Abstract

According to the literature, higher kernel content in a genotype may compensate for lower oil content and could even generate a higher income than a genotype with high oil content but lower kernel extraction. The objective of this study was to evaluate how kernel content affected income in oil palm. Deli x AVROS and Tanzania x AVROS were genotypes that fit into the above-stated assumptions, even though the increase in income was rather low (11,688 compared to 11,785 US$/100t of FFB). Two pairs (Deli x La Mé - Tanzania x La Mé and Deli x Ekona - Tanzania x Ekona) are examples in which reduced oil content was associated with a lower income. Deli x Ekona, the hybrid with the highest oil content in the bunch and the lowest kernel extraction rate produced the highest income (13,459 US$/100t FFB). No examples were found in which a high kernel content was associated with a particularly high income.

Introduction

The oil palm shows a wide variation in both kernel and oil extraction rates. In general, Deli mother plants yield more oil and less kernel than Tanzania genotypes, whereas oil content diminishes and kernel content increases when the male parent changes from Ekona to AVROS and La Mé (Alvarado and Sterling 1999). Kernel content is a trait with high heritability (Sterling et al . 1994). With both high variability and heritability it is feasible to obtain, through the appropriate crosses, a hybrid with a very high kernel content. Such genotype, with a lower oil but high kernel content could generate a higher income than a genotype with higher oil in the bunch but lower kernel content (Hartley 1983, Rajanaidu 1998). Thus, the objective of this study was to evaluate how kernel content in the bunch affected income in oil palm.

Materials and methods

Six hybrids (Deli x AVROS, Deli x Ekona, Deli x La Mé, Tanzania x AVROS, Tanzania x Ekona and Tanzania x La Mé) were planted in 1991 on experimental plots owned by ASD de Costa Rica S.A., located in the southwestern part of Costa Rica near the Pacific coast. Ripe bunches (tenera) were harvested biweekly from January to June 1998. The sample size necessary to obtain 8.5 % reliability and 95% confidence was between 17 and 37 bunches. However, a large number of bunches were analyzed in this study. Bunch components were determined according to Black et al.(1963). Income for 100 t of FFB was estimated from the percentage of oil and kernel in the bunch.

Results and discussion

The Hartley model

A genotype with 26 % oil and 6.5% kernel would produce a lower income than one with 24% oil and 9.75% kernel (Hartley 1983). To better understand the example, bunch components (rachis, spikelets, oil, fiber, shell and kernel) are expressed as percentages. Oil plus fiber in the bunch is equal to mesocarp in the bunch. In Hartley`s example ( Table 1 ), oil content was reduced in 1.62 units, and the fiber in 1.63 units (a total of 3.25 units). Kernel content also increased 3.25 units, which means that the amount reduced in mesocarp content corresponded to the increase in kernel content.

Model evaluation with real data

Deli x AVROS and Tanzania x AVROS are similar in terms of total fruits, mesocarp, shell and kernel in the bunch and oil in the mesocarp, which are the requisites in Hartley`s model. Oil content decreased from 28.5% to 27.2%, whereas kernel content increased from 4.6% to 6.8% (oil content was reduced by 1.3 units and fiber by 1.6 units for 2.9 units total, and kernel content increased by 2.2 units). These hybrids fit the model, even though 0.7 units were distributed among the rachis, spikelets and shell. Assuming a price of US$ 367/t of oil and US$ 263/t of kernel (Rajanaidu 1998), Tanzania x AVROS would generate a higher income than Deli x AVROS (11, 785 and 11, 688 US$/100 t FFB), which is predicted by the model.

The difference in income between Deli x AVROS and Tanzania x AVROS was rather small because the kernel contents are normal for commercial oil palm hybrids. A higher income will only occur if the ratios for fruits to bunch, shell to fruit and oil to mesocarp are kept constant, and at the same time, the kernel content is increased by a large proportion, for example from 6.5% to 9.7%. The reduction in oil content (with a higher price) is compensated for by a large increase in kernel content (with a lower price). Price compensates quantity and the closer these prices are the larger the impact on economic revenues.

Deli x La Mé and Tanzania x La Mé are different in terms of total fruits, mesocarp, shell and kernel in the bunch, as well as oil in the mesocarp. The reduction in oil content by 4.7 units and the 3.1-unit reduction in shell (for a total of 7.8 units) were associated with an increase of 4.7 units in the rachis and spikelets, 1.1 units in the fiber and only two units in kernel content (7.8 units in total). Tanzania x La Mé generated a lower net income than Deli x La Mé ($10, 912 vs $12, 125), despite the fact that its kernel content was a third lower (7.5% vs 5.5%).

Deli x Ekona and Tanzania x Ekona are also different in terms of total fruits, mesocarp, shell and kernel in the bunch, as well as oil in the mesocarp. The 7.4-unit reduction in oil content and 4.4-unit reduction in fiber (11.8 units in total) are related to an increase of 6 units in rachis and spikelets, 3.8 units in shell, and only two units in kernel content (11.8 units in total). Tanzania x Ekona had a lower net income than Deli x Ekona ($11, 275 vs $13, 459), in spite of having higher kernel content (5.3% vs 3.3%).

Hartley's model assumes that the reduction in oil content (and fiber) is directly related to an increase in kernel content, and that other bunch components remain constant. The pair Deli x AVROS and Tanzania x AVROS is an example that follows this model. However, it was found more frequently that the reduction in oil content was related to changes in other bunch components, which can have a negative effect on income. Deli x La Mé and Tanzania x La Mé, as well as Deli x Ekona and Tanzania x Ekona are good examples in this last category ( Table 2 ).

Kernel influence on income

A hybrid like B in Hartley`s model ( Table 1 ) may exist. However, none of the bunches analyzed that had high kernel content really produced outstanding income. Oil in the bunch has a large bearing on income (316 bunches analyzed, r²=0.91), whereas the situation is different for kernel content (r² = 0.0.4). On the other hand, 66% of the 32 palms that yielded the greatest net income were Deli x Ekona, which is a hybrid with low kernel content (3.3% average). Even without considering this particular hybrid, the rest of the palms also are characterized by high oil and low kernel contents ( Table 3 ). In conclusion, no evidence was found that high kernel content was related to higher income.