4. Recording of traits for selection

4.1 Technique of recording

The actual technique of recording yield and growth has often received little attention. There are numerous examples of well designed breeding programs which are recorded with awkward equipment, resulting in unreliable and labour-intensive yield and growth recording.

The following sections outline some relevant aspects of recording technique.

4.1.1 Bunch yield

Bunches are either weighed per palm (for example, in the source population of parent palms) or assembled per experimental plot, and weighed in bulk (testing of sources of planting material ).

Bunches are weighed per palm by using a spring balance attached to a tripod. The tripod is made of thin pieces of board, which are loosely tied together at the top. By extending one leg, the tripod can be used to lift a weight of about 150 kg by one worker, so that total plot yield can be recorded. Both methods are depicted in Fig. 2.

4.1.2 Bunch analysis

In a special building, bunches are analysed for the components of oil and kernel extraction, i.e. the ratio of fruit-to-bunch, mesocarp-to-fruit, oil-to-mesocarp and kernel-to-fruit, as follows.

Bunches are weighed at their arrival. After weighing, the spikelets are removed and weighed. A random sample of spikelets is then taken and weighed. The fertile and parthenocarpic fruit with oil-bearing mesocarp are removed and weighed to give the ratio of fruit to empty spikelets of the sample.

As the weight of the total spikelets of the bunch is known, the fruit-to bunch-ratio can be calculated.

A sample of normal fertile fruits is also taken shortly after arrival of the bunch. The fruit sam ple is weighed and the mesocarp is scraped off; the nuts are weighed and the weight of the mesocarp is calculated by difference. This gives the weight of the mesocarp at the moment of taking the fruit sample, that is the weight unaffected by drying of the fruit before scraping and during scraping of the mesocarp (the importance of determining the accurate weight of the mesocarp will be explained below).

The mesocarp-to-fruit ratio can be calculated from the weight of the fruit sample and the nuts. Nuts are air dried for about 3 days to facilitate cracking. The shell is removed and the kernels are weighed, so the kernel-to-fruit ratio can be calculated.

All of the scraped mesocarp is dried in an oven at 105° C to constant weight to determine the moisture content of the fresh mesocarp. As mentioned before, the fresh mesocarp is calculated by difference, so it is crucial that all mesocarp is carefully collected during scraping; any mesocarp spilled is calculated as moisture losses during drying in the oven, and thus overestimates the moisture content of the mesocarp.

Figure 3 presents a table, designed to eliminate losses during scraping. It shows how the mesocarp is directly collected in an alluminium tray fitted in a drawer under a hole in the top of the table.

The oil content of a sample of oven-dried mesocarp is determined by direct extraction with large Soxhlet extractors.

As the moisture content of the sample is determined earlier in the analysis, the ratio of oil-to fresh mesocarp content can be calculated.

Research centres show a wide variety of bunch analysis procedures, in terms of sampling methods, the weight of the sample and its method of preparation ( for a review see Hartley, 1988).

Bunch analysis is the most costly part of the breeding procedure. The ultimate aim is to determine the components of oil and kernel extraction of individual palms or families with sufficient precision at minimal costs. More research is still needed to achieve this objective.

4.1.3 Leaf measurements

An oil palm leaf is attached to the stem with the petiole. The leaf bears leaflets on each side of the leaf stalk. The latter may be divided into two zones: the rachis bearing the leaflets, and the petiole, which is much shorter than the rachis and bears only short lateral spines. This is clearly illustrated in Fig. 4. For a more specific description see Hartley (1988).

Leaf marking

Leaves assigned for measuring are marked at the date of opening, either by the number of the month (Fig. 5) or by a paint dot, the colour of which corresponds to the date of marking; the latter method is quicker and preferred for progeny testing. As leaf size does not change anymore after opening, measurements can then be done when convenient.

4.1.4 Trunk measurements

Height

Height increment for a given period is measured between the insertion of leaf bases of known opening date. Fig. 13 shows how the level of the lower reference leaf base is obtained by means of a fluid leveller because it is usually not directly above the leaf base marked at a later date; a movable bar is then brought to the insertion of the higher marked leaf base. The height difference between the two bases of the leaves marked at different ages (level of the growing point) is directly read from the attached sliding tape.

