Abstract

With the aim of finding out more about the factors that affect the variation in oil extraction per bunch at both laboratory and industrial levels, a study was carried out from 1990 to 1995 on an oil palm plantation located in the Southern Pacific area of Costa Rica. On a weekly basis, four to six mature bunches of Deli x AVROS origin were collected and analyzed on the basis of their components, and, at the same time, climatic data were collected and industrial oil extraction rates were recorded.

A seasonal pattern was noted in the average bunch weight and in fruit set. This pattern was associated with seasonal variations in climate. The variation in fruit set  may be associated with an excess of precipitation and the subsequent effect on the viability or availability of pollen, as well as on the activity or population of pollinating insects.

The percentage of oil in the mesocarp also showed annual variations, while the percentage of    oil in the bunch was higher during the second half of the year. The fluctuation of the industrial extraction rate  was similar to that observed under laboratory conditions. The lowest levels occurred in January and February, with an increase until July, and a reduction beginning in August.

Correlation analysis among bunch components in the laboratory showed that the variation in percentage of oil in the bunch depended mainly on the total numer of fruits and on the percentage of oil in the mesocarp. The correlation with climatic factors showed that the total number of fruits in the bunch was positively related to hours of sunlight, and negatively related to precipitation occurring five to six months before harvest, that is, during inflorescence pollination.

According to these results, the proportion of fruits in the bunch is the main determinant for the oil content in the bunch. More studies should be conducted on the fluctuation of pollinating insect populations, the seasonal pattern of the production and viability of pollen, and other aspects that may affect the percentage of fruit or other bunch components that contribute to the production of oil.

Introduction

Palm oil extraction mills in Costa Rica show annual fluctuations in their extraction rates.  This behavior has also been documented in other countries, especially in Malaysia, where it has been the object of considerable study in recent years.

The following factors are cited in the literature (Chan and Lee 1993; Chow 1993; Lee and Shawaluddin 1993) as the principal causes of these fluctuations:

• Type of planting material, palm age, and deterioration in commercial seed production techniques (contamination of dura palms).
• Variation in fruit set caused by seasonal climatic changes on pollination rates.  This last factor is closely related to fluctuations in the pollinating insect populations, as well as the availability and viability of pollen.
• Climatic and geographical factors (e.g., rain and sunlight) affecting plant physiology.
• Harvesting procedures, taking into account such factors as availability of labor, spacing of harvesting cycles, setting of maturity standards, and loss of loose fruits during harvest and transport.
• Mill efficiency.

With the aim of determining the main influences on various bunch characteristics and on the oil extraction rate, a long-term analysis of bunch composition was carried out, during which time some of the environmental factors were considered.

Materials and methods

The study was carried out from January 1990 to July 1995, on Compañía Palma Tica's plantation, Coto 47, on the southern Pacific coast of Costa Rica.

The study consisted of the weekly collection of four to six bunches, with maturity criteria for collection being at least three loose fruits. The bunches came from a group of 60 Deli x AVROS palms, all from the same progeny and planted in 1986.

The bunches were taken to the laboratory for analysis, using the methodology of Blaak et al . (1963). During this time, climatic information was collected at the meteorological station in Coto 47, and oil and kernel extraction rate records from the local processing plant were compiled.

A simple correlation analysis was done for bunch components, the extraction rate at the laboratory level, and different climatic variables, taking into consideration that the possible effects of external variables on the bunch characteristics could have occurred several months before the laboratory analysis was done.

Results and discussion

Laboratory study of bunch components

Table 1 shows the monthly averages of the bunch composition variables, as well as oil extraction rates in both laboratory and industrial settings. In Figures 1 , 2 , 3 , 4 , 5 , 6 , and 7 , the data for these factors are presented for the entire duration of the study.

The average bunch weight showed seasonal variation, with a marked increase beginning in April and lasting until the maximum value was reached in July ( Fig.1 ). The lowest average weight was observed between January and March, while the period between July and December was characterized by high average bunch weight ( Table 1 ). During the study, the bunch weight showed a long-term increase, related to increasing palm age.

The pollination efficiency, as measured by the proportion of fertile fruit per bunch (FF/B) and total fruit per bunch (F/B), was also affected by seasonal variations in climate ( Fig.2 ). These variables increased beginning in May, and reached a peak in August, while the lowest F/B value was seen between February and April ( Table 1 ).

Oil in the mesocarp (O/M) showed a tendency to steadely increase during the whole period, but, no tendency was clear within any particular year ( Fig. 3 ).

Oil per bunch (O/B) was low during the first half of the year, reaching its lowest values in the months of February and March.  During the second half of the year, the highest oil extraction rates were recorded, particularly in July and August ( Table 1 ).

