating and interpreting:

The harvest index is not a fixed attribute of a crop but can vary across environmental conditions. For this reason, using grain yield to estimate biomass is a rough approximation, not a precise calculation.