Actual height is measured to the base of leaf 25 at the end of yield recording height to compare progenies for mature height. This measurement is indicative for the economic life.

Trunk diameter

The diameter is measured at about 150 cm above the ground, that is, when the trunk has already decreased to and largely constant value. Fig. 13 shows how the widest distance between opposite sides of the trunk can be obtained between every 4th spiral. This proves to be a convenient guide for unskilled workers to prune obstructing leaf bases in order to expose the stem. Fig. 14 also depicts the equipment and technique of measuring the stem diameter.

4.2 Measurements to estimate growth parameters

Vegetative Dry Matter production (VDM)

VDM is restricted to dry matter incorporated in in trunk growth and leaf production (above-ground dry matter production).

Trunk dry matter production is estimated from measurements of trunk increment, trunk diameter and an estimate of dry matter per unit trunk volume (kg/dm3 ); the latter depends on palm age (years) as follows:

0.0076*years after planting + 0.083 (Corley et al., 1971).

Leaf dry matter production is the product of leaf production and mean weight of the leaves. The Rachis length

The rachis is measured from the start of the rudimentary leaflets to the split of the terminal leaflets, the apex, (Fig. 6).

Petiole cross-section area

The width and the depth of the petiole are measured at the initiation of rudimentary leaflets. As the centre of the rachis is usually below the edge, an extension is fitted to a conventional pair of calipers as shown in Fig. 7.

Fig. 7 also shows how the technician reads the measurements directly over the shoulder of the worker holding the calipers.

Length x width of leaflets

From both sides of the rachis a set of 10 leaflets are cut immediately under the region 2/5 from the apex (this region is the area with the larger leaflets). From each of these two sets the three longer leaflets are sampled (Fig. 8) and the length and width are measured (Fig. 9).

Counting leaflets

Only the number of leaflets on one side of the rachis are counted, including rudimentary leaflets at the base and fused leaflets at the apex, using a hand-counter (Fig. 10).

Leaf counting

Leaves open at regular intervals, usually in mature palms, about two leaves per month. In order to record the number of leaves which open at a certain time interval, the youngest fully openend leaf (leaf 1) is marked at the start and at the end of the period. As the marked leaf becomes older, a higher rank number is assigned. Since the difference in ranking (age) of sequential leaves on each spiral is eight leaves, the position of the first marked leaf in relation to the latest can easily be obtained from a specially designed diagram (Fig. 11). This gives the order of the leaves in the crown; the rate of leaf production is then obtained by the difference between the order of the leaves.

In order to number the spirals, the palm should first be classified whether its direction of the spiral is left- or right-handed (Fig. 12). The side at which inflorescences emerge from the leaf axils can be used for a final classification: inflorescences of right-handed palms emerge from the right side of the leaf axil and those of left-handed from the left. The spiral with the youngest opened leaf is assigned spiral 1.

Fig. 12 also shows how spirals can then be conveniently numbered by following the direction of the spiral downwards.

For example, the leaf was marked at opening in November 1992 and again one year later. Assume that leaf production for this period was recorded in December 1993. The leaf marked in 1992 is at that date leaf 4 in spiral 7 (leaf 27) and that marked in 1993 leaf 1 in spiral 4 (leaf 2). Leaf production for November 1992 to November 1993 is then 27-2=25 leaves.

4.1.4 Trunk measurements

Height

Height increment for a given period is measured between the insertion of leaf bases of known opening date. Fig. 13 shows how the level of the lower reference leaf base is obtained by means of a fluid leveller because it is usually not directly above the leaf base marked at a later date; a movable bar is then brought to the insertion of the higher marked leaf base. The height difference between the two bases of the leaves marked at different ages (level of the growing point) is directly read from the attached sliding tape.

Actual height is measured to the base of leaf 25 at the end of yield recording height to compare progenies for mature height. This measurement is indicative for the economic life.

