GENETIC IMPROVEMENT OF HAWAII’S PREMIER
HARDWOOD, ACACIA KOA ²²

by
X. Shi and J.L. Brewbaker ²³

Koa (Acacia koa A.Gray) is an endemic leguminous tree of the Hawaiian Islands. It is one of the most prestigious timber species in Hawaii, best known for its exceptionally fine wood. The wood has excellent working properties, ranges widely in rich colors (yellow to deep purple), and often shows exceptional figures of grain. Koa is legendary and symbolic in Hawaiian culture. It was famous for building canoes and surfboards, and is now widely used for instruments, crafts, furniture, and utensils (Whitesell, 1990; Sun, 1996). Genetic improvement is deemed essential for expanded use in plantations of this highly variable tree (Brewbaker, 1996).

INTRODUCTION

Koa is one of 1 200 species in the large leguminous genus, Acacia, a member of the thornless, phyllodinous subgenus Heterophyllum (Guinet and Vassal, 1978; Whitesell, 1990). It was originally classified into three species and several varieties by different authors, but only the species A. koa is now recognized
(Wagner et al., 1990). Koa is a tetraploid, with a chromosome number of 2n=52 (Atchison, 1948; Carr, 1978; Shi, 2003). No variation in chromosome structure and number within the species has been found.

Koa is an important component of Hawaii’s ecosystems. It is one of Hawaii’s few endemic nitrogen-fixing species, and nurtures soils through its rhizobial root system and leaf littering. It is commonly associated with ohia (Metrosideros polymorpha). Koa forests have protected Hawaii’s watersheds historically, and they are important habitats for many insects and birds (Whitesell, 1990). The tree once thrived on all Hawaiian Islands, but populations were reduced drastically by animal grazing, fire, land clearing, logging and insect pests. Between 1963 and 1978, stock volume of commercial koa dropped from 34.1 to 7.1 x 10⁵ m³ (Nelson and Wheeler, 1963; Metcalf et al., 1978). Few “superior” trees can be found in wild koa forests today (Brewbaker et al., 1991).

Koa forests presently occur on the major Hawaiian Islands - Kauai, Oahu, Molokai, Maui, Hawaii - at longitudes of 154^o to 160^o W and latitudes of 19^o to 22^o N. It grows on volcanic soils of all geologic ages and degrees of development (Whitesell, 1990), but appears to perform poorly on highly acid soils. Koa is concentrated at elevations between 600 to 1800 m, and thrives in a belt between 800 and 1600 m
(Whitesell, 1990; Sun, 1996). Koa wilt is implicated in the failure to survive at lower elevations (Shi, 2003). Most koa is found in high rainfall areas receiving 1900 to 5100 mm rainfall annually, and does not thrive in drier areas.

Germplasm collection

Germplasm collections of koa were initiated by Brewbaker and his students at the University of Hawaii in the 1960s and have continued to the present time. Accessions were usually collected from single trees, occasionally as bulked seeds from several mother trees. Since 1990, the collection has been expanded by former UH (University of Hawaii) student Nicolas Dudley, forester for HARC (Hawaii Agriculture Research Center), whose collection is focused on low-elevation provenances.

The germplasm collection is summarized in Table 1. It currently comprises 547 accessions housed by the two institutions, including some advanced-seed accessions from outstanding trees or families in performance trials. All accessions have been germinated and transplanted into trials on the islands of Oahu and Hawaii as the basis for evaluations and genetic improvement.

Table 1. Koa germplasm collection

Institution	Island	Number of families
UH	Oahu	29
	Kauai	132
	Hawaii	75
	Maui	18
	Advanced seeds	100
HARC	Most from Oahu and Kauai	75
Total		547

HARC (Hawaii Agriculture Research Center)

UH (University of Hawaii)

Genetic variation

Phenotypic variations are great among koa trees in natural populations, where the majority of trees are of poor form. However, these represent a complex of genetic and environmental influences. Large numbers of feral ruminant animals and pigs have damaged Hawaii’s natural ecosystems over the past two centuries, and koa is a highly digestible legume. Many koa populations have been damaged by hurricanes and tropical storms, and others affected by soil acidity, waterlogging, plant competition and drought. Many traits vary widely in these natural populations (Daehler et al., 1999).

