Introduction

Although several open pollinated Deli dura progenies from Sumatra and some West Africa germplasm samples were acquired between 1922 and 1969; the ASD Breeding Program commenced its oil palm breeding efforts in 1974, when ASD (former United Fruit Company) received a wide range of oil palm genetic material through exchange programs agreed with OPGL (Hardon, 1969), SOCFIN (Knetch, 1970), Uniliver (Blaak, 1970), IRHO (Meunier, J., 1978), DAMI (Rosenquist, 1978), Kade, Ghana (Wonkhy-Appyah, 1978). Simultaneously, some wild populations were prospected in Africa, including Tanzania (Kigoma), Cameroon highlands (Bamenda), Sierra Leone, and Malawi and Mobai (Table 1).

These exchanges of advanced oil palm breeding material and wild population sources provided ASD with one of the most complete and comprehensive oil palm genetic collections, in terms of both dura and pisifera sources. Probably only PORIM has a germplasm collection with such a degree of genetic diversity. 

ASD's breeding program aims to produce pollinated oil palm seeds with the highest genetic potential for oil yield and the best heritable secondary characteristics a premier product which will allow ASD to become the world's foremost oil palm seed supplier. The research objectives have been classified into four main areas: breeding and selection of compact planting material, seed production for high-yielding small palms, development of planting materials adapted to specific environments, and selection of outstanding (Ortet) individual plants for in vitru propagation and cloning.

Oil palm seed production in the PIPA/ASD program has been based on the Deli x AVROS cross for almost two decades. This high yielding and precocious material has also been the foundation of the oil palm industry in Asia for an even longer period. The weakness of the AVROS material is its excessive vegetative growth, both in terms of trunk height and leaf length.

The replacement of traditional commercial material has begun with the distribution of Deli x Ekona and Deli x Calabar hybrids which produce at least comparable oil yields per unit land area with smaller palms. There are also many other alternatives for new commercial hybrid planting material, and it is the purpose of this breeding program to identify and rank them.

• Develop a seed production program for compact palms with growth characteristics which will allow increased oil production from high density populations.
• Develop a seed production program for several high-yielding small palm hybrids which will produce more oil per hectare than current Deli x AVROS hybrids, but with better growth characteristics.
• Develop a seed production program for palm hybrids with special features to satisfy grower requirements and for adaptation to specific environments.
• Select Duras and pisiferas for clonal reproduction of seeds, aiming to reproduce those outstanding progenies identified in progeny test trials. The research also aims to select tenera ortets for in vitru propagation, and to evaluate and select clones for planting on a commercial scale.

The term compact, as used in our program, refers to a palm growth type characterized by short trunks and short leaves, which arose as a unique event in a backcross of an oleifera-guineensis interspecific hybrid. The usefulness of the genetic system controlling compact growth is that, unlike variation for plant size in non-compact palms, growth is controlled by few gene pairs. This allows the introduction of compact growth into a broad series of genetic backgrounds relatively rapidly, and, as such, probably offers the best potential for developing high-yielding small palms. For this reason, the development of a seed production program for compact palms is our priority.

Comparison of Deli AVROS (left) and BC3 compact oil palm planting materials four years after planting

The original compact palm (OCP) was self-pollinated and the palms were planted in trial CB80-10. These palms showed great variation in growth; some palms were the same size as the OCP, others were larger and still others considerably smaller.

The smaller palms, called "super-compacts," presumably arose from an accumulation of compact alleles due to random assortment from heterozygous loci. The tall palms arose from the same process, resulting in the loss of compact alleles in some gametes.

The self-pollinated descendants of the OCP are poor in yield components and have a high percentage of abnormal and disease-susceptible palms. Further breeding efforts at this level cannot be expected to accomplish much, agronomically speaking; however, this program could result in the maximum genetic preservation of compact growth characteristics. This is visualized as a very small program whose only goal is to preserve the source of this oil palm growth variant.

The Bx-1 generation refers to those progenies resulting from a cross between the original compact palm and an E. guineensis palm with normal growth. On the average, 50 percent of the genetic background of these palms originates from the original compact palm. As such, the palms at this genetic distance from the OCP do not excel in agronomic features or disease resistance and show a fairly high percentage of abnormal palms. On the other hand, the Bx-1 palms contain the compact genes in a relatively undiluted form, and by selecting the best recombinants, good compact growth can be recovered.

A series of about 20 elite Bx-1 generation compacts were selected in 1986 ( or "had been selected by 1986 "). Initially, two of these palms were used extensively in compact breeding (C288:20T, C333:122T). Other compact elites have been used in crosses more recently and 23 are undergoing cloning.

Two progenies (C9245, C9248), representing the F 1 of the Bx-1 generation, have been evaluated. C9245 resulted from the self-pollination of an elite Bx-1 compact, and C9248 resulted from a cross between two elite Bx-1 compacts. As had occurred in the selfed progeny of the OCP, super-compact palms also appeared in both F 1 progenies. The percentage of abnormal palms was high in the F 1 s and severe susceptibility to boron deficiency was common.