Trunk diameter

The diameter is measured at about 150 cm above the ground, that is, when the trunk has already decreased to and largely constant value. Fig. 13 shows how the widest distance between opposite sides of the trunk can be obtained between every 4th spiral. This proves to be a convenient guide for unskilled workers to prune obstructing leaf bases in order to expose the stem. Fig. 14 also depicts the equipment and technique of measuring the stem diameter.

4.2 Measurements to estimate growth parameters

Vegetative Dry Matter production (VDM)

VDM is restricted to dry matter incorporated in in trunk growth and leaf production (above-ground dry matter production).

Trunk dry matter production is estimated from measurements of trunk increment, trunk diameter and an estimate of dry matter per unit trunk volume (kg/dm³ ); the latter depends on palm age (years) as follows:

0.0076*years after planting + 0.083 (Corley et al., 1971).

Leaf dry matter production is the product of leaf production and mean weight of the leaves. The weight of an individual leaf is estimated as 0.1023*P + 0.2062 , where P is the mean petiole width * depth in cm² or petiole cross-section. (Corley et al., 1971).

Leaf area (LA)

The area of a single leaf is estimated, using the method of Hardon et al. (1969) as c*(n * lw), where

c = a correction factor which varies slightly with palm age (0.51 to 0.57 for palms of 1-2 to 8-11 years, respectively);

n= number of leaflets and

lw= the mean of length * mid-width of three leaflets sampled from each side of the rachis.

Leaf area ratio (LAR)

LAR is defined as N*A/V , where

N= leaf production,

A= mean area per leaf and

V= VDM.

Parameters of the logistic growth curve for leaf area at different years from planting

The logistic growth function f(t) = A/(1 + B* e-^Ct ) must be fitted to the data by the Least Squares Method to estimate the parameters A, B and C.

Rewriting the logistic growth function as follows:

with

A= Lm, B= (Lm - Li)/ Li and C= k,

where k = the relative rate of growth of the mean leaf area,

Lm = asymptotic maximum leaf area,

Li = leaf area at field planting.

For selection purposes this is conveniently expressed as the time to reach 95% of the maximum leaf area (t_0.95 ) as follows:

Bunch Dry Matter Production (Y)

This is calculated as 53% of the weight at harvesting.

Components of oil and kernel extraction

The percentage oil-to-bunch is calculated as the product of the ratios fruit-to-bunch, mesocarp-to-fruit and oil-to mesocarp. The percentage kernel extraction is the product of fruit-to-bunch and kernel-to-fruit.

Bunch Index (BI)

BI is the ratio of dry weight of fruit bunches to total above-ground dry matter production per palm.

BI=Y/(V + Y), where

Y=dry weight of bunches,

V=VDM.

Harvest Index (HI)

HI is the ratio of oil and kernel yield to total above-ground dry matter production.

HI= (ratio of oil-and-kernel to bunch) * bunch yield/(V + Y).

4.3 Timing of measurements

4.3.1 Bunch yield

Number of bunches and the total weight are determined at each harvesting round of 7 to 10 days. This gives the number of bunches and their mean weight per palm or per plot.

Yield in favourable environments increases rapidly during the first 3 years and then stabilises (Breure, 1988). As early yield greatly depends on the spread of the leaf area, another two years at mature yield level are recorded to obtain a fair estimate of the yield potential. A total of five years of yield recording is thus needed.

4.3.2 Bunch analysis

Oil content of the mesocarp increases sharply with age during the early period of bunch production. Conventionally, oil analysis starts therefore when the oil content by and large stabilises, about 2 to 3 years after the start of production. The rate of increase differs among progenies (Corley, personal communication), so precision is enhanced if the analysis of bunches starts shortly after the start of production. Early information on extraction rate is also desirable in case new sources of planting material are tested.

Until more information on the number of samples per progeny is available, a tentative schedule is to analyse 32 bunches/progeny/year during the first three years of production and 64/bunches/progeny/year in the fourth and fifth year.

Because of the wide diversity in bunch composition among palms in a progeny, the progeny sample must include as many palms as feasible.

It is equally important to analyse each month the same number of samples per progeny to eliminate known seasonal variations in oil content.