Clones occur in koa as a result of root-sprouting, however, and these often display genetic differences in tree form and morphology. Our replicated trials of seedlings derived from single mother trees provide the best evidence of genetic diversity. Genetic differences in these trials are evident in bole form, fluting and limbiness, growth rate (diameter and height), phyllode and pod shapes, seed size and arrangement, wood properties and resistance to koa wilt (Sun, 1996; Shi, 2003). Wood color and figure vary widely among individual koa trees, including the highly valued “fiddleback” figure. One variety, koaia, is not arboreal but shrubby, and reportedly has denser wood than other koas (Felling, pers. commun.).

Isozymic surveys of koa populations also revealed genetic variations in enzyme loci (Brewbaker, 1977; Conkle, 1996). Brewbaker noted variation among esterase and peroxidase loci intimating that outcrossing was the breeding system, while the six isozymic loci studied by Conkle had an average of three to seven alleles per locus. The calculated heterozygosity of these genes was 0.42, suggesting high genetic diversity. Self-pollinations by Sun (pers. commun.) also confirm that koa is largely or entirely outcrossed due to self-incompatibility.

Research on koa

Research on koa began in the nineteenth century, and focused on botany, ecology, forest management and reforestation (Whitesell, 1990). Later studies involved insect pests and diseases, phenotypic diversity and methods of propagation. With the erosion of koa resources, genetic improvement of koa became an issue of interest (Brewbaker, 1977; Brewbaker et al., 1991; Skolmen et al., 1991). Early koa plantations date back to 1910 on public and private lands. These were established from bulked, non-selected seeds from various sources (Skolmen et al., 1991). Very little genetic improvement of koa took place before the 1990s. An exception is the USDA Forest Service study that identified 52 “superior” trees with straight boles (Skolmen, 1977). A small progeny trial was planted at Keanaolu on the Island of Hawaii with seeds and clones derived from these trees (Skolmen et al., 1991). Genetic variations were not quantified, and no further selection took place (Sun, 1996).

Progeny trials

A genetic improvement programme for koa was started in 1990 based on seeds collected since the 1960s by Brewbaker and students. Since 1991, progeny trials have been planted annually at Hamakua (650 m elevation) on the Island of Hawaii. Trials usually consisted of about 50 families (half-sib seeds from single mother trees) in two replications of ten trees per rep planted at a spacing of 1 x 1.5 m. Four early trials were duplicated at Maunawili (160 m elevation) on the Island of Oahu, using the same experimental design, with more recent HARC trials in the region of Opaeula on Oahu. Two seed orchards at Hamakua have been established in recent years based on seeds of selected high-performing families. Plantings at Hamakua comprise 12 replicated yield trials (planted annually), one demonstration orchard, and two seed orchards. Trials were planted into herbicide-managed grass pastures, irrigated as needed and protected from weed competition for the year of establishment. Ruminant animals were excluded, but pig damage has occurred. Research has focused on: evaluation of growth rate, tree form and quality, survival rates and resistance to koa wilt and seed production of improved varieties based on selected families.

Progeny trials prove that koa is a fast-growing species under these experimental conditions, averaging 1.5 m height increment annually the first 5 years. Canopy closure could be achieved in six months. The narrow spacing we use (1 x 1.5 m) encouraged straighter boles and less forking than trees growing in open canopies. Average DBHs at age of four years ranged from 6.8 to 9.5 cm. Some families achieved heights >15 m and DBHs of >40 cm after 12 years.

Significant differences in growth occurred among families, with 90 percent of the families assessed as “genetic junk” (Brewbaker, 1996). Diversity within the half-sib families was often impressively small, suggesting some inbreeding within provenances. Genetic variance components for DBH ranged from 32.7 to 70.3 percent of total phenotypic variance for four trials at age of four years. High mortality caused by koa wilt disease precluded accurate estimations of heritability among older trees. Families from the Islands of Oahu and Kauai generally performed better than those from the Islands of Hawaii and Maui. However, superior families were identified from all Islands.

A seed orchard established in 1999 started flowering in 2002. This seed orchard consisted of single-tree plots of 15 superior families selected from early progeny trials. The seed from this seed orchard has been designated ‘UH Koa Comp 1’, and samples of seeds have been disseminated to foresters for evaluation of growth and resistance to koa wilt disease.

Breeding for resistance to koa wilt

Many pathogens and insects can impact on koa forests, but few are able to kill mature trees (Brewbaker et al., 1991). However, the koa wilt disease kills mature trees in natural forests and in plantations, and has raised deep concerns among foresters. This poorly understood disease injures trees of all size classes and occurs on all islands (Anderson, et al., 2002). Koa wilt was found by Gardner (1980) to be caused by the pathogen Fusarium oxysporum f. sp. koae. Early symptoms of the disease include leaf chlorosis and defoliation. Major branches or the entire crown of an affected tree can then wilt and die (Anderson, et al., 2002).