Compact palms at this level (Bx-1) were planted in semi-commercial trials in the three divisions in 1991. While the Bx-1 is distinctly inferior to more advanced generations in agronomic characteristics, the Bx-1 had a lower probability of losing the genes which contribute to the compact growth habit. Future improvements at the Bx-1 level will probably be easier in terms of growth but difficult in terms of yield and bunch composition.

Beginning with trials planted in 1985, compact palms of the Bx-2 generation were evaluated. These palms were produced by backcrossing two elite Bx-1 generation palms with normal guineensis palms. These progenies showed much better yield and bunch composition and fewer abnormalities than was characteristic of the Bx-1 progenies. Three of the best progenies were :

• C9236 CAM236:31T     x  C288:20T

• C9252 (Deli x AVROS) x   C288:20T

• C9269 CAM236:31T     x    C333:122T

Those progenies involving CAM236 had especially good bunch composition.

Since the best progenies of the Bx-2 generation are agronomically very good, the main concern is recovering an optimal level of compact growth and, especially, ensuring uniform growth among the progeny.

Elite individual palms of the Bx-2 generation are being cloned in order to establish clone trials. Crossing programs are also underway with elite Bx-2 compacts.

Three sub-programs for the development of Bx-2 compacts are underway. These involve making crosses of putative Bx-2 x Bx-2 combinations, along with the selfing of the parental palms for future seed production purposes.

It is yet to be determined whether compact palms at planting densities of about 180 palms per hectare or super-compact palms with considerably higher planting densities will result in the better yield, and more importantly the higher profit. If super-compacts can be correctly identified at the nursery stage, this important question will be able to be answered from evaluations of Bx-2 x Bx-2 progenies.

The Bx-3 generation of compacts is genetically removed from the OCP. At this level only one-eighth of the genetic code of the progeny is derived from the OCP. The retention of the genes controlling the compact growth habit depends on selection of the Bx-2 elites for growth. This is the case since we know that the Bx-1 elites used had a sufficiently complete complement of compact genes as to yield super-compacts upon selfing.

In the Bx-3 generation, we expect excellent agronomic characteristics. The evaluation of the Bx-3 progenies will help determine which parental palms are agronomically adequate for seed production. As a further requirement, the selfing of the putative parental Bx-2 palm used for producing the Bx-3 generation must segregate for super-compact palms. It is these super-compact dura or pisifera individuals which must be used as parental palms in seed production if progeny with near-uniform growth are to be obtained.

Although they carry certain defects such as excessive growth and low oil to mesocarp, the oil palm type "Deli dura" is used almost universally as a female parent because they are unique palms with outstanding fruit traits. None of the African duras available match the bunch yield and the fruit quality of Deli duras.

On the other hand, male parents used in producing oil palm seeds have been more diverse.

Some well-known male parents are :

Palm Line Program

2/2311T Ekona Unilever

32.3005T Calabar NIFOR

L2T La Me IRHO

SP 540 Dj'ongo AVROS

Derivatives of these 4 widely-used male parents are found in the Coto program. Although other promising pollen sources are also available, because of the proven value of these four sources, they should probably receive priority attention

Current oil palm breeding programs are focussed on producing planting materials for high oil yield . It is generally recognized that narrowness of gene-pool has been the major obstacle for rapid breeding and selection progress. ASD attempts to broad the genetic diversity of the current planting materials offering at least four of the most important and world wide distributed commercially oil palm planting materials as Deli x AVROS, Deli x Ekona, Deli x La Me and Deli x Calabar).

Drought tolerance in commercial oil palm hybrids is probably the most universally sought-after of the special characteristics.

Other useful characteristics are :

• resistance to "Lethal Spear Rot"

• tolerance to cool temperatures

• oil composition with high unsaturation

The preservation of the E. oleifera collection is also necessary for future use.

Contrasting behavior of drought-tolerant genotypes (left) and very susceptible Deli x AVROS (right) under growing conditions of severe drought

The policy has been to produce the best seeds possible in light of the most up-to-date knowledge available.

The seeds which are produced by ASD's Palm Research Program come from 3797 trees selected from among 16,000 Deli dura palms. They were chosen on the basis of bunch yields recorded over many years, bunch composition data obtained by the standard method of bunch analysis, and growth characteristics. The selected trees (approximately 25%) were all above of the mean for the plot in which they were found with respect to the characteristic examined, and for one or more of the traits they were outstanding. Pisifera x AVROS (80), Calabar (36), La Me (59) and Ekona (60) have been selected on the basis of progeny-testing results.

The pollination is controlled in order to insure the genetic identity in the crosses produced. The actual seed production potential approaches 38 million seeds a year.