4.3.3 Vegetative Growth

Clearly, growth recording should be completed by the end of the five-year period of yield recording.

The start of production depends on environmental conditions, age of seedlings etc. It is therefore more convenient to follow initially a recording schedule according to the time after field planting, but change this later on to the time after harvesting.

4.3.4 Leaf measurements

Measurements are required for the following objectives:

(i) To establish the logistic growth curve of leaf area against age.

The leaf area fits the logistic growth function f(t)=A/(1+B*e^Ct) as explained in section 2.7.4. A set of measurements at 6, 42, 66 and 90 months after field planting is recommended for a step 2 progeny test (first screening).

For a step 3 progeny test an additional measurement is recommended 12 months after field planting.

(ii) To obtain a reliable estimate of Leaf Area Ratio (LAR).

(iii) To obtain the petiole cross-section in order to estimate dry leaf weight.

The timing will be described in the following sections.

4.3.5 Leaf marking

The first measurements are done on leaves marked at 6 months from planting, that is when leaf size starts to increase after a period of so-called "transplanting shock" due to root disturbance during the movement of seedlings to the field. This leaf is assigned L0.

The second leaf, marked one year after the start of bunch production (L1), serves as a reference point for height measurements. The latest fully opened leaves are marked at the end of the third and fifth year of production, and assigned L2 and L3, respectively. Note that these leaves refer to the date of opening while leaf 25 and leaf 41 (see section 2.7.6 and section 4.1.3) are actually assigned according to their ranking in the crown.

4.3.6 Production of leaves

Leaf production is determined between L1 and L2; and between L2 and L3.

4.3.7 Height measurements

Height increment is measured from L1 (reference point) to the insertion of L2 (stem increment for a period of two years). Also to leaf 25 at the end of the fifth year of production.

4.3.8 Trunk diameter

Measurements are done only once at the end of the fifth year of production, when the lower leaves are about 150 cm above the ground.

4.3.9 Leaf magnesium level

To assess magnesium status of individual palms, samples are collected in 6 successive months and bulked (selection of parent palms in the source population). Samples to determine progeny means are collected for each palm once and bulked per plot. Samples thus obtained are analyzed for magnesium content.

Sampling should be done when symptoms of magnesium deficiency are most pronounced. This is usually at the end of the second year of production when lower leaves are still exposed to light and palms are bearing the first heavy crop (stress due to high fruiting activity).

4.3.10 Crown disease

Breure & Soebagjo (1991) observed the first symptoms of crown disease on newly emerged leaves at 8 months from planting. Severity reached a peak at 12 months; thereafter, severity gradually diminished until it by and large stabilised between 22 and 35 months. Note, that severity, in terms of the degree of bending of the leaves, was recorded on the nine youngest leaves. The bend is permanent, so symptoms persist for at least a year when scoring on the nine youngest leaves is done.

Incidence can therefore conveniently be scored at 18, 30 and 42 months after planting. Once crown disease has been observed, the palm is marked to avoid double counting in another round.

At 42 months the percentage of affected palms, recorded during the three rounds, can be established for each progeny.

4.4 Components of growth

Table 9 presents an example of mean measurements of one progeny assembled at various periods during the first 90 months after field planting; it is assumend that bunch production starts 30 months after planting.

The following sections illustrate how the relevant growth parameters are estimated.

4.4.1 Bunch yield (kg /palm)

Bunch yield for year 1 to 5 of production are 65, 140, 230, 185 and 190. The dry weight is estimated as

Year 1: 0.53 * 65 = 34.75

Year 2: 0.53 * 140 = 74.20

Year 3: 0.53 * 230 = 121.90

Year 4: 0.53 * 185 = 98.05

Year 5: 0.53 * 190 = 100.70

4.4.2 Oil and kernel extraction ratio

From the bunch analysis results it is calculated that the average extraction of the fresh fruit bunches is 27.3% mesocarp oil and 2.7% kernels.