Koa wilt is epidemic in the progeny trials at Hamakua and Maunawili. Data from 11 of the annual trials planted at Hamakua are summarized in Table 2. Each year around 10 percent of trees in the trials die, largely due to wilt disease. Genetic variations in tolerance to the wilt have been evident. Certain families have shown exceptional tolerance to the disease in repeated plantings, while many do not survive. An almost linear regression occurs when survival is plotted against time (R² = 95 percent).

Table 2. Individual and average survival rates of the trials of Hamakua at different ages.

SET no.	# of family	Overall survival rate (%), at the age (years) of
		1	2	3	4	5	6	7	8	9	10	11	12
91-1	48	88.2	77.0	46.5	43.1	40.4	-	-	-	16.2	8.4	7.8	7.2
93-1	15	67.5	58.6	55.4	49.8	-	-	24.0	22.0	18.1
94-1	45	68.0	-	-	-	-	-	22.8	19.7	17.9
95-1	53	96.0	-	-	74.7	39.9	27.0	27.4	24.5
96-1	59	-	-	37.7	28.1	28.8	25.4	22.5
96-2	8	-	-	73.3	-	42.0
97-1	80	72.4	-	50.1	43.0	38.6	32.2
98-1	27	81.5	-	41.3	30.7	29.7
99-1	66	58.7	47.9	40.0	34.5
00-1	47	76.1	54.5	46.3
01-1	48	54.4	41.5
Avg.		73.6	61.8	48.8	43.4	36.6	28.2	24.2	22.1	17.4	8.4	7.8	7.2

Family selection for resistance to wilt disease has been conducted based on survival data of this type. Thirty families with the highest survival rates at the age of for or five years have been identified. They will be further examined for resistance at multiple locations in Hawaii, and incorporated into seed orchards.

Vegetative propagation

Early studies of vegetatively propagating koa met with limited success. Skolmen (1977) conducted an extensive study on vegetative propagation. Stem cuttings of juvenile seedlings and air-layers of juvenile shoots successfully produced roots, but no success was achieved from cuttings of older trees. Rooting ability varied from tree to tree, indicating the involvement of genetic factors. Grafting and root cuttings were not successful. Callus induction in tissue cultures was successful, and shoots could be regenerated from the callus (Skolmen, 1977). Only a few mature clones survived. Meristematic multiplication from young koa seedlings seemed promising in a study by Nagai and Ibrahim (1996), but also has not provided techniques suitable for large-scale cloning.

In recent studies of Shi (2003), propagation of koa cuttings was very successful from juvenile seedlings. Variations in rooting ability were observed among families tested. Rooting ability declined quickly as trees aged, most losing rooting ability when trees developed phyllodinous leaves. Cuttings failed to respond to auxin treatments. Etiolation treatments on young shoots of trees entering mature stage seemed very promising. Cuttings survived much longer, and rooting frequencies were increased in some families. Cuttings of a few mature trees from the Island of Hawaii rooted well after plant hormone treatments, confirming genetic variation in rooting ability. Vegetative propagation can play an important role in forest tree breeding (Zobel and Talbert, 1984). A successful vegetative propagation method could enable koa breeders to establish clonal plantations or to exploit self-sterility in the production of hybrid trees.

Future priorities

Early evaluation is desired for the koa composites derived from Cycle 1 selections from the UH and HARC trials. A high priority is to identify field resistance to koa wilt, a difficult and widespread pathogen of many crops. More research is needed on koa’s wood quality and its heritability, notably of the highly-figured koa prized by craftsmen. Proof is needed of the presence of a self-incompatible system in koa that could be the basis for hybrid tree production. Any exploitation of improved germplasm will rely on improved management methods, as unimproved plantations of koa often succumb to the stresses imposed by weeds, animals, diseases and insects. The high value of this prized tree and the devastating loss of high-value tropical hardwoods underscore the importance of this continued research.

REFERENCES

Anderson, R.C., Gardner, D.E., Daehler, C.C. & Meinzer, F.C. 2002. Dieback of Acacia koa in Hawaii: Ecological and pathological characteristics of affected stands. Forest Ecology and Management 162: 272-289.

Atchison, E. 1948. Studies on the Leguminosae. II. Cytogeography of Acacia (Tourn.) L. American Journal of Botany 35(10): 651-655.

Brewbaker, J.L. 1977. Acacia koa project. Final report. Cooperative Project 21-191. University of Hawaii and USDA Forest Service, Institute of Pacific Islands Forestry, Honolulu. p. 6.