Correlation analysis of bunch components ( Table 2 ) shows that fruit per bunch (F/B) was the main component in the total variation of oil per bunch (O/B). The correlation coefficient between these variables was 0.76 (p<0.01). Oil to mesocarp, with a correlation coefficient of 0.35 (p<0.05), was the second-most-important determining factor in the variation of oil per bunch.

Fruit per bunch showed a positive correlation with hours of sunshine, and a negative correlation with precipitation received five or six months prior to harvest, the period during which pollination would have occurred. Oil per mesocarp showed a negative correlation with the evaporation rates recorded three to four months before harvest ( Table 3 ).

During the rainy season, the proportion of fertile fruit per bunch was consistently higher than in the dry season. This variable is associated with events that happen five to five and a half months before the bunches are harvested, which is to say, during pollination. Thus, for example, the determination of the total fertile fruit per bunch (FF/B) harvested between February and April occurs between the months of September and November.  These are the rainiest months, which is considered to be the cause of the poor pollination of these bunches.  This period coincides with the smaller population of pollinating insects, and possibly with the rapid decline in pollen viability partially due to high humidity in the enviroment (Chinchilla y Richardson 1989).

The number of male inflorescences is normally greater between June and November. This is the period during which fruit set is determined for those bunches harvested between December and April. This is an indication that the reason for a low F/B ratio during the dry season, is not due to a lack of pollen during the rainy season but the low viability of such pollen. During the months of high rainfall, raindrops may knock the pollen of the inflorescences, or too-wet pollen may not adhere to insects. These are areas to be investigated.

In summary, the low proportion of fruit to bunch and the low rate of oil per mesocarp are responsible for the low rate of oil per bunch (O/B) which occurs during the first months of the year.  The drop in oil per bunch which is seen in September seems to be related to the low oil per mesocarp values recorded in August ( Table 1 ).

Industrial oil extraction rates in coto 54

The variation in industrial extraction rates is clearly seasonal.   It begins with the lowest values in January and February, which then steadily increase from March to July, and which finally decline beginning in September ( Table 1 ). Furthermore, the industrial extraction rates also show another seasonal trend with a cycle of two to three years ( Fig. 5 ), which is evident after 1991, when mill efficient was improved.

Bunches processed at the mill are highly heterogeneous, and those analyzed at the laboratory during this study, were choosen for their homogeneity. Neverless, the variation in the industrial oil extraction rate and that determined at the laboratory were similar. This is a good reason to assume that there are important enviromental factors determining the bahavior of this variable.

A high load of fruit on the trees, under limited light conditions, may lead to a low amount of oil produced, which is then distributed among many bunches. This causes a large variation in O/B and the industrial oil extraction rate (Chow 1993). More energy is needed to produce oil than carbohydrates, which mean that a high oil production can not be compatible with a high bunch load (carbohydrates) in any particular palm (Henson 1993).

Conclusion

The proportion of fruit to bunch is the main factor in the oil content per bunch.  One of the possible ways to improve the low industrial extraction rate in Coto between February and April would be to increase the efficiency of pollination between August and October.

Additional research is required on aspects that hamper satisfactory pollination, such as:

• Annual fluctuation in the population of pollinating insects.
• Seasonal patterns in production, and viability of pollen.
• Nutritional aspects affecting the bunch composition.
• Relationship between the industrial oil extraction rate and FFB produced.

Bibliography

BLAAK, G. et al .  1963.  Breeding and inheritance in the oil palm Elaeis guineensis Jacq.  Part II. Methods of bunch quality analysis. J.W. Afr. Inst. Oil Palm Res., 4:146-155.

CHAN, K.W.; LEE, K.H. 1993.  OER:  A concern facing the oil palm industry.  Proc. Nat. Sem. Palm Oil Extr. Rate:  Problems and Issues.  PORIM, Malaysia.  Pp. 1-16.

CHINCHILLA, C.; RICHARDSON, D.L.  1989.  Situación actual de los insectos polinizadores y la polinización en palma aceitera en Centro América.  Boletín Técnico OPO - United Brands 3(2): 29-48.

CHOW, C.S.  1993.  The variability of oil extraction rate in the Malaysian palm oil industry.  Proc. Nat. Sem. Palm Oil Extr. Rate:  Problems and Issues, PORIM, Malaysia.  Pp. 17-26.

HENSON, I.E.  1993.  Factors determining mesocarp oil to bunch ratio in the oil palm:  A physiological perspective.  Proc. Nat. Sem. Palm Oil Extr. Rate:  Problems and Issues, PORIM, Malaysia.  Pp. 27-35.

LEE, C.H.; SHAWALUDDIN, T.  1993.  Recent trends in OER in relation to MPD analyses in Golden Hope.  Proc. Nat. Sem. Palm Oil Extr. Rate:  Problems and Issues, PORIM, Malaysia.  Pp. 79-90.