4.4.3 Leaf measurements

Leaf weight (kg)

LO : 0.1023 * 6.78 + 0.2062 = 0.90

LO*: 0.1023 * 9.26 + 0.2062 = 1.15

L1 : 0.1023 * 17.15 + 0.2062 = 1.96

L2 : 0.1023 * 23.60 + 0.2062 = 2.62

L3 : 0.1023 * 34.15 + 0.2062 = 3.70

Leaf area (m²)

L0 : 0.51 * 151 * 159 = 1.22

L0*: 0.51 * 180 * 207 = 1.90

L1 : 0.53 * 275 * 420 = 6.12

L2 : 0.55 * 340 * 515 = 9.63

L3 : 0.56 * 375 * 547 = 11.49

4.4.4 Stem measurements

Height increment (cm/year)

The distance from the reference point (L1) to the insertion of leaf L2 is 152 cm.

The annual stem increment is thus 152/2 = 76 cm.

Trunk diameter (cm)

The diameter of the exposed stem is 55.2 cm.

4.4.5 Production of leaves

Leaf production between those marked at opening 42 and 66 months from planting (L1 and L2, respectively) is 56.2. Annual leaf production for this period is thus 56.2/2 =28.1.

In the same way the production between months 66 and 90 amounts to 49.2/2 =24.6.

4.5 Calculation of growth parameters

4.5.1 Bunch dry matter production (kg/palm/year)

Month 42 to month 66 (years 2 and 3 of production):(74.20 + 121.90)/2 = 99.38.

4.5.2 Vegetative dry matter production (kg/palm/year)

Leaf dry matter production

Months 42 (L1) to 66 (L2) =

(1.96 + 2.62)/2 * 56.2/2 = 64.35

Months 66 (L2) to 90 (L3) =

(2.62 + 3.70)/2 * 49.2/2 = 77.74.

Trunk dry matter production

Height increment and trunk diameter are assumed to be by and large stabilised at the time of measuring.

For months 42 to 66 and for months 66 to 90 trunk dry matter is estimated as follows:

Volume increase:

(55.2/2)² * 76 = 181.79 dm³

Weight per volume

= 0.0076 * (90 - 42)/12 + 0.083

= 0.11 kg/dm³

Trunk dry matter production:

181.79 * 0.11 = 20.0.

VDM

Months 42 to 66:

64.35 (leaf DM) + 20.0 (trunk DM) = 84.35

Months 66 to 90:

77.74 (leaf DM) + 20.0 (trunk DM) = 97.74.

4.5.3 Bunch Index

Months 42 to 66:

98.05/(84.35 + 98.05) = 0.538

Months 66 to 90:

99.37/(97.74 + 99.37) = 0.504.

4.5.4 Harvest Index

Months 42 to 66:

(185.0 * 0.30)/(84.35 + 98.05) = 0.304

Months 66 to 90:

(187.5 * 0.30)/(97.74 + 99.37) = 0.285.

4.5.5 Leaf area ratio (m² /kg)

Months 42 to 66:

{28.1 * (6.12 +9.63)/2}/84.35 = 2.62

Months 66 to 90:

{24.6 * (9.63 + 11.49)/2}/97.74 = 2.66.

4.5.6 Parameters of the logistic growth curve

The curve is fitted through estimated leaf area (y) at the following months after planting (t); the data pairs (t,y) were as follows:

(6, 1.22); (12, 1.90); (42, 6.12); (66, 9.63) and (90, 11.49).

The logistic growth function f(t) = A/(1 + B* e^-Ct ) has been fitted to these data by the Least Squares Method and gives A= 12.180772, B= 11.286396, C= 0.057470.

Rewriting the logistic growth function as follows:

with A= L_m, B= (L_m - L_i)/ L_i and C= k,

where k = the relative rate of growth of the mean leaf area,

L_m = asymptotic maximum leaf area,

L_i = leaf area at field planting.

For selection purposes this is conveniently expressed as the time to reach 95% of the maximum leaf area (t_0.95 ) as follows:

This gives the following parameters for selection:

L_m = A =12.180772, L_i = A/(1+B) = 0.991403, k= C =0.057470.

Now 0.95*L_m = 11.571733 and hence

= -17.400383 * (-5.368037) = 93.4059 months after planting.