Brewbaker, J.L. 1996. Genetic improvement, a sine qua non for the future of koa. p. 24-26In L. Ferentinos and D.O. Evans (eds.) Koa, a decade of growth. Proc. of the Hawaii Forest Industry Association 1996 Annual Symposium, Honolulu.

Brewbaker, J.L., Conrad, C.E., Shebata, S., Masaki, C., Simmons, P., Conkle, T. & Buck, M. 1991. Why koa improvement. p. 5-6. In J. L. Brewbaker, N. Glover, and E. Moore. (eds.) Improvement of Acacia koa: Resource documents. US Forest Service Institute of Pacific Islands Forestry, State of Hawaii Department of Land and Natural Resource, and Nitrogen Fixing Tree Association, Honolulu, Hawaii.

Carr, G.D. 1978. Chromosome numbers of Hawaiian flowering plants and the significance of cytology in selected taxa. American Journal of Botany 62(2): 236-242.

Conkle, M.T. 1996. Isozyme studies of genetic variability. p. 27-32. In L. Ferentinos and D.O. Evans (eds.) Koa, a decade of growth. Proceedings of the Hawaii Forest Industry Association 1996 Annual Symposium, Honolulu.

Daehler, C., Yorkston, M., Sun, W. & Dudley, N. 1999. Genetic variation in morphology and growth characters of Acacia koa in the Hawaiian Islands. International Journal Plant Sci. 160(4): 767-773.

Gardner, D.E. 1980. Acacia koa seedling wilt caused by Fusarium oxysporum f. sp. koae, f. sp. nov. (New taxa). Phytopathology 70(7): 594-597.

Guinet, P. & Vassal, J. 1978. Hypothesis of the differentiation of the major groups in the genus Acacia (Legumnosae). Kew Bulletin 32(3): 509-527.

Metcalf, M.E., Nelson, R.E., Petteys, E.Q.P. & Berger, J.M. 1978. Hawaii's timber resources-1970. USDA Forest Service, Resource Bulletin PSW-15. Pacific Southwest Forest and Range Experiment Station, Berkeley, CA. p. 20.

Nagai, C. & Ibrahim, R.. 1996. Clonal propagation of Acacia koa. p. 30-32. In L. Ferentinos and D.O. Evans (eds.) Koa: A decade of growth. Proceedings of the symposium. Hawaii Forest Industry Association 1996 Annual Symposium. Honolulu, Hawaii.

Nelson, R.E. & Wheeler, P.R. 1963. Forest resources in Hawaii. Hawaii Department of Land and Natural Resources in cooperation with U.S. Forest Service, Pacific Southwest Forest and Range Experiment Station, Honolulu, HI. p. 48.

Shi, X. 2003. Genetic improvement of Leucaena and Acacia koa Gray as high-value hardwoods. Ph.D. Dissertation. Department of Tropical Plant and Soil Sciences, University of Hawaii at Manoa, Honolulu.

Skolmen, R.G. 1977. Clonal propagation of Acacia koa. Ph. D. dissertation. University of Hawaii. Honolulu, Hawaii.

Skolmen, R.G., Horiuchi, H., Goo, D. & Masaki, C. 1991. Field trials, plantations, and arboreta. p. 19-21. In J.L. Brewbaker, N. Glover, and E. Moore (eds.) Improvement of Acacia koa: Resource documents. US Forest Service Institute of Pacific Islands Forestry, State of Hawaii Department of Land and Natural Resource, and Nitrogen Fixing Tree Association, Honolulu, Hawaii.

Sun, W. 1996. Genetic improvement of Leucaena and Acacia koa Gray. Ph. D. Dissertation. Department of Horticulture, University of Hawaii at Manoa, Honolulu.

Wagner, W.L., Herbst, D.R. & Sohmer, S.H. 1990. Manual of the flowering plants of Hawaii. Vol. 1. University of Hawaii Press, Bishop Museum, Honolulu, Hawaii.

Whitesell, C.D. 1990. Acacia koa Gray. p. 17-28. In R.M. Burns and B.H. Honkala (eds.) Silvics of North America. Vol. 2. Hardwood. USDA Forest Service Agricultural Handbook 654, Washington, D.C.

Zobel, B. & Talbert, J. 1984. Applied forest tree improvement. John Wiley and Sons.

²² RECEIVED FEBRUARY 2004.

²³ Department of Tropical Plant and Soil Sciences. College of Tropical Agriculture and Human Resources, University of Hawaii. 3190 Maile Way, Honolulu, HI 96822 USA